| *** vladimir3p has quit IRC | 00:01 | |
| annegentle | vidd: I think you can use keystone-manage commands to create EC2 creds - refer to lines 42 and 43 here: https://github.com/cloudbuilders/devstack/blob/master/files/keystone_data.sh | 00:01 |
|---|---|---|
| *** Bwong has quit IRC | 00:02 | |
| *** Telamon has quit IRC | 00:03 | |
| *** veepee has joined #openstack | 00:08 | |
| *** patelna has quit IRC | 00:09 | |
| *** robix has left #openstack | 00:09 | |
| *** stewart has joined #openstack | 00:12 | |
| *** ohnoimdead_ has joined #openstack | 00:15 | |
| *** dragondm has quit IRC | 00:17 | |
| *** ohnoimdead has quit IRC | 00:18 | |
| *** ohnoimdead_ is now known as ohnoimdead | 00:18 | |
| *** sdake has quit IRC | 00:18 | |
| *** veepee has quit IRC | 00:19 | |
| *** helfrez has quit IRC | 00:21 | |
| *** HowardRoark has joined #openstack | 00:24 | |
| vidd | annegentle, are you the one that writes the documentation on http://docs.openstack.org/ | 00:25 |
| vidd | ? | 00:25 |
| vidd | if so kudos !!!! and TY for your hard work | 00:27 |
| notmyname | vidd: annegentle is responsible for all the awesome parts of the documentation | 00:27 |
| notmyname | the rest of it was stuff we devs threw up there ;-) | 00:28 |
| vidd | notmyname, you have any part in the keystone app? | 00:28 |
| notmyname | nope, sorry. just swift | 00:28 |
| vidd | i dont have enough equiptment to even LOOK at swift =] | 00:29 |
| notmyname | heh. that's something I hope we make better in the future. swift needs to scale down better | 00:29 |
| vidd | trying to get a demo up and working so the ppl with the purse can give me some servers from this century | 00:30 |
| notmyname | but you can run it in one VM ;-) | 00:30 |
| vidd | but swift looks like something i definantly want | 00:30 |
| Glaceee | notmyname: Do you have a minute to take a look at this? weird 404 account error when I run swauth-prep ??http://pastebin.com/vRFPQrxi | 00:31 |
| notmyname | Glaceee: the key part is "could not create container". now as to the why.... ;-) | 00:32 |
| Glaceee | hehe ok :) | 00:32 |
| notmyname | vidd: I'm a big fan of swift :-) I think more people should use it | 00:32 |
| notmyname | Glaceee: what version of swift are you using? | 00:33 |
| Glaceee | diablo.. the latest | 00:33 |
| Glaceee | I downloaded swauth on github | 00:33 |
| notmyname | Glaceee: are you using swauth because you have need for it or.... (what's the use case? if this is just POC, then use tempauth) | 00:33 |
| *** nati2 has quit IRC | 00:33 | |
| vidd | Glaceee, diablo is NOT the latest...they are working on essex | 00:33 |
| *** nati2 has joined #openstack | 00:33 | |
| notmyname | vidd: it's the last release we had | 00:34 |
| notmyname | for swift, diablo == 1.4.3 | 00:34 |
| Glaceee | thats what i meant sorry | 00:34 |
| notmyname | we haven't released 1.4.4 yet | 00:34 |
| vidd | my bad | 00:34 |
| Glaceee | its a dev env notmyname | 00:34 |
| notmyname | and there haven't been any milestone releases for essex yet | 00:34 |
| *** sdake has joined #openstack | 00:35 | |
| Glaceee | yeah its POC.. i will try tempauth then | 00:35 |
| notmyname | Glaceee: ah. then I'd strongly recommend tempauth unless you have some particular need for swauth | 00:35 |
| *** pixelbeat has quit IRC | 00:35 | |
| Glaceee | will it work anyway with cyberduck? | 00:35 |
| notmyname | tempauth and swauth are functionaly the same (or very close). the biggest difference is that tempauth uses a config file and memcached and swauth uses swift itself for a backing store | 00:36 |
| Glaceee | oh ok.. thanks I didnt know that the new docs.. are really pointing to swauth or keystone | 00:36 |
| *** npmapn has quit IRC | 00:36 | |
| *** npmapn has joined #openstack | 00:37 | |
| vidd | everyone but nova wants a peice of keystone =] | 00:37 |
| notmyname | Glaceee: keystone == the future, but not ready yet. swauth == prod-ready, but not widely used. tempauth == very simple, perfect for small numbers of accounts in dev or POC environments | 00:38 |
| Glaceee | thanks | 00:39 |
| *** suntzupr has joined #openstack | 00:41 | |
| suntzupr | cloud? | 00:42 |
| matt_joyce | yes? | 00:43 |
| Glaceee | sun? | 00:43 |
| suntzupr | lol just joking. starting on openstack. | 00:43 |
| *** suntzupr has quit IRC | 00:46 | |
| Glaceee | will it be possible to have encprytion at rest with swift in the future? | 00:46 |
| vidd | is there some kind of plugin between euca and keystone? | 00:46 |
| notmyname | Glaceee: so tell me more about what you want there | 00:46 |
| notmyname | (I have my own thoughts) | 00:47 |
| Glaceee | well something like a sha-256 encryption at the file level.. you can launch an api command to ecrynpt a file lets say | 00:48 |
| *** PeteDaGuru1 has joined #openstack | 00:49 | |
| *** livemoon has joined #openstack | 00:49 | |
| notmyname | Glaceee: so I'm the swift deployer and you're my customer. what's the process look like? and what benefit are you gaining? | 00:50 |
| Glaceee | AES o meant | 00:50 |
| Glaceee | i meant | 00:50 |
| vidd | EaaS =] | 00:51 |
| vidd | Encryption as a service | 00:51 |
| Glaceee | notmyname.. Im not sold to the idea.. but i have been asked by potential customers :0 | 00:51 |
| *** PeteDaGuru has quit IRC | 00:52 | |
| notmyname | Glaceee: I think it's a marketing bullet point that doesn't do much (if anything) for real security ;-) | 00:52 |
| vidd | so something like the "encrypt your home directory" option in a linux install? | 00:52 |
| Glaceee | agreed.. but still requted.. :/ | 00:52 |
| *** adrianmoya has joined #openstack | 00:53 | |
| livemoon | hi, all | 00:53 |
| livemoon | morning | 00:53 |
| adrianmoya | Hi all, quick newbie question: Dashboard says Hypervisor: QEMU( lahf_lm, xtpr, cx16, tm2, est, vmx, ds_cpl, pbe, tm, ht, ss, acpi, ds). This machine is setup with libvirt_type=kvm. Is it OK or something is failng? should it say KVM? | 00:54 |
| Glaceee | lets say that I dont want raskspace or the governement opening my file | 00:55 |
| Glaceee | would that be a use case? | 00:55 |
| notmyname | Glaceee: if you (the customer) are asking me (the deployer) to encrypt the data, then you are either giving me the keys or trusting me to use my own keys. so, regardless of what I say, it's only the honestly of my employees that prevents me from reading the data | 00:55 |
| notmyname | (and that doesn't even get in to the key management issues on the deployer side) | 00:56 |
| *** koolhead17|zzZZZ is now known as koolhead17 | 00:56 | |
| vidd | livemoon, you have keystone working....are you using for nova too? | 00:56 |
| vidd | and if so, how did you do it? | 00:56 |
| Glaceee | hehe | 00:56 |
| notmyname | so if you want encrypted data stored in swift, encrypt it before you upload it | 00:56 |
| *** joeyang has joined #openstack | 00:56 | |
| *** dysinger has joined #openstack | 00:56 | |
| Glaceee | hehe yeah :) thats a good idea | 00:57 |
| notmyname | Glaceee: so, I think it's a marketing bullet point that doesn't give any real security ;-) | 00:57 |
| *** tsuzuki_ has joined #openstack | 00:58 | |
| Glaceee | hehe. no I agree.. never was a big fan.. but its the type of thing that a customer would ask | 00:58 |
| notmyname | so to answer you original question, I don't foresee at-rest encryption being added to swift any time soon | 00:58 |
| Glaceee | hehe | 00:58 |
| vidd | adrianmir, if kvm fails (not enabled in BIOS for example) then it automatically fails back to qemu | 01:00 |
| *** obino has quit IRC | 01:01 | |
| *** misheska has quit IRC | 01:02 | |
| adrianmoya | vidd, what could I check? is there a command I could run to see if everything is ok for kvm? | 01:02 |
| adrianmoya | I runned kvm-ok and it says INFO: /dev/kvm exists KVM acceleration can be used | 01:03 |
| vidd | did you run the devstack script? | 01:04 |
| *** worstadmin has quit IRC | 01:04 | |
| dysinger | I remember at at Swift Recon talk @ openstack boston I heard mention of collectd or rrd compat - somebody have more info for me ? | 01:05 |
| adrianmoya | yep, but to be honest, it failed early so I finished runnint it manually | 01:05 |
| vidd | i dont think the devstack script even checks if kvm is possible | 01:05 |
| *** mgius is now known as mgius-away | 01:06 | |
| vidd | but since it runs as a user, it might not be able to confirm kvm | 01:06 |
| uvirtbot | New bug: #878602 in openstack-dashboard "Settings "gear icon" image sprite broken on hover" [Undecided,In progress] https://launchpad.net/bugs/878602 | 01:06 |
| vidd | with a manual install, you can set nova-compute.conf --libvirt=kvm | 01:06 |
| *** ohnoimdead has quit IRC | 01:07 | |
| *** ohnoimdead has joined #openstack | 01:07 | |
| *** rsampaio has joined #openstack | 01:07 | |
| vidd | livemoon, you there? | 01:08 |
| adrianmoya | the script has a kvm section, but let me check carefully if there's a configuration overridding my nova.conf | 01:09 |
| *** rods has quit IRC | 01:11 | |
| *** doorlock has joined #openstack | 01:11 | |
| uvirtbot | New bug: #878603 in openstack-qa "/nova/rpc/impl_kombu.py" [Undecided,New] https://launchpad.net/bugs/878603 | 01:15 |
| adrianmoya | well it seems it working with kvm, it could be a dashboard bug | 01:16 |
| adrianmoya | I have two instances and two /usr/bin/kvm process running | 01:16 |
| vidd | adrianmoya, i'd say its morlikely a devstack script bug | 01:19 |
| *** sdake has quit IRC | 01:20 | |
| *** obino has joined #openstack | 01:21 | |
| vidd | annegentle, adding those fields to keystone gives me something to authenticate with, but i cant find a way to get euca to look at keystone for the credentials | 01:21 |
| *** miclorb_ has quit IRC | 01:22 | |
| adrianmoya | but it seems the server is using kvm... so it's a dashboard issue that it reports QEMU... | 01:22 |
| vidd | its dashboard-on-devstack-script | 01:22 |
| vidd | unless you actually set up dashboard on your own, you cant confirm its dashboard and not the way devstack makes dashboard that is causing the reporting issue | 01:24 |
| *** nati2_ has joined #openstack | 01:25 | |
| *** nati2 has quit IRC | 01:25 | |
| *** miclorb_ has joined #openstack | 01:26 | |
| adrianmoya | ok, I'll check a manual install and see how it goes... thanks again! | 01:29 |
| vidd | np | 01:29 |
| *** adrianmoya has quit IRC | 01:32 | |
| *** nati2_ has quit IRC | 01:32 | |
| *** sdake has joined #openstack | 01:33 | |
| *** jdurgin has quit IRC | 01:34 | |
| *** huslage has quit IRC | 01:35 | |
| *** Narayanan has joined #openstack | 01:35 | |
| Narayanan | hi, i was trying to find which hypervisor technology that openstack has the most support | 01:36 |
| Narayanan | is it kvm or xen (xcp) | 01:36 |
| rmk | kvm | 01:36 |
| vidd | kvm | 01:36 |
| Narayanan | when compared to vmware esx? | 01:37 |
| vidd | when compared to "all other options" | 01:38 |
| Narayanan | cool | 01:39 |
| Narayanan | what will lose if i decide xcp | 01:39 |
| vidd | kvm is built into the linux kernel...all others are extra | 01:39 |
| *** bencherian has quit IRC | 01:39 | |
| vidd | i dont think you'll "lose" anything...its just that kvm is more well documented | 01:40 |
| vidd | support usually means "we can help you with that" as opposed to "this function, that function, the other function" | 01:41 |
| *** bencherian has joined #openstack | 01:42 | |
| Narayanan | that also means this is very well tested as opposed to other platforms? | 01:42 |
| vidd | but....is there something xcp can do that kvm annot do? | 01:42 |
| Narayanan | anyone knows what is commonly used hypervisor with open stack? is it kvm still? | 01:43 |
| Narayanan | we want to go for something that open stack has good support and also widely used | 01:43 |
| vidd | kvm is widely used, tested to the gills (and then tested some more) and has excelent "we can help you with that" support | 01:44 |
| Narayanan | great. thanks | 01:45 |
| livemoon | hi, vidd | 01:45 |
| vidd | also, since its built into the linux kernel, you dont have to pay extra for it | 01:45 |
| vidd | livemoon, !!!! | 01:46 |
| Narayanan | right | 01:46 |
| Narayanan | :-) | 01:46 |
| vidd | livemoon, what did you build your stack on? *buntu? | 01:46 |
| livemoon | nova,glance,keystone,novaclient in production | 01:47 |
| livemoon | dashboard in test | 01:47 |
| livemoon | vidd: Have you create windows image? | 01:47 |
| vidd | right....but on centOS, redhat, *buntu, some-vesion-of-linux-noone-has-ever-heard-of....what? | 01:48 |
| *** doorlock has quit IRC | 01:48 | |
| vidd | and no...i havent created any images yet because i cant get keystone to talk to euca2ools | 01:49 |
| vidd | or i should say, euca2ools to authenticate via keystone | 01:49 |
| livemoon | I don't use euca2ools | 01:50 |
| livemoon | I just use novaclient | 01:50 |
| livemoon | if you make euca2ools via keystone, please tell me | 01:51 |
| *** Narayanan has left #openstack | 01:51 | |
| vidd | *sigh* nova client isnt in the ubuntu repos =\ | 01:53 |
| Glaceee | I found my problem notmyname.. my bad.. swift_cluster_default was not in the proxy.conf file | 01:54 |
| *** ewindisch_ has joined #openstack | 01:56 | |
| notmyname | Glaceee: whoops | 01:58 |
| *** ewindisch has quit IRC | 01:58 | |
| *** ewindisch_ is now known as ewindisch | 01:58 | |
| vidd | livemoon, you use novaclient to put images into glance? | 01:59 |
| *** bencherian has quit IRC | 02:01 | |
| Glaceee | :0 | 02:02 |
| *** hugokuo has joined #openstack | 02:02 | |
| livemoon | images only use glance | 02:10 |
| vidd | whereis keystone.middleware.ec2_token:EC2Token.factory | 02:12 |
| vidd | whoops | 02:12 |
| vidd | livemoon, can you pastebin me your api-paste.ini? | 02:16 |
| vidd | for nova | 02:16 |
| *** vernhart has joined #openstack | 02:17 | |
| *** kayyagari_ has joined #openstack | 02:17 | |
| *** kernelfreak has quit IRC | 02:17 | |
| *** ohnoimdead has quit IRC | 02:18 | |
| *** kayyagari_ has quit IRC | 02:26 | |
| vidd | nova wants to use nova-manage and euca2ools for everything and keystone for nothing | 02:33 |
| livemoon | vidd: http://paste.openstack.org/show/2785/ | 02:33 |
| vidd | same as i have | 02:36 |
| *** sdake has quit IRC | 02:36 | |
| vidd | how did you get images for glance to use? | 02:36 |
| vidd | did you make the novarc with the .pems? | 02:37 |
| livemoon | vidd: glance -A "999888777666" index | 02:38 |
| vidd | before i can index glance, i need images for glance to look at | 02:38 |
| vidd | those are in nova-objectstorage | 02:39 |
| vidd | and to get stuff in nova-objectstorage, i need euca2ools to work | 02:39 |
| vidd | or am i totally missing something? | 02:39 |
| livemoon | you can use glance upload images | 02:39 |
| *** vernhart has quit IRC | 02:39 | |
| livemoon | my novaclient have some error too | 02:40 |
| vidd | right...but how do i make the images to upload? | 02:40 |
| livemoon | use "glance add" | 02:40 |
| vidd | ok...i have "some cool server.iso | 02:41 |
| vidd | how do i get that into glance so glance can use it | 02:42 |
| *** praefect has quit IRC | 02:42 | |
| *** wariola has joined #openstack | 02:45 | |
| *** ewindisch_ has joined #openstack | 02:45 | |
| *** ewindisch has quit IRC | 02:48 | |
| *** ewindisch_ is now known as ewindisch | 02:48 | |
| livemoon | glance help | 02:58 |
| livemoon | hi, anyone know "nova list" error "KeyError: 'public'" | 02:59 |
| *** hadrian has quit IRC | 03:01 | |
| *** g2 has quit IRC | 03:03 | |
| *** tokuzfunpi has joined #openstack | 03:04 | |
| *** joeyang has quit IRC | 03:07 | |
| mjfork | full error? | 03:10 |
| *** thingee has joined #openstack | 03:12 | |
| *** Ryan_Lane has joined #openstack | 03:17 | |
| *** miclorb_ has quit IRC | 03:25 | |
| *** miclorb_ has joined #openstack | 03:26 | |
| *** supriya has joined #openstack | 03:31 | |
| livemoon | mjfork: http://paste.openstack.org/show/2786/ | 03:32 |
| livemoon | mjfork: it's my error log | 03:32 |
| mjfork | something is missing a pulblic address | 03:33 |
| mjfork | is this a trunk or releqase? | 03:33 |
| *** matiu_ has joined #openstack | 03:35 | |
| livemoon | what's public address? | 03:36 |
| livemoon | I use git install novaclient | 03:36 |
| mjfork | one of your servers | 03:36 |
| mjfork | if it is from git, something could be out of sync | 03:36 |
| mjfork | have you run nova-manage syncdb ot make sure it is up to date? | 03:36 |
| livemoon | I got it.thanks hugo | 03:39 |
| mjfork | what was it? | 03:40 |
| livemoon | need use --region_name | 03:40 |
| mjfork | ha, pretty sure i didn't help with that :-0 | 03:40 |
| *** obino has quit IRC | 03:43 | |
| vidd | so it looks like i need to find some way to to ceate the following: cacert.pm, pk.pem, cert.pm, and match them to my username and password info in keystone, and upload them into keystone =\ | 03:45 |
| vidd | oh...and get keystone to push all this data into my env Nova_ and env OS_AUTH | 03:47 |
| *** naehring has joined #openstack | 03:47 | |
| *** wariola has quit IRC | 03:48 | |
| *** joeyang has joined #openstack | 03:49 | |
| *** sdake has joined #openstack | 03:53 | |
| *** tokuzfunpi has quit IRC | 03:53 | |
| *** HowardRoark has quit IRC | 03:56 | |
| livemoon | mjfork: thank you anyway | 03:57 |
| *** wilmoore has joined #openstack | 04:00 | |
| jasona | don't suppose anyone here knows anything about bnt switches and lacp ? :-) | 04:00 |
| *** Vadim2 has joined #openstack | 04:03 | |
| *** corrigac_ has joined #openstack | 04:03 | |
| *** crayon has joined #openstack | 04:03 | |
| livemoon | I found nova cannot pause or suspend server? | 04:04 |
| *** llang629 has joined #openstack | 04:04 | |
| *** soosfarm_ has joined #openstack | 04:04 | |
| *** llang629 has left #openstack | 04:04 | |
| *** _cerberu` has joined #openstack | 04:04 | |
| *** vish1 has joined #openstack | 04:05 | |
| *** annegentle has quit IRC | 04:05 | |
| *** Vadim has quit IRC | 04:05 | |
| *** vishy has quit IRC | 04:05 | |
| *** soosfarm has quit IRC | 04:05 | |
| *** thingee has quit IRC | 04:05 | |
| *** axisys has quit IRC | 04:05 | |
| *** axisys has joined #openstack | 04:05 | |
| *** crayon_z has quit IRC | 04:05 | |
| *** _cerberus_ has quit IRC | 04:05 | |
| *** floehmann has quit IRC | 04:05 | |
| *** ogelbukh has quit IRC | 04:05 | |
| *** annegentle has joined #openstack | 04:05 | |
| *** jaypipes has quit IRC | 04:05 | |
| *** corrigac has quit IRC | 04:05 | |
| *** tr3buchet has quit IRC | 04:05 | |
| *** tr3buchet has joined #openstack | 04:05 | |
| *** deshantm has quit IRC | 04:05 | |
| *** ogelbukh has joined #openstack | 04:06 | |
| *** deshantm has joined #openstack | 04:06 | |
| *** floehmann has joined #openstack | 04:07 | |
| *** wilmoore has quit IRC | 04:16 | |
| *** jaypipes has joined #openstack | 04:18 | |
| *** perryh has joined #openstack | 04:22 | |
| *** Ryan_Lane has quit IRC | 04:25 | |
| *** joonwon has quit IRC | 04:27 | |
| livemoon | Is there a way that vm can run on not controller but node? | 04:28 |
| *** TreeStump has quit IRC | 04:29 | |
| *** TreeStump has joined #openstack | 04:30 | |
| *** koolhead17 has quit IRC | 04:31 | |
| vidd | livemoon, sure....on the controller, service nova-compute stop | 04:32 |
| vidd | then it has no choice but to run on the node | 04:32 |
| *** worstadmin has joined #openstack | 04:34 | |
| vidd | livemoon, sudo glance -A [keystone admin token] index | 04:35 |
| vidd | comes back.... "Not authorized to make this request. Check your credentials (OS_AUTH_USER, OS_AUTH_KEY, ...). | 04:35 |
| vidd | " | 04:35 |
| *** rsampaio has quit IRC | 04:45 | |
| *** reed has quit IRC | 04:45 | |
| *** fysa has quit IRC | 04:50 | |
| mjfork | livemoon: no problem | 04:51 |
| *** lionel_ has quit IRC | 04:53 | |
| *** perryh has quit IRC | 04:54 | |
| *** PeteDaGuru1 has quit IRC | 04:56 | |
| *** egant has quit IRC | 04:56 | |
| *** egant has joined #openstack | 04:57 | |
| *** bengrue has joined #openstack | 04:57 | |
| *** obino has joined #openstack | 04:57 | |
| *** fysa has joined #openstack | 04:58 | |
| *** lionel_ has joined #openstack | 04:58 | |
| *** tokuz has joined #openstack | 05:00 | |
| *** royh has quit IRC | 05:01 | |
| jasona | hmm | 05:02 |
| jasona | mjf: don't suppose you know anyone who knows a bit about bnt kit do you ? :) | 05:02 |
| *** royh has joined #openstack | 05:02 | |
| *** joeyang has quit IRC | 05:03 | |
| *** HugoKuo_ has joined #openstack | 05:04 | |
| *** hugokuo has quit IRC | 05:07 | |
| vidd | what kind of bnt kit? | 05:10 |
| vidd | jasona, like switches | 05:13 |
| *** ewindisch_ has joined #openstack | 05:14 | |
| *** ewindisch has quit IRC | 05:18 | |
| *** ewindisch_ is now known as ewindisch | 05:18 | |
| *** ewindisch has quit IRC | 05:21 | |
| *** ewindisch has joined #openstack | 05:22 | |
| *** sahuram has joined #openstack | 05:25 | |
| *** sahuram is now known as ramesh_hp | 05:25 | |
| *** _cerberu` is now known as _cerberus_ | 05:28 | |
| *** maplebed has quit IRC | 05:34 | |
| *** thingee has joined #openstack | 05:37 | |
| livemoon | hi | 05:43 |
| livemoon | vidd? | 05:43 |
| vidd | yes? | 05:44 |
| livemoon | do you know how to use volume via novaclient? | 05:44 |
| vidd | no | 05:44 |
| vidd | i'd like to know how you got keystone and glance to talk to each other | 05:45 |
| vidd | did you install keystone from repo's or from git? | 05:46 |
| livemoon | from git | 05:46 |
| vidd | maybe thats my issue then =\ | 05:47 |
| livemoon | no,I used to use repo's . it work well with glance | 05:47 |
| vidd | you on ubuntu server? | 05:48 |
| livemoon | yes | 05:50 |
| livemoon | 11.10 | 05:50 |
| vidd | me too...and i installed keysone and glance from ubuntu repos and they dont want to talk to each other | 05:51 |
| vidd | i think there is a file or two missing in the repo build | 05:52 |
| *** TheOsprey has joined #openstack | 05:52 | |
| *** bengrue has quit IRC | 05:54 | |
| *** joeyang has joined #openstack | 05:58 | |
| *** tokuz has quit IRC | 06:01 | |
| *** matiu_ has quit IRC | 06:07 | |
| *** kaigan_ has joined #openstack | 06:08 | |
| *** ramesh_hp has quit IRC | 06:08 | |
| *** stewart has quit IRC | 06:09 | |
| *** lionel_ has quit IRC | 06:09 | |
| *** lionel__ has joined #openstack | 06:09 | |
| *** bodepd has quit IRC | 06:11 | |
| *** bodepd has joined #openstack | 06:11 | |
| livemoon | vidd | 06:15 |
| livemoon | my blog: livemoon.dyndns.org | 06:15 |
| livemoon | It has keystone and glance conf file | 06:15 |
| *** hugokuo has joined #openstack | 06:17 | |
| *** HugoKuo_ has quit IRC | 06:17 | |
| *** guigui1 has joined #openstack | 06:18 | |
| vidd | interesting..... | 06:19 |
| vidd | your [pipeline:glance-registry] is set to "dont use keystone" | 06:20 |
| livemoon | really? | 06:23 |
| livemoon | pipeline = authtoken keystone_shim registryapp | 06:23 |
| vidd | just from what the config file says....but your stuff works | 06:23 |
| *** nerens has joined #openstack | 06:24 | |
| *** stewart has joined #openstack | 06:24 | |
| vidd | the registry | 06:24 |
| vidd | but perhaps you copied the correct lines from the examples packaged with the keystone git script | 06:25 |
| *** thingee has quit IRC | 06:28 | |
| *** nerens has quit IRC | 06:28 | |
| *** vidd_ has joined #openstack | 06:29 | |
| *** vidd_ has joined #openstack | 06:29 | |
| *** vidd has quit IRC | 06:30 | |
| *** vidd_ is now known as vidd | 06:31 | |
| *** nerens has joined #openstack | 06:31 | |
| *** mgoldmann has joined #openstack | 06:32 | |
| *** sahuram has joined #openstack | 06:33 | |
| uvirtbot | New bug: #878673 in openstack-qa "/nova/image/glance.py" [Undecided,New] https://launchpad.net/bugs/878673 | 06:36 |
| *** popux has joined #openstack | 06:36 | |
| *** jedi4ever has quit IRC | 06:39 | |
| *** andy-hk has quit IRC | 06:42 | |
| *** jedi4ever has joined #openstack | 06:44 | |
| *** joeyang has quit IRC | 06:51 | |
| *** joeyang has joined #openstack | 06:51 | |
| hugokuo | does any one success to keystone + SWIFT(SSL) + Cyberduck ? | 06:53 |
| uvirtbot | New bug: #878680 in openstack-qa "nova.scheduler.base_scheduler" [Undecided,New] https://launchpad.net/bugs/878680 | 06:56 |
| *** Oneiroi has joined #openstack | 07:04 | |
| *** Oneiroi has quit IRC | 07:04 | |
| *** Oneiroi has joined #openstack | 07:04 | |
| *** siwos has joined #openstack | 07:07 | |
| *** nerens has quit IRC | 07:11 | |
| *** reidrac has joined #openstack | 07:11 | |
| *** pradeep has joined #openstack | 07:21 | |
| *** bencherian has joined #openstack | 07:22 | |
| *** ewindisch_ has joined #openstack | 07:24 | |
| vidd | keystone + swift = DEMO ONLY according to keystone's README.md | 07:25 |
| Oneiroi | someone should tell rackspace then | 07:25 |
| Oneiroi | as it was on the training xD | 07:25 |
| *** supriya has quit IRC | 07:26 | |
| *** ewindisch has quit IRC | 07:26 | |
| *** ewindisch_ is now known as ewindisch | 07:26 | |
| vidd | i just installed it via the git script and read the README | 07:26 |
| vidd | perhaps there is a different version? | 07:26 |
| Oneiroi | ah yes git will always be the trunk aka the development version | 07:27 |
| Oneiroi | think we were using a deelopment milestone | 07:27 |
| *** miclorb_ has quit IRC | 07:27 | |
| Oneiroi | there's alaso diablo release milestones 1 through 4 iirc | 07:27 |
| vidd | the ubuntu repo wasnt working | 07:27 |
| *** osier has quit IRC | 07:29 | |
| *** nacx has joined #openstack | 07:31 | |
| *** twisla_ is now known as twisla | 07:32 | |
| *** sahuram has quit IRC | 07:42 | |
| *** bencherian has quit IRC | 07:44 | |
| *** halfss has joined #openstack | 07:47 | |
| halfss | hi is anybody resize sucess? | 07:48 |
| *** vdo has joined #openstack | 07:49 | |
| *** popux has quit IRC | 07:49 | |
| *** uksysadmin has joined #openstack | 07:49 | |
| halfss | does resize an instance need the instance id shutdown? | 07:50 |
| *** dobber has joined #openstack | 07:57 | |
| *** hingo has joined #openstack | 07:58 | |
| *** koolhead17 has joined #openstack | 07:59 | |
| *** osier has joined #openstack | 07:59 | |
| siwos | halfss: I guess you need to shut it down to resize the filesystem | 08:02 |
| *** osier has quit IRC | 08:04 | |
| *** pradeep has left #openstack | 08:04 | |
| halfss | siwos:you sure resize should shutdown the instance? | 08:05 |
| siwos | halfss: it definitely needs if you want to shrink the partition | 08:06 |
| halfss | is that xen can resize instance with the instance is live? | 08:06 |
| siwos | don't know about xen | 08:06 |
| siwos | (googlin) | 08:06 |
| halfss | siwos:how about if i don't shrink the partition? | 08:07 |
| Oneiroi | anyone got nova running with percona btw? | 08:08 |
| siwos | you can extend the fs with resize2fs | 08:08 |
| siwos | the proper way to do it is to fsck your drive first | 08:09 |
| zykes- | Oneiroi: isn't that easy ? | 08:10 |
| *** FabriceB has joined #openstack | 08:10 | |
| zykes- | siwos: do you have swift? | 08:10 |
| halfss | oh, i want make sure wheather shoud showdown the instance when i want resize it? | 08:10 |
| siwos | zykes-: it's being configured by my colleague | 08:10 |
| zykes- | are you going to setup keystone with it and use it with glance ? | 08:10 |
| *** halfss has quit IRC | 08:12 | |
| *** halfss has joined #openstack | 08:12 | |
| Oneiroi | zykes-: don't know I've seen it not fulfil dependancy issues on rpm installs because of naming differences | 08:12 |
| Oneiroi | but imho it's mysql so should just work | 08:12 |
| siwos | I guess yes ;-) but still did not dive into this | 08:12 |
| halfss | siwos: oh ,thank you | 08:12 |
| halfss | :) | 08:12 |
| siwos | afaik there's also possibility to use sparse images with xen | 08:13 |
| siwos | but I'm not an expert on that - I use kvm | 08:13 |
| zykes- | Oneiroi: we use Percona for a publication solution and it works straigt with mysql connection drivers (sqlalchemy) for example | 08:13 |
| Oneiroi | indeed I think it must just be broken rpm .spec requirements | 08:14 |
| Oneiroi | but I'm rolling oneiric now anyway xD will see how I get on | 08:14 |
| zykes- | ;) | 08:14 |
| zykes- | i'm still hanging at swift / glance - can't upload images to glance when using swift as the backend, glance gives me a swift auth erorr | 08:14 |
| halfss | here is my conf , i user swift store glance | 08:17 |
| halfss | default_store = swift | 08:17 |
| halfss | swift_store_auth_address =192.168.1.101:8080/auth/v1.0 | 08:17 |
| halfss | swift_store_user = system:root | 08:17 |
| halfss | swift_store_key = testpass | 08:17 |
| halfss | swift_store_container = glance | 08:17 |
| halfss | swift_store_create_container_on_put = true | 08:17 |
| halfss | swift_enable_snet = False | 08:17 |
| vidd | halfss, please use paste | 08:17 |
| vidd | http://paste.openstack.org | 08:18 |
| *** uksysadmin has quit IRC | 08:18 | |
| zykes- | halfss: wat is the system:root ? <user><tenant> ? | 08:18 |
| vidd | you paste a page like that and my notify pop=up takes up half my screen =\ | 08:18 |
| zykes- | vidd: you on a android or ? :p | 08:19 |
| halfss | swift'user | 08:19 |
| zykes- | not following you there | 08:19 |
| *** sahuram has joined #openstack | 08:19 | |
| zykes- | i use it with keystone, so it's a bit different | 08:19 |
| vidd | no....im half blind....800x600 resolution | 08:20 |
| halfss | oh | 08:20 |
| vidd | 15 inch monitor =] | 08:20 |
| siwos | now I have a question ;-) | 08:21 |
| *** osier has joined #openstack | 08:21 | |
| siwos | I'm still confused about two ways to upload custom images | 08:21 |
| siwos | using glance-upload vs euca-upload-bundle | 08:22 |
| *** nerens has joined #openstack | 08:22 | |
| siwos | euca-upload-bundle does not seem to interact with glance | 08:22 |
| vidd | it doesnt | 08:22 |
| siwos | anyone knows how to make euca-upload-bundle work with glance? | 08:22 |
| vidd | i thought the whole point of glance was to not have to use euca since euca doesnt work with keystone =\ | 08:23 |
| *** pixelbeat has joined #openstack | 08:23 | |
| vidd | siwos, you can try setting glance's destination folder to the same place as euca is putting stuff | 08:25 |
| vidd | but i dont know if it will register with glance that it's there | 08:25 |
| siwos | vidd: thanks for clarifying things ;-) | 08:25 |
| siwos | I'll stick to glance at the moment | 08:25 |
| siwos | I was just confused | 08:26 |
| *** osier has quit IRC | 08:26 | |
| vidd | also, glance API listens on 9292, and registry listens on 9191 | 08:26 |
| siwos | yes | 08:26 |
| halfss | is anybody resize sucess ? | 08:26 |
| zykes- | hmmm, is the newest glance client broken i wonder :| | 08:27 |
| vidd | if you somehow get euca to tell 9191 that "hey...new stuff over here" glance might be able to use it | 08:27 |
| *** worstadmin has quit IRC | 08:27 | |
| *** worstadmin has joined #openstack | 08:27 | |
| *** nerens has quit IRC | 08:28 | |
| *** katkee has joined #openstack | 08:31 | |
| *** nerens has joined #openstack | 08:31 | |
| *** supriya has joined #openstack | 08:31 | |
| livemoon | hi, siwos | 08:34 |
| halfss | resize don't not need the instance is shutdown :Resize (ie, migrate) a running instance. | 08:34 |
| livemoon | so bad, I delete my disk which keystone installed on that | 08:34 |
| siwos | hi | 08:35 |
| siwos | livemoon: pitty | 08:35 |
| livemoon | what are doing these days? | 08:35 |
| *** sahuram_ has joined #openstack | 08:36 | |
| *** timr1 has joined #openstack | 08:37 | |
| *** sahuram has quit IRC | 08:37 | |
| siwos | not much on openstack | 08:37 |
| zykes- | halfss: i take it you are not using keystone with swift? | 08:37 |
| livemoon | I have not found novaclient can control nova-volume. Is it? | 08:38 |
| siwos | I keep working on cluster automation | 08:38 |
| *** osier has joined #openstack | 08:38 | |
| livemoon | cluster of red hat? | 08:38 |
| halfss | zykes： is there any wrong ? my work well | 08:38 |
| zykes- | halfss: are you using keystone ? | 08:38 |
| halfss | zykes: i just use swift store glance's imgae. | 08:39 |
| halfss | when i use keystone manage swift, does't work well | 08:39 |
| zykes- | hence my point ;) | 08:40 |
| *** benner has joined #openstack | 08:42 | |
| siwos | livemoon: openstack cluster automation | 08:44 |
| siwos | testing modules for puppet and chef | 08:44 |
| siwos | none of them working out of the box though ;-) | 08:44 |
| *** AhmedSoliman has joined #openstack | 08:45 | |
| siwos | something different gusy | 08:49 |
| siwos | guys | 08:49 |
| siwos | I highly reccommend you a tool called boxgrinder | 08:49 |
| siwos | you can use it to create custom AMIs | 08:50 |
| siwos | it's easy as hell ;-) | 08:50 |
| Oneiroi | ok wtf ... nova-network installed, but" `restart nova-network` gives: unknown instance | 08:50 |
| Oneiroi | :S | 08:50 |
| siwos | http://boxgrinder.org/ | 08:50 |
| Oneiroi | +1 boxgrinder | 08:50 |
| Oneiroi | I wrote some yamls for it that got upstreamed :) | 08:51 |
| siwos | Oneiroi: +1 ;-) | 08:51 |
| siwos | waiting for ubuntu templates ;-) | 08:51 |
| Oneiroi | yeh I'll write one for openstack | 08:51 |
| mgoldmann | someone would need to write an ubuntu plugin first :) | 08:51 |
| Oneiroi | as soon as I have my manual oneiric working and know what's to script | 08:51 |
| mgoldmann | morning! | 08:51 |
| Oneiroi | mgoldmann: hello there you sneaking in :D | 08:52 |
| mgoldmann | always :) | 08:52 |
| vidd | Oneiroi, try "service nova-network restart" | 08:52 |
| mgoldmann | siwos: thanks for kind words about BG | 08:52 |
| Oneiroi | you may be glad to know I was "pimping" boxgrinder at the openstack training to everyone :P | 08:52 |
| mgoldmann | awesome! :) | 08:52 |
| mgoldmann | and I have an openstack plugin written locally | 08:52 |
| mgoldmann | :) | 08:52 |
| Oneiroi | gief! | 08:52 |
| Oneiroi | :D | 08:52 |
| mgoldmann | it pushes the images to glance, nothing more yet | 08:53 |
| Oneiroi | vidd: unkown instance :( | 08:53 |
| siwos | mgoldmann: - you're a Polish guy ;-) | 08:53 |
| vidd | then try start | 08:53 |
| siwos | me too | 08:53 |
| mgoldmann | heh :) | 08:53 |
| siwos | Poznan area | 08:53 |
| Oneiroi | vidd: *facepalm* | 08:53 |
| Oneiroi | I'm sucha tit | 08:53 |
| Oneiroi | xD | 08:53 |
| *** uksysadmin has joined #openstack | 08:54 | |
| mgoldmann | siwos: I'm currently from Wrocław | 08:54 |
| vidd | heh...what time of day is it for you and how long have you been up? | 08:54 |
| * vidd is working on his 3 barrel of coffee | 08:54 | |
| Oneiroi | vidd: 09:54 on my first coffee ;-) | 08:54 |
| vidd | *3rd | 08:54 |
| Oneiroi | ~3hrs sleep | 08:54 |
| Oneiroi | crap ... yeh nova-network keeps bombing out | 08:55 |
| vidd | sleep...that's the animal wool comes from .... right? | 08:55 |
| Oneiroi | lol | 08:55 |
| Oneiroi | yup forgot the db sync ... do nova-* is crapping out | 08:55 |
| * Oneiroi facepalm | 08:55 | |
| vidd | if nova-network keeps dropping out on you then you probably have a network config issue | 08:56 |
| Oneiroi | vidd: table nova.networks does not exist ;-) | 08:57 |
| vidd | its 4:56 am and ive been up since 8:55 am | 08:57 |
| Oneiroi | mgoldmann: I'm going to write some BG yamls to build openstack on oneiric, unelss someone allready has / is ? | 08:57 |
| vidd | but i did have an amazing 4 hours of rest | 08:57 |
| mgoldmann | Oneiroi: feel free to throw anything you have on me :) | 08:58 |
| Oneiroi | vidd: sleep is inefficient and overrated ;-) | 08:58 |
| livemoon | siwos: what is openstack cluster? | 08:58 |
| vidd | yeah...wool makes me itch anyway | 08:58 |
| Oneiroi | mgoldmann: I have a few windows machines I could throw at you if you really wanted :P | 08:58 |
| mgoldmann | too heavy | 08:59 |
| mgoldmann | :) | 08:59 |
| siwos | LOL | 08:59 |
| Oneiroi | livemoon: an openstack deploy using more than just one machine | 08:59 |
| vidd | windows machines will run linux...throw them my way =] | 08:59 |
| livemoon | Oneiroi: I know ,but mutil nodes does not mean cluster | 09:00 |
| Oneiroi | vidd: indeed >) recycling 2 vistaquad core desktops into a lab now | 09:00 |
| vidd | mgoldmann, once you strip out the proprietary stuff, those machines would be lighter =] | 09:00 |
| livemoon | siwos: chep, do you have some tech docs about it? | 09:00 |
| Oneiroi | livemoon: a nova one = cluster, i.e. 5 compute nodes 1 controller etc ... | 09:01 |
| vidd | ohhhhh quad-cores | 09:01 |
| Oneiroi | nove zone* | 09:01 |
| siwos | we plan 100 nodes | 09:01 |
| siwos | (puppet) | 09:01 |
| Oneiroi | vidd: my first build is with recycled kit :P looking at 16 cores ~40GB ram 1.83TB RAID10 sata ~300GB RAID 10 15K SAS :D | 09:01 |
| livemoon | siwos: how many controller? | 09:01 |
| Oneiroi | siwos: someone has suggested I look @ juju | 09:02 |
| livemoon | glance, keystone, dashboard ? | 09:02 |
| vidd | Oneiroi, i have to get two systems up with the dashboard so the purse-holders can send some purchase orders my way | 09:03 |
| livemoon | siwos: Is each node running compute and volume? | 09:03 |
| siwos | brb | 09:03 |
| *** mayfield has quit IRC | 09:03 | |
| Oneiroi | livemoon: each worker node should be running compute +_ network at the very least | 09:03 |
| vidd | Oneiroi, network is not needed if you use vlan | 09:04 |
| Oneiroi | vidd: if you don't have network on each compute node, network is routed through the one node that is | 09:04 |
| halfss | yes network is not needed | 09:04 |
| Oneiroi | according to the training material | 09:04 |
| Oneiroi | thus bottleneck + single failiure point | 09:05 |
| siwos | Oneiroi: brb | 09:05 |
| halfss | when use multi-host,you shoud need network on each compute node | 09:05 |
| vidd | ive found it helpful to have rabbitqm on all machines | 09:05 |
| *** koolhead17 has quit IRC | 09:05 | |
| livemoon | Oneiroi: I know, I want to know dose siwos want to running volume in nodes? | 09:05 |
| Oneiroi | ah | 09:06 |
| Oneiroi | vidd: indeed, anythign you can replicate do so :) so that if you were say to drop kick one machine the whole cluster doesn't die :) | 09:07 |
| vidd | Oneiroi, from what i understand, if you tell the nodes give this vm this address, give that vm that address" and you have the gateways set to your main gateway, there is less bottleneck | 09:07 |
| vidd | oh, and less node overhead | 09:08 |
| Oneiroi | In theory if you deploy it into an existing vlan + range then yes it _should_ use the native network equipment, but out of the box deploys switch to HAFlatDHCP manager, and run network on every compute node at the very least :) | 09:09 |
| vidd | no way | 09:10 |
| vidd | flatdhcp is a nightmare | 09:10 |
| Oneiroi | just going on what was on the course | 09:11 |
| Oneiroi | there is probably a better manager now | 09:11 |
| Oneiroi | esp with Quantum coming | 09:11 |
| vidd | troubleshooting a hiccup with flatdhcp has got to be a headache and a half | 09:11 |
| siwos | livemoon: I've tested the setup you mentioned | 09:11 |
| siwos | but finally we are buying dedicated hardware for nova-volume service | 09:12 |
| Oneiroi | siwos: you going to sit nova-volume on swift? or just use as standalone? (not using nova-volume yet myself) | 09:12 |
| zykes- | siwos: what company you work for ? | 09:12 |
| vidd | Oneiroi, "out-of-the-box" is usually the worst "solution" =] | 09:12 |
| Oneiroi | vidd: agreed | 09:12 |
| *** supriya has quit IRC | 09:12 | |
| vidd | look at windows for a prime example =] | 09:13 |
| Oneiroi | even not otb, windows is still <insert reference to excrement here> | 09:13 |
| vidd | idk...my no longer out of the box windows has a penquin on it =] | 09:14 |
| Oneiroi | windows 8 though ... from what I can tell pulls a lot from the linux kernel, which is why MS upstreamed so much recently ... though I'm not convinced it's still windows | 09:14 |
| Oneiroi | vidd: therefor it's a linux pc and not windows ;-) | 09:14 |
| halfss | how did you monitor instance ? | 09:16 |
| *** eferenr has joined #openstack | 09:16 | |
| Oneiroi | snmp + zenoss / nagios personally | 09:16 |
| vidd | ive used every version of windows since 3.11 and they all are little more than malware magnets | 09:16 |
| Oneiroi | _but_ there are a variety to choose from | 09:16 |
| Oneiroi | vidd: good to practising my pentesting on, that's all I use it for ;-) | 09:17 |
| vidd | how do you remove a user with keystone-manage? | 09:17 |
| zykes- | Oneiroi: why do you need nagios + | 09:17 |
| Oneiroi | nothing in keystone-manage --help ? | 09:17 |
| zykes- | when you got zenoss | 09:18 |
| Oneiroi | zykes-: I said zenoss / nagios, depenedant on personal preference, either or :) | 09:18 |
| Oneiroi | I'm using Zenoss | 09:18 |
| vidd | no...keystone-manage --help is about as useful as DOS | 09:18 |
| Oneiroi | lol | 09:18 |
| zykes- | ah | 09:19 |
| Oneiroi | vidd: keystone-manager user delete username | 09:19 |
| eferenr | i have a dummy question regarding glance, but i do not find documentation about it. Once you have Glance Authentication with Keystone, how do u authenticate (e.g. which headers should I use in the rest call)? thx! | 09:19 |
| Oneiroi | keystone-manage* | 09:19 |
| Oneiroi | eferenr: auth with keystone and get your token, that token can now be used to auth with glance | 09:19 |
| Oneiroi | as for headers ... X-Auth-Token I think | 09:20 |
| Oneiroi | check the API docs @ http://wiki.openstack.org | 09:20 |
| eferenr | ok, I thought there was a way to specify user and pasw and glance automatically request the keystone token.... | 09:20 |
| vidd | eferenr, im having a hard enough time just getting glance and keystone to talk to each other | 09:20 |
| Oneiroi | vidd: bugger https://bugs.launchpad.net/keystone/+bug/877504 | 09:20 |
| vidd | eferenr, http://docs.openstack.org/diablo/openstack-identity/admin/content/curl-examples.html | 09:21 |
| eferenr | thx vidd, but I am already familiar with KeyStone calls, I was looking for a glance http call where you specify user and pasw. and it automatically requests for a keystone token and uses it in the next requests | 09:23 |
| vidd | oh... | 09:24 |
| vidd | eferenr, what you do is....use curl to get your token | 09:24 |
| vidd | then glance -H [glance host] -A [token] [the rest of your glance request] | 09:25 |
| livemoon | how to control volume | 09:26 |
| eferenr | I got it, and then I'm trying something like curl -X GET -H "X-Auth-Token: $TOKEN" http://localhost:9292/v1/images | 09:26 |
| eferenr | ok vidd, I'm gonna try first the cli | 09:26 |
| eferenr | and see if it works, thx ;-) | 09:26 |
| vidd | ive been trying to get it to work for the last 16 hours | 09:26 |
| eferenr | yes.... it's not very intuitive... and the lack of documentation doesn't help | 09:28 |
| *** javiF has joined #openstack | 09:28 | |
| vidd | these plenty of documentation for glance =] | 09:28 |
| *** osier has quit IRC | 09:30 | |
| halfss | is there anybody resize sucess? | 09:34 |
| *** Oneiroi has quit IRC | 09:35 | |
| *** Oneiroi has joined #openstack | 09:35 | |
| *** koolhead17 has joined #openstack | 09:37 | |
| *** mus has quit IRC | 09:37 | |
| koolhead17 | hi all | 09:37 |
| livemoon | hi,koolhead17 | 09:38 |
| livemoon | you are so late | 09:38 |
| koolhead17 | livemoon: yeah. slept late my friend. how have you been | 09:39 |
| *** shaon has joined #openstack | 09:42 | |
| livemoon | I am so tired | 09:42 |
| livemoon | work 14 hours everyday | 09:42 |
| livemoon | Do you use volume, koolhead17? | 09:42 |
| zykes- | has anyone tested swift with newest keystone? | 09:43 |
| koolhead17 | livemoon: zykes- nopes | 09:44 |
| * koolhead17 will grab some lunch and come back | 09:44 | |
| Oneiroi | brb head's burried in oneiric | 09:46 |
| *** Razique has joined #openstack | 09:47 | |
| *** osier has joined #openstack | 09:47 | |
| *** nerens has quit IRC | 09:47 | |
| Daviey | Hola, has anyone seen http://pb.daviey.com/ynK1/ before ? | 09:49 |
| Daviey | leaves instances stuck in pending. | 09:49 |
| Daviey | Restarted rabbitmq, with no joy. | 09:49 |
| vidd | Daviey, is this for a node or the controller? | 09:50 |
| Daviey | vidd: good question.. | 09:52 |
| *** nerens has joined #openstack | 09:52 | |
| vidd | i suspect its for a node | 09:52 |
| *** npmapn has quit IRC | 09:52 | |
| *** osier has quit IRC | 09:52 | |
| *** Oneiroi_ has joined #openstack | 09:54 | |
| Daviey | I'm wondering if it is an sqlalchemy issue, rather than messaging. | 09:54 |
| vidd | its a quota issue | 09:54 |
| vidd | you have a quota of 10 | 09:54 |
| vidd | there are 10 things waiting | 09:54 |
| vidd | your trying # 11 | 09:55 |
| Daviey | yes, but the queue doesn't seem to reducing | 09:55 |
| *** Oneiroi has quit IRC | 09:55 | |
| *** Oneiroi_ is now known as Oneiroi | 09:55 | |
| vidd | what is it your trying to do? | 09:56 |
| vidd | i dont know what nova.rpc is | 09:56 |
| Daviey | vidd: It's a in-use cloud where instances are being stuck in pending | 09:57 |
| Daviey | Not a new deployment. | 09:57 |
| halfss | is there anybody resize sucess? | 09:57 |
| vidd | do you have rabbitMQ installed on all machines? | 09:58 |
| Daviey | vidd: I don't have access to the machines.. not sure. :/ | 09:58 |
| zykes- | http://paste.openstack.org/show/2787/ anyone seen that before ? | 09:58 |
| vidd | how many jobs does this tenent have active? | 09:59 |
| Oneiroi | looks to me like someones trying to use https over http | 09:59 |
| Oneiroi | resulting in "bad request" | 09:59 |
| zykes- | Oneiroi: yeah, i fixed that by enabling ssl in the swift proxy | 10:00 |
| zykes- | but then a new one: http://paste.openstack.org/show/2788/ | 10:00 |
| zykes- | if i use the credentials with swift -A http://keystone.stack.lan:5000/v1.0 -U glance -K glance_pw stat -v it works fine | 10:01 |
| vidd | zykes-, the error is that you are tying to NOT authenticate | 10:02 |
| vidd | thats why when you do, it works | 10:02 |
| vidd | ive been battling with keystone for hours....im VERY familiar with the 401 error | 10:03 |
| zykes- | eh, vidd i've set the credentials in glance-api.conf the same ones in that command... | 10:03 |
| zykes- | yet another glance bug or ? | 10:03 |
| zykes- | or keystone | 10:03 |
| vidd | the way i understand it....the auth key in the config file is for things like dashboard | 10:04 |
| vidd | not for CLI use | 10:05 |
| *** nerens has quit IRC | 10:05 | |
| zykes- | vidd: in swift -K it's your password... | 10:05 |
| zykes- | and in the config it's the password for the user of glance in swift | 10:05 |
| vidd | in glance config? | 10:06 |
| *** nerens has joined #openstack | 10:06 | |
| zykes- | yeah, when you use swift backing you need to give it a user to login to swift with ? | 10:06 |
| zykes- | how else should it auth ;) | 10:06 |
| vidd | right | 10:06 |
| vidd | swift -A http://keystone.stack.lan:5000/v1.0 -U glance -K glance_pw stat -v is fine and swift -A http://keystone.stack.lan:5000/v1.0 -U glance stat -v fails | 10:07 |
| vidd | right? | 10:07 |
| zykes- | no vidd, using 'glance' with 'swift' as a backend fails | 10:08 |
| zykes- | when I do glance -A <token> name=test < <image file> | 10:08 |
| vidd | is swift set up with keystone? | 10:08 |
| *** osier has joined #openstack | 10:09 | |
| zykes- | yes, that's what i'm trying to say | 10:09 |
| zykes- | glance api > keystone OK | 10:10 |
| zykes- | swift proxy > keystone OK | 10:10 |
| zykes- | glance api > swift proxy > keystone != compute | 10:10 |
| vidd | right.... | 10:10 |
| vidd | it needs to do glance>keystone>swift | 10:10 |
| *** fmancinelli has joined #openstack | 10:11 | |
| halfss | zykes:you use dashboard mangage swift suecss? | 10:11 |
| zykes- | no | 10:11 |
| halfss | zyeks:oh | 10:12 |
| vidd | your first 2 ok's were actually glance>keystone>glance and swift>keystone>swift | 10:12 |
| zykes- | glance client to glance api with keystone is ok | 10:13 |
| zykes- | swift client to swift proxy with keystone auth is ok | 10:13 |
| zykes- | but using a glance backed by swift is negative | 10:13 |
| vidd | actually....its glace client to keystone to glance api | 10:13 |
| *** osier has quit IRC | 10:13 | |
| halfss | zykes: you can see api.log | 10:13 |
| zykes- | halfss: nothing exciting there going on | 10:14 |
| *** osier has joined #openstack | 10:14 | |
| halfss | ? what you config in glance-api.conf | 10:14 |
| vidd | try this: glance add -A [glance auth key] -H [swift destination] | 10:15 |
| halfss | do not use -A | 10:15 |
| vidd | fine --auth_key[glance auth key] | 10:16 |
| halfss | just like this glance add name="ubuntu1104" is_public=true disk_format=raw < ubuntu-11.04-server-amd64.iso | 10:16 |
| zykes- | halfss: I can't do that... swift auth remember ? | 10:16 |
| halfss | you use swift store image,don't need use keystone | 10:16 |
| zykes- | ehm, glance auth | 10:16 |
| halfss | ? | 10:16 |
| zykes- | i can't just do a glance add with no token | 10:17 |
| *** uksysadmin has quit IRC | 10:17 | |
| vidd | halfss, his keystone is fine | 10:17 |
| Razique | zykes-: | 10:17 |
| Razique | glance with keystone ok :) | 10:17 |
| zykes- | nor can i do -H in glance towards swift, that doesn't make any sense | 10:17 |
| vidd | why not? thats where you want it to go isnt it? | 10:17 |
| Razique | http://paste.openstack.org/show/2789/ | 10:17 |
| zykes- | vidd: no, the host needs to be your glance api server... | 10:18 |
| zykes- | what I'm telling you is "glance -A <token> name=ubuntu < ubuntu-image.tar.gz" gives an auth failure with the glance client or glance api | 10:18 |
| halfss | ? | 10:19 |
| vidd | so your have the same issue im having | 10:19 |
| halfss | no | 10:19 |
| zykes- | vidd: and that is ? | 10:19 |
| koolhead17 | Vadim2: hey man | 10:19 |
| vidd | glance add -H localhost -A [authkey] name=xxx < ~/image.tar | 10:20 |
| vidd | ive been battling it since monday | 10:21 |
| zykes- | and further more the -A auth key is the auth key for your admin account, not the one that's used towards swift | 10:22 |
| halfss | http://hi.baidu.com/vdaos/blog/item/3bcc0436bc47e2abd1a2d309.html here is my process when use swift store image, but is chinese | 10:22 |
| halfss | :> | 10:22 |
| vidd | the auth key is whatever auth key your supposed to use from where your ising it | 10:22 |
| siwos | hey koolhead17! | 10:22 |
| *** timr has joined #openstack | 10:22 | |
| vidd | if your the admin, you can use the admin key | 10:23 |
| siwos | guys - I still try to figure out which puppet module is better for automated openstack deployments | 10:23 |
| vidd | if your a tech, you use the tech key | 10:23 |
| vidd | and so on | 10:23 |
| siwos | anyone using openstack with puppet? | 10:23 |
| siwos | you | 10:23 |
| halfss | i config swift as here http://swift.openstack.org/howto_installmultinode.html | 10:23 |
| siwos | ve got 2 repos in general | 10:23 |
| zykes- | vidd: yeah, but still the token in -A doesn't necassarely have anything to do with the stuff in glance-api.conf | 10:23 |
| siwos | one from cloudbuilders team and one from puppetlabs team | 10:23 |
| zykes- | the stuff in glance-api.conf can be totally different | 10:24 |
| siwos | don't know which is better | 10:24 |
| vidd | right | 10:24 |
| Oneiroi | ok found out why my lab hardware isn't booting vms .... the intel quad chips (Q8300) have 2 versions, one with VT one without, take a guess which ones I have ... | 10:24 |
| zykes- | the ones without ? ;) | 10:24 |
| zykes- | halfss: yes, that's without keystone | 10:24 |
| Oneiroi | indeed :( | 10:24 |
| vidd | it can be any key that keystone says your endpoint is allowed to use | 10:24 |
| halfss | en | 10:24 |
| halfss | oh | 10:24 |
| *** timr has left #openstack | 10:24 | |
| zykes- | * | 10:24 |
| zykes- | *hears a tada* | 10:25 |
| zykes- | i wonder if there might be anything todo with the glance swift module | 10:25 |
| *** joeyang has quit IRC | 10:26 | |
| vidd | it probably has something to do with the endpoints or endpoint template for glance in keystone | 10:26 |
| vidd | thats where im looking | 10:26 |
| *** naehring has quit IRC | 10:27 | |
| vidd | i dont have swift and i have the same issue | 10:27 |
| zykes- | vidd: i wouldn't think so since then i don't think the swift client should work ? | 10:27 |
| zykes- | vidd: what issie is that ? | 10:27 |
| Razique | what the nova-manage scrub command does exactly | 10:27 |
| Razique | I mean I know it removes a network from a project before removing the project itsefl but | 10:27 |
| zykes- | if you're on a newer glance vidd then do: $cmd endpointTemplates add RegionOne glance http://glance.stack.lan:9292/v1.0/files http://glance.stack.lan:9292/v1.0/files http://127.0.0.1:9292/v1.0/files 1 0 something like this? | 10:28 |
| vidd | if the endpoint for swift is correct and the endpoint for glance has a typo, then its quite possible | 10:28 |
| Razique | does it only allow me to revmoe a network or does the scrubing deletes other stuff ? | 10:28 |
| Razique | ok… il also remove security groups | 10:28 |
| Razique | good to know | 10:28 |
| *** nerens_ has joined #openstack | 10:28 | |
| zykes- | vidd: my glance index with admin token works fine | 10:28 |
| *** nerens has quit IRC | 10:28 | |
| *** nerens_ is now known as nerens | 10:28 | |
| zykes- | so i believe there is something going wrong in between | 10:29 |
| *** worstadmin has quit IRC | 10:29 | |
| vidd | if your index is working then perhaps the syntax your using is off | 10:29 |
| vidd | can you paste exactly what your using [dummy info for keys of cource]? | 10:30 |
| *** nikolaij has joined #openstack | 10:30 | |
| zykes- | i wonder if the swift.py in glance isn't updated for use with keystone ? | 10:31 |
| *** perestrelka has quit IRC | 10:32 | |
| vidd | it should authenticate with keystone before it even bothers to see where its supposed to put stuff | 10:32 |
| nikolaij | I everybody, good morning. i have a cloud running with 1 controller and one node. when i run an instance on a node, i receive "wget: can't connect to remote host (169.254.169.254): Network is unreachable". how can i fix it? | 10:32 |
| *** naehring has joined #openstack | 10:33 | |
| vidd | zykes-, do you have default_sore = swift in glance-api? | 10:34 |
| nikolaij | and before that i can see: "Sending discover..." which fails... please help | 10:34 |
| zykes- | yes | 10:35 |
| uvirtbot | New bug: #878759 in keystone "better manual page with description needed for keystone." [Undecided,New] https://launchpad.net/bugs/878759 | 10:35 |
| vidd | and you have all that authentication crap that keysone replaced commented out? | 10:36 |
| zykes- | oh | 10:37 |
| zykes- | amazing | 10:37 |
| *** nerens has quit IRC | 10:37 | |
| zykes- | if i set the swift auth url to http://keystone:5000/v1.0 | 10:37 |
| zykes- | it works | 10:37 |
| vidd | right...because it goes glance>keystone>swift | 10:38 |
| zykes- | yeh | 10:38 |
| *** BuZZ-T has quit IRC | 10:39 | |
| vidd | so you good? | 10:39 |
| zykes- | dunno yet, checking stuff | 10:39 |
| vidd | hehe | 10:39 |
| vidd | ive gotten quite good at this pathing stuff trying to find my error =] | 10:40 |
| *** BuZZ-T has joined #openstack | 10:40 | |
| *** perestrelka has joined #openstack | 10:41 | |
| zykes- | should be documented that if you're using swift you need to go to keystone first if you have it ;p | 10:41 |
| vidd | theres ALOT that should be documented =] | 10:42 |
| zykes- | *notes down* | 10:42 |
| zykes- | i'm doing a small setup now but going larger later i hope | 10:43 |
| vidd | im doing a demo of two junk machines | 10:43 |
| vidd | when i pull this off i hope to get funding for two brand new servers | 10:44 |
| *** Razique has quit IRC | 10:44 | |
| zykes- | ;) | 10:46 |
| zykes- | i could always rend a server for 200$ a month with 4 * 2*6 cores | 10:46 |
| zykes- | but that would mean my $ | 10:47 |
| zykes- | :( | 10:47 |
| vidd | drop the entire database AGAIN and add from scratch | 10:48 |
| vidd | i have the thing mapped out this time so i know what is going where =] | 10:49 |
| zykes- | i made a .sh script with all the keystone-manage commands, got tired of having to go in and change endpoints ;p | 10:49 |
| nikolaij | i thought this was a support channel | 10:52 |
| vidd | this is a support channel | 10:53 |
| nikolaij | but nobody gives me support never. | 10:54 |
| *** halfss has quit IRC | 10:54 | |
| nikolaij | i do not want to be the complainer of the situation, i know people has their own business and stuff to do, but this is not the first time i join, and nobody ever helped me :( | 10:55 |
| vidd | 169.254.169.254 is an ip address owned by amazon | 10:55 |
| nikolaij | i thought that was an internal ip in the cloud to retrieve metadata. isn't it? | 10:56 |
| *** perestrelka has quit IRC | 10:57 | |
| *** kernelfreak has joined #openstack | 10:57 | |
| vidd | i just tried to access it fom my home computer and i cant get to it | 10:57 |
| nikolaij | and actually that range /16 is reserved, | 10:57 |
| nikolaij | like 192.168 | 10:58 |
| nikolaij | when an instance is booting, it retrieves some info from the controller at that address | 10:58 |
| nikolaij | afaik | 10:58 |
| FabriceB | http://en.wikipedia.org/wiki/Private_network#Link-local_addresses | 10:59 |
| *** halfss has joined #openstack | 10:59 | |
| nikolaij | exactly | 10:59 |
| FabriceB | a never routed adress | 10:59 |
| livemoon | hi.all | 11:00 |
| vidd | nikolaij, that is the "my system cant find a dhcp server to get an address from" ip address | 11:00 |
| vidd | i KNEW it looked familiar =] | 11:00 |
| nikolaij | maybe this can clarify: in my compute-node into openstack, i have 2 NICs one for private and 1 for public. when i run an instance, then br100 is created but no address is assigned. is that correct behavior? | 11:01 |
| vidd | no, that is not correct behavior | 11:01 |
| nikolaij | ok thanks, so i already know something is wrong. | 11:02 |
| vidd | you should have your system set up so that local ips are assigned to you VMs | 11:02 |
| nikolaij | the strange thing is that if i run the image from the controller itself, then everything works | 11:02 |
| *** ksteward has joined #openstack | 11:02 | |
| vidd | well yeah....because the controller has nova-network installed on it so it created the proper local ip for the vm | 11:03 |
| vidd | or at least an ip that is routable | 11:03 |
| nikolaij | my setup is: controller (all nova components) eth0 192.168.56.2 eth1 10.0.3.15 br0 10.0.0.1 while the nova compute machine has br100 no ip, 192.168.56.3 eth1 10.0.3.15 i just followed http://cssoss.wordpress.com/2011/04/27/openstack-beginners-guide-for-ubuntu-11-04-installation-and-configuration/ and everything worked. while on part 2 i had the abouve issues | 11:04 |
| vidd | that is wrong | 11:05 |
| zykes- | now just to get up nova ;) | 11:06 |
| nikolaij | which config should be good? i tried to manually assign to br100 the same ip that the controller has (10.0.0.1) just for test, but of course does not work (2 ips same net) | 11:06 |
| *** gohko_ has quit IRC | 11:07 | |
| nikolaij | my nova.conf looks like this (excuse my verbosity) --daemonize=1 | 11:08 |
| nikolaij | --dhcpbridge_flagfile=/etc/nova/nova.conf | 11:08 |
| nikolaij | --dhcpbridge=/usr/bin/nova-dhcpbridge | 11:08 |
| nikolaij | --logdir=/var/log/nova | 11:08 |
| nikolaij | --state_path=/var/lib/nova | 11:08 |
| nikolaij | --verbose | 11:08 |
| nikolaij | --libvirt_type=qemu | 11:08 |
| nikolaij | --sql_connection=mysql://root:nova@192.168.56.2/nova | 11:08 |
| nikolaij | --s3_host=192.168.56.2 | 11:08 |
| nikolaij | --rabbit_host=192.168.56.2 | 11:08 |
| nikolaij | --ec2_host=192.168.56.2 | 11:08 |
| nikolaij | --ec2_url=http://192.168.56.2:8773/services/Cloud | 11:08 |
| nikolaij | --fixed_range=10.0.0.0/8 | 11:08 |
| nikolaij | --network_size=64 | 11:08 |
| nikolaij | --num_networks=1 | 11:08 |
| nikolaij | --FAKE_subdomain=ec2 | 11:08 |
| nikolaij | --public_interface=eth0 | 11:08 |
| nikolaij | --state_path=/var/lib/nova | 11:08 |
| nikolaij | --lock_path=/var/lock/nova | 11:08 |
| *** kaigan_ has quit IRC | 11:08 | |
| nikolaij | --glance_api_servers=192.168.56.2:9292 | 11:08 |
| nikolaij | --image_service=nova.image.glance.GlanceImageService | 11:08 |
| vidd | nikolaij, use paste | 11:08 |
| livemoon | nikolaij: you should use pastebin | 11:08 |
| nikolaij | ok sorry i will next time | 11:08 |
| GheRivero | ec2_url is deprecated | 11:09 |
| GheRivero | the same as fake_subdomain | 11:10 |
| *** tsuzuki_ has quit IRC | 11:10 | |
| nikolaij | actually seems that i'm not the only one with this problem: https://answers.launchpad.net/nova/+question/152089 but it does not fix for me | 11:10 |
| nikolaij | i'm removing deprecated flags, and try again to run | 11:11 |
| vidd | have you read the cookbook? | 11:11 |
| nikolaij | you mean docs.openstack.org? | 11:12 |
| *** gohko_ has joined #openstack | 11:13 | |
| vidd | no...im looking for the URL | 11:13 |
| nikolaij | tnx | 11:13 |
| nikolaij | also: how can i know the release of openstack i'm running? | 11:13 |
| livemoon | --version | 11:14 |
| livemoon | nikolaij, form china? | 11:14 |
| nikolaij | lol, nope | 11:15 |
| nikolaij | no livemoon i'm from italy, living in holland | 11:15 |
| vidd | nikolaij, http://cssoss.files.wordpress.com/2011/10/openstackbookv2-0_csscorp.pdf | 11:15 |
| *** ahasenack has joined #openstack | 11:15 | |
| vidd | nothing but net there | 11:15 |
| nikolaij | tnx a lot | 11:16 |
| *** Razique has joined #openstack | 11:16 | |
| vidd | no keystone, no dashboard, | 11:16 |
| vidd | i dont think it has swift either | 11:16 |
| nikolaij | my version btw is 2011.3 (2011.3-nova-milestone-tarball:tarmac-20110922115702-k9nkvxqzhj130av2) | 11:17 |
| vidd | wow...it does have swift | 11:17 |
| livemoon | bye all. Go home | 11:17 |
| nikolaij | bye livemoon tnx so far | 11:19 |
| *** naehring has quit IRC | 11:19 | |
| eferenr | Hi, when trying to integrate nova with keystone I'm getting the following error (when launching nova-api). Does anyone have any idea of the reason and how to solve it? Thx! http://pastebin.com/Fe9DR6Bh | 11:21 |
| *** naehring has joined #openstack | 11:22 | |
| vidd | eferenr, i had that one! | 11:22 |
| eferenr | and did u solve it? :-) | 11:22 |
| eferenr | I'll owe you a beer if u help me with that. hehe | 11:23 |
| vidd | yes i solved it....i just going through my bash history | 11:23 |
| vidd | two lines fixed it | 11:23 |
| vidd | 1.) sudo apt-get install python-pip | 11:26 |
| vidd | 2.) sudo pip install [missing file] | 11:26 |
| Razique | eferenr: what is your version of nova ? | 11:26 |
| vidd | in your case "auth" | 11:26 |
| livemoon | Razique: you are late today | 11:26 |
| *** sahuram_ has quit IRC | 11:27 | |
| eferenr | I'm gonna try it, vidd ;-). @Razique, I'm using the 2011.3 version tagged at github | 11:27 |
| *** AhmedSoliman has quit IRC | 11:27 | |
| *** orclev has joined #openstack | 11:27 | |
| eferenr | @vidd, seems that it does not work. Downloading/unpacking auth | 11:28 |
| larissa | eferenr: Error: "vidd," is not a valid command. | 11:28 |
| eferenr | Could not find any downloads that satisfy the requirement auth | 11:28 |
| eferenr | No distributions at all found for auth | 11:28 |
| eferenr | Storing complete log in /root/.pip/pip.log | 11:28 |
| livemoon | eferenr: you can not use auth module | 11:29 |
| livemoon | Razique: Have you used nova-volume in your work? | 11:29 |
| eferenr | then how should I proceed to integrate keystone? All I've done is to add the paste_config file in keystone... and I'm getting this "module named auth not found" error. | 11:30 |
| *** lts has joined #openstack | 11:31 | |
| livemoon | eferenr: you should use api-paste.ini in your nova directory | 11:32 |
| vidd | eferenr, what command did you try to run that caused that error? | 11:32 |
| eferenr | pip install auth | 11:32 |
| vidd | no...the error you originally had | 11:33 |
| eferenr | ./bin/nova-api | 11:33 |
| eferenr | when adding the api-paste.ini into the nova.conf file | 11:33 |
| livemoon | you paste your api-paste.ini | 11:33 |
| vidd | and what did you do b4 that? | 11:33 |
| eferenr | before that it was working, but without keystone authentication. livemoon I'm gonna try the paste file you say, instead of the one provided with keystone | 11:34 |
| livemoon | eferenr: actually, I have not use auth in my api-paste.ini conf file. | 11:35 |
| vidd | eferenr, you dont add anything to your nova.conf file | 11:35 |
| *** cereal_bars has joined #openstack | 11:36 | |
| vidd | eferenr, did you install nova from packages or from git? | 11:36 |
| eferenr | from the tar available in github | 11:37 |
| eferenr | the 2011.3 | 11:37 |
| vidd | then open the README ad follow the instruction there | 11:37 |
| vidd | the README in keystone | 11:38 |
| eferenr | livemoon: if I use the api-paste.ini file it works, but I presume it does not include KeyStone authentication | 11:38 |
| eferenr | vidd: I've followed such instructions | 11:38 |
| eferenr | but still, I get this "module named 'auth' not found" error. | 11:38 |
| vidd | thos instruction say to create a symbolic link and nothing else | 11:39 |
| livemoon | eferenr: I haven't use auth in api-paste.ini | 11:39 |
| livemoon | you can delete "auth" | 11:39 |
| *** perestrelka has joined #openstack | 11:40 | |
| *** AhmedSoliman has joined #openstack | 11:41 | |
| eferenr | livemoon: that's the api-paste.ini I'm using, what do you mean by deleting auth? http://pastebin.com/zgSk7qud | 11:41 |
| vidd | what does your nova.conf file look like? | 11:42 |
| *** nerens has joined #openstack | 11:42 | |
| *** markvoelker has joined #openstack | 11:44 | |
| eferenr | problem solved, it was a bug in the keystone/middleware/nova_keystone_context.py | 11:45 |
| eferenr | I'll check if it has been reported, otherwise I will add a ticket | 11:45 |
| livemoon | do you install keystone from apt? | 11:47 |
| eferenr | no, from github | 11:48 |
| vidd | livemoon, no, he got it from git | 11:48 |
| eferenr | I'm using the tagged version | 11:48 |
| eferenr | I supposed it was the most stable version | 11:49 |
| halfss | is there anybody resize success? | 11:50 |
| livemoon | bye all, I will go home. | 11:50 |
| *** livemoon has quit IRC | 11:50 | |
| eferenr | bye, I thx for the help ;-) | 11:50 |
| zykes- | how does nova do multiple networks ? like subnets etc and seperate traffic on them | 11:51 |
| *** ewindisch_ has joined #openstack | 11:53 | |
| vidd | like group one VM's are on 192.168.2.0/24, group two is on 192.168.3.0/24, group three is on .....and so on? | 11:53 |
| vidd | all from one compute node? | 11:53 |
| zykes- | the first | 11:54 |
| zykes- | vlan backed ? | 11:54 |
| *** ewindisch has quit IRC | 11:54 | |
| *** ewindisch_ is now known as ewindisch | 11:54 | |
| vidd | so each node has iths own subnet? | 11:54 |
| zykes- | nono, i mean how can i seperate networks ? | 11:55 |
| zykes- | say customer 1 and customer 2 ? | 11:55 |
| vidd | so cust 1 has *.1.0/24, cust 2 has *.2.0/24? | 11:56 |
| zykes- | for example | 11:57 |
| zykes- | not sure how one would do it | 11:57 |
| vidd | i suppose until Quantum is stable, you would have a seperate NIC for each subnet | 11:58 |
| vidd | pipe each subnet through its own eth port | 11:59 |
| vidd | and set the floating IPs based on who gets what subnet | 11:59 |
| vidd | Quantum is supposed to handle all that ove a single nic | 12:01 |
| vidd | you would definantly go with vlan | 12:01 |
| *** halfss has quit IRC | 12:02 | |
| zykes- | yeah, really looking forwards for openvswitch support | 12:04 |
| *** livemoon has joined #openstack | 12:07 | |
| livemoon | hi,all | 12:07 |
| shaon | livemoon: you are home already? :D | 12:07 |
| livemoon | undergroud railway | 12:08 |
| livemoon | shaon: hi | 12:09 |
| livemoon | I want to know how to use volume via novaclient. | 12:10 |
| *** livemoon_ has joined #openstack | 12:12 | |
| Razique | livemoon: hi :) | 12:13 |
| Razique | hi shaon | 12:13 |
| livemoon_ | Razique: hi | 12:13 |
| livemoon_ | I can use keystone | 12:14 |
| *** livemoon has quit IRC | 12:15 | |
| livemoon_ | but not know how to with euca2ools together | 12:15 |
| Razique | livemoon_: let me go back at desk | 12:15 |
| livemoon_ | Razique: :) | 12:16 |
| Razique | I've been able to make euca2ools work with keystone yesterday | 12:16 |
| Razique | i've all the notes there | 12:16 |
| *** Razique has left #openstack | 12:16 | |
| zykes- | Razique: submit a patch ? ;p | 12:16 |
| eferenr | ufff... that's the never ending story..... :/ hehehe does anyone have KeyStone + Nove + Glance up and running? I've manage to link these components (the authentication flow has changed and now it does not crash when running the different components). But I'm not able to authenticate myself in order to use the RESTful interface neither using X-Auth-Token: XXXXXX or by providing a User and Pasw. | 12:20 |
| eferenr | I have the feeling that something might be wrong with the KeyStone end-point ... or how Glance/Nova try to validate the provided token | 12:20 |
| eferenr | because I'm always getting the following response: This server could not verify that you are authorized to access the document you requested. Either you supplied the wrong credentials (e.g., bad password), or your browser does not understand how to supply the credentials required. | 12:21 |
| vidd | eferenr, did you set up your keystone database? | 12:22 |
| eferenr | yeps, I can send request to keystone | 12:23 |
| vidd | or did you just pump in the sample data | 12:23 |
| eferenr | and authenticate by sending request directly to it | 12:23 |
| eferenr | no no, I'm using my custom DB data | 12:23 |
| livemoon_ | eferenr:first, does glance can use with keystone? | 12:24 |
| vidd | did you use the keystone-manage db_setup or did you just throw up the database? | 12:24 |
| eferenr | I guess it tries to use it, but it cannot authenticate. Since I added keystone I'm always getting a 404 error | 12:24 |
| vidd | db_sync | 12:25 |
| eferenr | ehms.... I just called ./bin/sampledata | 12:25 |
| eferenr | but if I try to authenticate with my custom user it works (if I work direcly towards keystone) | 12:25 |
| livemoon_ | drop earlier database, create a new and run keystone, it will set up tables auto | 12:26 |
| eferenr | that's what I've done. I'm able to log in into KeyStone with my custom user | 12:26 |
| vidd | so the curl to your custom database works but not the keystone auth with apps? | 12:27 |
| livemoon_ | then use keystone-manage to add your data | 12:27 |
| eferenr | exactly vidd | 12:27 |
| eferenr | if I send request directly to KeyStone, it works | 12:27 |
| livemoon_ | which apps? | 12:27 |
| eferenr | glance | 12:27 |
| vidd | that means you never ran keystone-manage db_sync so keystone has no clue how to talk to its database | 12:27 |
| eferenr | or usign CURL towards nova-api // glance-api | 12:27 |
| eferenr | if I use this it works : | 12:28 |
| *** livemoon_ has quit IRC | 12:28 | |
| eferenr | curl -X POST -d '"auth":{"passwordCredentials":{...}"}' http://0.0.0.0:35357/v2.0/tokens | 12:28 |
| eferenr | and I get a token | 12:29 |
| vidd | eferenr, curl is the back-end | 12:29 |
| vidd | glance needs to use the front end | 12:29 |
| vidd | and keystone cant translate front end spanish into back-end german | 12:29 |
| *** Razique has joined #openstack | 12:29 | |
| vidd | does that help you to understand? | 12:30 |
| eferenr | :/ not at all.... I'm just using CURL to access the RESTful interface | 12:30 |
| vidd | right | 12:31 |
| koolhead17 | vidd:can you mention the issues you found on http://cssoss.files.wordpress.com/2011/10/openstackbookv2-0_csscorp.pdf will look at it and try to update it. You can comment on the blog page :) | 12:31 |
| eferenr | glance command should be just a CURL-like wrapper | 12:31 |
| vidd | but its NOT | 12:31 |
| vidd | glance uses a glance command | 12:31 |
| eferenr | the problem is that when I'm curling keystone, it seems to work (I get a token), but when I try to curl glance-api with the X-Auth-Token header it does not work | 12:31 |
| eferenr | but then, how do you use glance HTTP api? | 12:32 |
| vidd | it tells keystone this is what my password should be | 12:32 |
| vidd | keystone tries to read the database | 12:32 |
| eferenr | yeps, but if keystone is working, then glance index --host=$HOST --auth_token=$TOKEN should work | 12:32 |
| vidd | NO NO NO NO | 12:33 |
| eferenr | aha | 12:33 |
| vidd | keytstone is NOT!!!!!!!!!!!!!!!!!!!!!!!!!!! working.....CURL is working | 12:33 |
| eferenr | how is then that I'm getting tokens from keystone? | 12:34 |
| *** ziyadb has joined #openstack | 12:34 | |
| vidd | keystone WILL NEVER work unless you use keystone -manage to tell keystone this is the database, and keystone says this is how the database be be configured | 12:34 |
| vidd | you are NOT getting tokens from keystone | 12:34 |
| eferenr | who is sending me then the tokens? | 12:35 |
| vidd | you are getting tokens of the curl scripts against your database info | 12:35 |
| vidd | run that curl command against anything in your nova database and you will get a "token" | 12:36 |
| eferenr | when I try to curl glance-api I get an unauthorized message | 12:37 |
| vidd | ive been fighting with this keystone database for 3 days now....i know a thing or twop about iot | 12:37 |
| eferenr | I do not get any token | 12:37 |
| eferenr | hehe | 12:37 |
| eferenr | ls | 12:37 |
| vidd | delete the tables in your keystone database | 12:38 |
| vidd | run "keystone-manage db_sync" | 12:38 |
| vidd | do you have phpmyadmin installed on your machine? | 12:39 |
| eferenr | nops, but I have sqlitebrowser to analyze the db | 12:39 |
| vidd | fine....i suppose it is similar | 12:39 |
| vidd | look at your database structure | 12:40 |
| eferenr | sudo ./bin/keystone-manage db_sync | 12:40 |
| eferenr | db_sync is not a supported obj type | 12:40 |
| zykes- | why do you need to do a db_sync vidd ? | 12:40 |
| *** nhubbard_away is now known as nhubbard | 12:40 | |
| zykes- | eferenr: are you using sampledata or creating your own ? | 12:40 |
| eferenr | creating my own | 12:40 |
| zykes- | hmm | 12:41 |
| vidd | zykes-, you have to do the sync otherwize keystone wont understand it | 12:41 |
| vidd | i dunno why, its just that way | 12:41 |
| eferenr | when I try to sync I get this error: db_sync is not a supported obj type | 12:41 |
| Razique | vidd: could you sum up your issue to me briefly ? | 12:42 |
| zykes- | vidd: no, keystone doesbn't have a db_sync... | 12:42 |
| Razique | I just landed | 12:42 |
| Razique | zykes-: looks like it does actually, | 12:42 |
| vidd | i backed up my database and did a clean install, imported the database and keystone said "NOPE not using that....I didnt make it" | 12:42 |
| vidd | it does so | 12:42 |
| *** halfss has joined #openstack | 12:42 | |
| zykes- | vidd: keystone usually recreates the db on start | 12:43 |
| vidd | the "not a supported obj type" is the expected outcome of that command | 12:43 |
| vidd | zykes-, no it most certainly does NOT | 12:43 |
| vidd | if it recreates the database every time it starts, how could you possible store ppl's login info on it?!? | 12:44 |
| *** superbobry has joined #openstack | 12:45 | |
| zykes- | vidd: ehm, pardon doesn't recreate, it "creates" | 12:45 |
| vidd | zykes-, it READS the database every time it starts | 12:46 |
| zykes- | vidd: yeah, checking that all tables etc are there | 12:46 |
| vidd | right | 12:47 |
| zykes- | and if you do not have a database creates using sqlite it will create an empty one on start or when you run keystone-manage | 12:47 |
| zykes- | s/creates/it creates | 12:47 |
| vidd | and if you have one but never ran db_sync, keystone will have no idea how to read it | 12:48 |
| zykes- | dude, there is seriously no db_sync command i think | 12:48 |
| zykes- | if you look at the python keystone.backends or keystone/backends/__init__.py you see that it runs configure_backends which in terms sets up the db if needed | 12:49 |
| eferenr | I agree with zykes.... whatever you type outputs the same msg | 12:49 |
| vidd | http://docs.openstack.org/diablo/openstack-identity/admin/content/dependencies.html | 12:49 |
| eferenr | yeps, in the documentation the talk about such command | 12:49 |
| vidd | the error message is a know bug | 12:49 |
| eferenr | but when you type it, ti does not work | 12:49 |
| eferenr | ah | 12:49 |
| eferenr | oko | 12:49 |
| zykes- | vidd: the error message is not a "error" message in terms that there is no code to handle that command if you read through the code | 12:50 |
| vidd | i never said it was an error message | 12:50 |
| vidd | i said the result was expected =] | 12:50 |
| zykes- | If you read around in the other projects you see that "db sync" sets up (updates your db) to the most recent version, in other words if you use keystone from github and you do a git pull origin master or whatever then you should need to upgrade you db | 12:52 |
| zykes- | but if you skim the source code you see that there is no "migrate" stuff in keystone yet | 12:53 |
| zykes- | so the command that you mention is useless | 12:53 |
| *** hingo has quit IRC | 12:53 | |
| *** Glaceee has quit IRC | 12:53 | |
| *** hingo has joined #openstack | 12:53 | |
| *** cmagina has quit IRC | 12:54 | |
| zykes- | as in glance for example a db sync is used to upgrade the db to the most recent version of the schema spec | 12:54 |
| zykes- | probably will be in later keystone versions as well | 12:54 |
| *** cmagina has joined #openstack | 12:54 | |
| vidd | zykes-, if i hadnt see proof to the contrary with my own eyes, id tend to agree with you | 12:54 |
| zykes- | so you don't agree ? ;) | 12:54 |
| vidd | no | 12:55 |
| Razique | hey guys, I discovered that http://www.aeolusproject.org/ | 12:55 |
| Razique | seems great | 12:55 |
| vidd | i have already did a clean install, dropped my backed up database in and keystone refused to accept it until i sync'd it | 12:55 |
| zykes- | i'll check with devs later to clearify ;p | 12:56 |
| Razique | ok so does db_sync is a command or not ? :D | 12:56 |
| vidd | try it yourself....fire up a vm, and test it | 12:57 |
| vidd | Razique, yes it is | 12:57 |
| zykes- | vidd: what's the scenario, that after keystone is installed it doesn't create the db? | 12:57 |
| vidd | upon first launch | 12:57 |
| zykes- | lemme see :) | 12:58 |
| Razique | zykes-: by running keystone & in order to launch it | 12:58 |
| Razique | it creates and populate the database | 12:58 |
| vidd | on a clean install | 12:58 |
| Razique | what would be an unclean on in such case ? | 12:59 |
| uvirtbot | New bug: #878826 in openstack-manuals "Update doc for Cactus to Diablo upgrade needed" [Undecided,In progress] https://launchpad.net/bugs/878826 | 13:01 |
| vidd | if keystone-manage db_sync has already been run | 13:01 |
| Razique | what does it gives you when you run it another timE ? | 13:01 |
| Razique | i think i'll act as the nova does when you run nova-manage db sync | 13:02 |
| Razique | of glance-control db_sync | 13:02 |
| Razique | it must log the version uprade, that it :) | 13:02 |
| *** msivanes has joined #openstack | 13:02 | |
| *** hingo has quit IRC | 13:02 | |
| *** PeteDaGuru has joined #openstack | 13:02 | |
| *** hingo has joined #openstack | 13:03 | |
| Razique | vidd: so what is your issue now ? | 13:03 |
| Razique | keystone doesn't start ? | 13:03 |
| vidd | if you put up a database with data in it, and never ran keystone-manage db_sync, you get the error that the browser doesnt recognize or something like that | 13:04 |
| Razique | ok | 13:04 |
| vidd | my issue is the same issue i had yesterday and the day b4 | 13:04 |
| Razique | and if afterward you run the db_sync ? | 13:04 |
| *** superbobry has left #openstack | 13:04 | |
| Razique | sorry, it's just keystone is a pain for so many people here :D | 13:04 |
| vidd | if you run db_sync then all is well | 13:04 |
| vidd | yeah it is | 13:05 |
| *** hadrian has joined #openstack | 13:05 | |
| vidd | i wish the *buntu ppl would have fixed up thier deb so it did more than just put half the keystone stuff on the system | 13:06 |
| vidd | they are usually much better than that | 13:06 |
| vidd | but the realease was so close to the 11.10 release | 13:07 |
| koolhead17 | vidd: you would have checked for bugs before trying keystone from ubuntu repo :) | 13:08 |
| vidd | i expect come january, the deb will be fine as frog fuzz | 13:08 |
| * koolhead17 wonders if keystone code changes everyday | 13:08 | |
| *** lorin1 has joined #openstack | 13:08 | |
| koolhead17 | vidd: yes i think the same. | 13:08 |
| vidd | koolhead17, yes....it does | 13:08 |
| koolhead17 | vidd: so how you you expect it to be packaged and work perfectly, so most of us are using it from github :) | 13:09 |
| vidd | the script i downloaded 5 hours again set up a different database scheme than the onle i installed yesterday | 13:09 |
| Razique | vidd: use keysotne from github | 13:10 |
| vidd | it would be nice if it just sorta worked | 13:10 |
| Razique | I've been able to integrate it only from github | 13:10 |
| Razique | and nova/glance PPA's :) | 13:10 |
| *** AhmedSoliman has quit IRC | 13:10 | |
| vidd | least they could have done was include the examples folder that has the info for fixing api-paste | 13:11 |
| koolhead17 | Razique: glance is stable so if nova. | 13:11 |
| *** livemoon has joined #openstack | 13:11 | |
| Razique | glance from the PPA's yup :) | 13:11 |
| koolhead17 | *is | 13:11 |
| koolhead17 | :) | 13:11 |
| Razique | vidd: You want a working database scheme ? | 13:11 |
| koolhead17 | as vidd said even am hoping by jan keystone will be stable and painless | 13:12 |
| koolhead17 | :D | 13:12 |
| vidd | im using 11.10....the ppa's are for older versions | 13:12 |
| Razique | for Github's keystone, not the ubuntu one | 13:12 |
| koolhead17 | vidd: you dont need PPA for 11.10 | 13:13 |
| koolhead17 | :D | 13:13 |
| vidd | Razique, you had the db up for at least 2 days....it probably isnt compatable with the current install script =] | 13:13 |
| Razique | vidd: which install script are we talking about here ? | 13:14 |
| vidd | the one from github that is updated 6? hours ago | 13:14 |
| *** naehring has quit IRC | 13:14 | |
| *** vernhart has joined #openstack | 13:15 | |
| *** ziyadb has quit IRC | 13:16 | |
| Oneiroi | Nova,Glance,Keystone,Swift,Quantum .... did I miss anything? | 13:17 |
| * Oneiroi is preppring slides to run a trianing session here | 13:17 | |
| vidd | Oneiroi, dashboard | 13:17 |
| Oneiroi | thank you | 13:17 |
| zykes- | nerens: you doing quantum as well ? | 13:18 |
| Razique | vidd : do you have a link for that install script ? | 13:18 |
| Oneiroi | devstack ? | 13:18 |
| Razique | I generally don't use scripted installation :D | 13:18 |
| vidd | Razique, its burried in my bah history | 13:18 |
| *** ziyadb has joined #openstack | 13:19 | |
| zykes- | Oneiroi: using devstack ? | 13:19 |
| vidd | (how many lines does that go back?) | 13:19 |
| *** Shentonfreude has quit IRC | 13:19 | |
| *** Slav has joined #openstack | 13:19 | |
| Oneiroi | zykes-: nope no worky on oneiric | 13:19 |
| *** crussell has joined #openstack | 13:19 | |
| vidd | Oneiroi, it most certyainly does | 13:20 |
| Razique | vidd: here is what I did in order to make keystone running http://paste.openstack.org/show/2790/ | 13:20 |
| Oneiroi | vidd: it most certianly does not, reboot and dies | 13:20 |
| Oneiroi | nothing comes back | 13:20 |
| Oneiroi | fine so long as you never reboot | 13:20 |
| vidd | what do you expect....it a demo! | 13:20 |
| Oneiroi | it to not die horribly if I have a power brown out | 13:21 |
| Oneiroi | that'd be nice esp when I'm giving a demo of said demo ... | 13:21 |
| Oneiroi | demo death never fun ;-) | 13:22 |
| *** ianloic has quit IRC | 13:22 | |
| *** jingizu_ has quit IRC | 13:22 | |
| *** scottsanchez has quit IRC | 13:22 | |
| *** fujin has quit IRC | 13:22 | |
| *** Tribaal has quit IRC | 13:22 | |
| Razique | guys, what does that mean : http://paste.openstack.org/show/2791/ | 13:22 |
| Razique | NoMoreNetworks ? | 13:22 |
| Oneiroi | it's run out of ip's to allocate? | 13:22 |
| *** shaon has quit IRC | 13:23 | |
| Oneiroi | assuming a network exists and you've been using it | 13:23 |
| Razique | private ips or floating ones ? | 13:23 |
| Razique | I've a /24 network | 13:23 |
| Razique | and only two instances running | 13:23 |
| Razique | :D | 13:23 |
| vidd | hey...bill gates sold millions of copies of '98 even though it BSOD'd 32 seconds into booting | 13:23 |
| * Oneiroi shrug | 13:23 | |
| Oneiroi | vidd: thus proving windows users are idiots, I'd like to think better of *nix lusers ;-) | 13:24 |
| vidd | hehe | 13:24 |
| *** ianloic has joined #openstack | 13:24 | |
| vidd | the ppl paying us to set these things up are 94-1 windows users | 13:24 |
| vidd | 5% of them never touch a computer more complex then a calculator | 13:25 |
| Oneiroi | a guy on the training course complained citing "I didn't knwo we'd be using command line ..." | 13:25 |
| FabriceB | ho please leave this rant to leet groups | 13:26 |
| vidd | we're not ranting | 13:26 |
| *** ianloic has quit IRC | 13:26 | |
| FabriceB | no you're doing useless troll | 13:26 |
| Oneiroi | xD | 13:26 |
| Razique | no room for trolls here :o | 13:27 |
| Oneiroi | yeh we haven't settup br100 for them to live under yet | 13:27 |
| Oneiroi | /endpun | 13:27 |
| * Razique don't get it :D | 13:28 | |
| vidd | Razique, br100 is a bridge.... | 13:28 |
| vidd | homeless ppl live under bridges | 13:29 |
| *** jeh has joined #openstack | 13:29 | |
| vidd | oh...and trolls =] | 13:29 |
| Razique | ok | 13:29 |
| Oneiroi | so that raises the question | 13:29 |
| Oneiroi | where do you find a homeless troll? | 13:29 |
| Razique | I'm not a native english, so puns are not as easily interpreted by my brain as urs | 13:29 |
| Razique | :D | 13:29 |
| vidd | on freenode? | 13:30 |
| Oneiroi | it would seem so | 13:30 |
| Oneiroi | ^_^ | 13:30 |
| uvirtbot | Oneiroi: Error: "_^" is not a valid command. | 13:30 |
| Oneiroi | uvirtbot: quiet you | 13:30 |
| uvirtbot | Oneiroi: Error: "quiet" is not a valid command. | 13:30 |
| vidd | Oneiroi, its a bot...let it be | 13:30 |
| Razique | dudes, how can I work while u make fun here | 13:32 |
| Razique | :o | 13:32 |
| Razique | have* maybe | 13:32 |
| Oneiroi | Multiple monitors ? | 13:32 |
| Oneiroi | :D | 13:32 |
| livemoon | hi | 13:32 |
| Oneiroi | ih | 13:32 |
| Razique | Too lazy to plug the 60$ apple display cable | 13:32 |
| Razique | :D | 13:32 |
| livemoon | Razique: where is you doc? | 13:32 |
| vidd | Oneiroi, no...multiple monitors is not helpng me get this databse set up =] | 13:32 |
| livemoon | Oneiroi: hi | 13:32 |
| Razique | livemoon: sitill in progress, I've finished the Glance part | 13:33 |
| Razique | and almost the keystone | 13:33 |
| Razique | tonight I'll do the nova one | 13:33 |
| Oneiroi | Yeh Razique seriously how long does it take you to do something for free ... jeez | 13:33 |
| livemoon | Razique: good job | 13:33 |
| vidd | ive been waiting MINUTES already! | 13:33 |
| vidd | =] | 13:34 |
| *** tacitone has joined #openstack | 13:34 | |
| *** bcwaldon has joined #openstack | 13:34 | |
| *** pixelbeat has quit IRC | 13:34 | |
| zykes- | i'll have nova and dashboard up tonight | 13:34 |
| zykes- | gonna be fun ;) | 13:34 |
| *** pixelbeat has joined #openstack | 13:34 | |
| livemoon | I said: Razique is my teacher | 13:34 |
| zykes- | Oneiroi: why doesn't quantum work on oneiric | 13:35 |
| Oneiroi | HAving to order new hardware here ... :( | 13:35 |
| Razique | zykes-: cook! | 13:35 |
| tacitone | Hey all, configuring swift and I can't get tempauth to load on swift-init proxy start (diablo, Debian Squeeze) | 13:35 |
| vidd | i hope to get to dashboard before i colopse into a heap on the floor | 13:35 |
| Razique | I dunno if the dashboard worths it actually | 13:35 |
| Razique | Oneiroi: ahah | 13:35 |
| Oneiroi | zykes-: *shrug* not tried yet, did you divide by zero yet? | 13:35 |
| zykes- | divide by zero ? | 13:35 |
| Razique | must be another pun | 13:35 |
| Razique | :D | 13:35 |
| * Razique thinking | 13:35 | |
| zykes- | oh bah | 13:35 |
| Oneiroi | http://i270.photobucket.com/albums/jj105/callatov/Divided_by_zero.jpg | 13:36 |
| livemoon | Now I am not ready to use dashboard | 13:36 |
| Razique | broken link :D | 13:36 |
| livemoon | I need to use CLI and learn nova-api | 13:36 |
| *** kaigan_ has joined #openstack | 13:36 | |
| Razique | livemoon: yah, dashboard is cool and sexy for demos | 13:36 |
| livemoon | then I want to write my own web with api | 13:36 |
| Razique | but u won't use it I think :D | 13:36 |
| tacitone | http://paste.openstack.org/show/2792/ | 13:37 |
| Oneiroi | Razique: ok try http://i302.photobucket.com/albums/nn97/XxharmlessXsinxX/588px-Divide_by_zer0.jpg | 13:37 |
| tacitone | I figure I'm missing the tempauth module somehow. | 13:37 |
| Razique | Oneiroi: sorry, but I don't get that one either :/ | 13:37 |
| Oneiroi | >< | 13:37 |
| Razique | stupid native puns | 13:38 |
| Razique | :p | 13:38 |
| Oneiroi | not native, rule #52 of the internet iirc | 13:38 |
| livemoon | I am going to watch nikita. | 13:38 |
| livemoon | bye all | 13:38 |
| Razique | hehe enjoy | 13:38 |
| livemoon | good work | 13:38 |
| Oneiroi | I'm going to watch nikita from behind your sofa | 13:38 |
| vidd | division by zero is the quickest, fastest, easies way to blow up a program | 13:38 |
| Razique | ah ok | 13:39 |
| Razique | thanks | 13:39 |
| tacitone | I ran a grep on /usr/lib/pymodules/python2.6 & /usr/lib/pymodules/python2.7 - I thought having two concurrent versions of Python installed would cause a problem, but the grep shows there are no files present with a mention of tempauth. | 13:39 |
| Razique | cya livemoon :) | 13:39 |
| Oneiroi | tacitone: grep -i ? | 13:39 |
| tacitone | Oneirio: good idea 8) | 13:40 |
| vidd | the two versions are there to support python scripts that are ...... behind the times | 13:40 |
| vidd | i deliberatly refrained from making a microsoft jab | 13:41 |
| tacitone | :D | 13:41 |
| koolhead17 | haha | 13:42 |
| tacitone | Yep, no eggs in /usr/lib/pymodules containing pattern 'tempauth' [grep -R -i] | 13:42 |
| tacitone | Built my debs from bzr | 13:42 |
| vidd | somebody said something about tempauth a couple of hours ago....dont remember what | 13:43 |
| tacitone | Maybe I can get tempauth from pypi? | 13:43 |
| Oneiroi | well if they did it'll be in teh logs | 13:43 |
| Oneiroi | http://eavesdrop.openstack.org/irclogs/ | 13:43 |
| *** kbringard has joined #openstack | 13:43 | |
| tacitone | swauth and tempauth not registered on PyPi O.o | 13:44 |
| *** siwos has quit IRC | 13:44 | |
| vidd | its befor 7am and afte 2am EST | 13:44 |
| *** jingizu_ has joined #openstack | 13:44 | |
| vidd | oh...it was swauth | 13:45 |
| zykes- | what is aeolus ? kind of like deltacloud or ? | 13:45 |
| Razique | zykes-: yah, kinda puppet-stuff | 13:45 |
| Razique | but for instances | 13:46 |
| tacitone | Many refs to swauth in /usr/lib/pymodules.. Perhaps I should switch to swauth - advantages over tempauth, anyone? | 13:46 |
| *** ejat has joined #openstack | 13:46 | |
| *** ejat has joined #openstack | 13:46 | |
| Razique | you create xml manifests, and it handles for you the while image creation | 13:46 |
| Oneiroi | tacitone: advantage is you have swauth? | 13:46 |
| tacitone | Oneirio: Point well taken X] | 13:46 |
| zykes- | Razique: doing what? | 13:46 |
| Oneiroi | :D | 13:46 |
| *** mdomsch has joined #openstack | 13:46 | |
| Razique | ooops | 13:47 |
| Razique | sorry I wrote like I was drunk | 13:47 |
| vidd | zykes-, puppet is a tool for auto-configuring servers to expand your cloud | 13:47 |
| Razique | zykes-: You create xml fies that define how you image will be | 13:48 |
| tacitone | #!/bin/bash | 13:48 |
| *** Tribaal has joined #openstack | 13:48 | |
| tacitone | :P | 13:48 |
| Razique | like Os, packages installed into it, user to create | 13:48 |
| tacitone | Ah, so puppet is an imaging service? | 13:48 |
| Razique | then it does everything for you…. instead of launchink kvm yourself with as long flags as my arm | 13:49 |
| vidd | more like a provisioning service | 13:49 |
| Razique | tacitone: Puppet is a post deplpyment tools | 13:49 |
| zykes- | yeah, i know what puppet is ;p, but what does aeolus do, like build ontop or ? | 13:49 |
| tacitone | #!/bin/bash is also a post deployment tool :P | 13:49 |
| Razique | it customizes a system and make sure for a given server, all the instruction are satisfied | 13:49 |
| jeh | puppet is configuration management. A single interface that controls all the configuration (users, files, packages, services, etc) for all your systems. Cloud based or not | 13:49 |
| FabriceB | a cfengine on steroid | 13:50 |
| Razique | for instance, it makes sure you have the apache service installed on a server | 13:50 |
| Razique | I myself used Puppet for a client | 13:50 |
| Razique | I've been able to provide around 60 VMs easily | 13:50 |
| Razique | without it, it would have taken me days | 13:50 |
| zykes- | we use it here as well | 13:51 |
| *** imsplitbit has joined #openstack | 13:51 | |
| FabriceB | an alternative to puppet is chef | 13:52 |
| Razique | it's a cool tool yah | 13:52 |
| FabriceB | both are ruby based | 13:52 |
| *** fujin has joined #openstack | 13:52 | |
| Razique | FabriceB: I read a nice article that claims Puppet is better o/ | 13:52 |
| Razique | http://bitfieldconsulting.com/puppet-vs-chef | 13:52 |
| Razique | :p | 13:52 |
| FabriceB | mm i think i can find an article saying chef is better | 13:52 |
| Razique | ahah I think it' not about the tools | 13:53 |
| Razique | it's about the one who use them :) | 13:53 |
| vidd | what part of puppet development did the writer create? | 13:53 |
| Razique | vidd: what do you mean ? | 13:53 |
| FabriceB | but any way those kind of tools are almost mandatory to manage a cloud | 13:53 |
| *** Oneiroi has quit IRC | 13:53 | |
| Razique | FabriceB: depends of what kind of cloud we are talking about | 13:55 |
| vidd | it implies the writer is bias because he may have participated in making the product he was endorcing | 13:55 |
| crussell | Guys, i'm trying to figure out the relationship to some pieces of openstack. | 13:55 |
| Razique | I guess it's not that important on a small IAAS platform :) | 13:55 |
| Razique | hu crussell just ask :) | 13:55 |
| crussell | WHen an image runs.. it reads from a parent image housed in glance? | 13:55 |
| crussell | and then the differencing disk, is wrote down to nova-volume? | 13:55 |
| vidd | like when stev jobbs used to talk about how much better the iphone was compared to the blackberry | 13:55 |
| Razique | crussell: nope | 13:56 |
| crussell | glad i asked then | 13:56 |
| vidd | [please note, yet another avoidance of a microsoft jab] | 13:56 |
| Razique | crussell: 1 - you have images in Glance 2- when you ask to run one , it is sent to a node that has been elected to run that image | 13:56 |
| *** patelna has joined #openstack | 13:56 | |
| Razique | 3- the image is sent to that server, which is then stored locally | 13:57 |
| Razique | (in the _base dir) | 13:57 |
| vidd | is nova-volume even used anymore? | 13:57 |
| vidd | or did glance replace it | 13:57 |
| Razique | 4- that file, called a "backing file" 5- from that file is created a smaller one that is your instance | 13:57 |
| crussell | WHich is essentially a difference disk. | 13:58 |
| Razique | 5- nova-volumes allows you to attach an extra disk to that server (similary to the amazon s3 stuff) | 13:58 |
| Razique | vidd: nova-volumes <> glance | 13:58 |
| Razique | the benefit here is to have an instance, and a disk that doesn't depend on it (nova-volume) | 13:59 |
| vidd | ah....storage space | 13:59 |
| Razique | thus, you can delete your instance, that won't delete the volume you attached | 13:59 |
| Razique | vidd: exactly, which has it's own life | 14:00 |
| Razique | when you terminate an instance, the backing file is not removed | 14:00 |
| crussell | Ok, with that said then.. my /var/lib/nova/instnace dir can be a mounted shared volume sourced from a san and shared amoung hosts since ext4 knows how to singularly lock a file with out concurency problems | 14:00 |
| Razique | and when you launch annother instance with the same base image, you re-create an image from that backing file | 14:00 |
| Razique | and voila :) | 14:00 |
| *** hacim has left #openstack | 14:01 | |
| crussell | Good explanation, ty | 14:01 |
| *** sdake has quit IRC | 14:01 | |
| *** sdake has joined #openstack | 14:02 | |
| crussell | And if i can.. then my volumes can be replicated by the san across datacenters. | 14:02 |
| FabriceB | you're /var/lib/nova/instnace is mounted on many hosts at the same time ? | 14:02 |
| FabriceB | instance | 14:02 |
| crussell | I'm asking can it.... | 14:02 |
| FabriceB | ho that's not possible | 14:03 |
| *** PeteDaGuru has quit IRC | 14:03 | |
| *** PeteDaGuru has joined #openstack | 14:03 | |
| crussell | if i have a block stoage san.. such as equalogic, i that is attached via iscsi | 14:03 |
| FabriceB | you can't share a fs unless it's build for | 14:03 |
| FabriceB | like gfs from rehdat | 14:03 |
| Razique | FabriceB: in fact of you want to do live migration you need to share that directory | 14:03 |
| crussell | I thought ext4 could handle it without corruption | 14:04 |
| FabriceB | for cache coherency problem from example | 14:04 |
| Razique | FabriceB has a point here | 14:04 |
| FabriceB | Razique: not from common fs | 14:04 |
| crussell | i know darn well ntfs can not handle it | 14:04 |
| Razique | FabriceB: you are totally right | 14:04 |
| Razique | crussell: you can use NFS, but perf will sux | 14:04 |
| FabriceB | ntfs, xfs, ext[1..4] | 14:05 |
| Razique | you should use GFS or | 14:05 |
| FabriceB | but gfs, gpfs, and some other | 14:05 |
| Razique | I don't recall the other one | 14:05 |
| jeh | gluster | 14:05 |
| FabriceB | lustre i think | 14:05 |
| Razique | nope | 14:05 |
| Razique | another one :D | 14:05 |
| Razique | DRBD ? | 14:05 |
| FabriceB | ceph is on the way to do it | 14:05 |
| Razique | yup drbd | 14:05 |
| FabriceB | drbs is for block sharing afaik | 14:05 |
| Razique | http://www.openstack.org/blog/2011/07/announcing-the-gluster-connector-for-openstack/ | 14:06 |
| Razique | time to cofee | 14:06 |
| crussell | I just have this hesitation.. that my customers data is residing on some commoditoy mirror... and in first case use for the cloud, it would be an entire VM. - How do i repliacate this locally sotred data then | 14:06 |
| Razique | djee I was supposed to work | 14:06 |
| Razique | I just did nothing but talk here ahah | 14:06 |
| *** dprince has joined #openstack | 14:06 | |
| *** robbiew has joined #openstack | 14:06 | |
| Razique | crussell: you would have a shared FS for that then :) | 14:07 |
| crussell | The Gluster Connector for OpenStack …supports the virtual motion of the VMs within the OpenStack compute environment. | 14:07 |
| crussell | this is me.. | 14:07 |
| crussell | exactly what i'm looking for. | 14:07 |
| Razique | crussell: You could use I think an ugly hack via rsync and it's sparse files option | 14:07 |
| Razique | but that wouldn't be cool at all :D | 14:08 |
| crussell | doesn't sound predictable or reliable =) | 14:08 |
| koolhead17 | Can i pull a particular repo on basis of date ? | 14:08 |
| koolhead17 | from github? | 14:08 |
| *** tacitone has left #openstack | 14:08 | |
| * koolhead17 is no github export | 14:08 | |
| *** fmancinelli has quit IRC | 14:08 | |
| *** ejat has quit IRC | 14:09 | |
| *** gnu111 has joined #openstack | 14:10 | |
| *** nerens has quit IRC | 14:10 | |
| *** Razique has left #openstack | 14:16 | |
| crussell | Gluster/Openstack seems entirely too simple to implement | 14:17 |
| gnu111 | yes, much easier than swift :D | 14:18 |
| jtanner | that's good | 14:18 |
| notmyname | gnu111: :-( | 14:18 |
| *** scottsanchez has joined #openstack | 14:19 | |
| gnu111 | notmyname: I got swift up and running. just head to read the doc carefully. :) | 14:19 |
| vidd | notmyname, someone just had a question about auth for swift | 14:19 |
| *** nelson has quit IRC | 14:20 | |
| * notmyname scrolls up | 14:20 | |
| notmyname | tacitone. looks like he left | 14:20 |
| vidd | like 10 minutes ago | 14:21 |
| *** deshantm has quit IRC | 14:21 | |
| *** fmancinelli has joined #openstack | 14:21 | |
| gnu111 | So, I am still trying to understand what this means: "Assuming there are 5 zones with 1 node per zone, ZONE should start at 1 and increment by one for each additional node." I have 8 nodes to play with, so far I added three storage and one proxy. I basically did export ZONE=1 then 2 and built the ring for each zone and storage node. is that right? | 14:22 |
| notmyname | gnu111: you have to name your zones numbers starting at 1. so zone1, zone2, zone3, etc | 14:24 |
| *** lvaughn_ has joined #openstack | 14:25 | |
| notmyname | gnu111: with 8 nodes, you could do 2 proxies, 1 load balancer, and 5 storage nodes (for all kinds of HA) | 14:25 |
| gnu111 | ok. I understand now. I named them "1" "2" '3" ...how many nodes can be in each zone? | 14:25 |
| notmyname | gnu111: ya, I think that's right for building the ring | 14:26 |
| gnu111 | is there a command to see all the zones? | 14:26 |
| vidd | looks like they made some nice improvements to the keystone database | 14:26 |
| gnu111 | swift list just lists the container. | 14:27 |
| vidd | we got a credentials table with a "type" "key" and "secret" fields | 14:27 |
| notmyname | gnu111: swift-ring-builder <ring builder file> | 14:27 |
| vidd | this thing might actually come to something next year! | 14:28 |
| *** lvaughn has quit IRC | 14:28 | |
| *** hingo has quit IRC | 14:29 | |
| *** hingo has joined #openstack | 14:29 | |
| crussell | man, i'm a glutton for punishment.. i slick and reloaded again to start a new... | 14:29 |
| crussell | this time with gluster! | 14:29 |
| *** praefect has joined #openstack | 14:31 | |
| gnu111 | notmyname: ok that worked! thanks. | 14:31 |
| gnu111 | notmyname: 262144 partitions, 3 replicas, 3 zones, 3 devices, 0.00 balance. I uploaded only few files so far. | 14:32 |
| praefect | hi guys, I use --auto_assign_floating_ip=true and it works but it cannot release the floating ips, even if I do a euca-disassociate-address and euca-release-address it complains that the floating address is still in use... | 14:32 |
| *** deshantm has joined #openstack | 14:33 | |
| praefect | I think this only happen with auto_assign_floating_ip but I'm sure most of you use this switch | 14:33 |
| vidd | praefect, there is a nova.conf flag to set the "auto-release" of those IPs | 14:34 |
| livemoon | I am coming | 14:34 |
| livemoon | I have watched nikita | 14:34 |
| praefect | vidd: no way!!! I'll look into it right now | 14:35 |
| vidd | praefect, give me a sec | 14:35 |
| livemoon | what are you talking about? | 14:36 |
| praefect | vidd: the truth is that the really useful auto_assign_floating_ip switch is not even in the official "Reference for flags in nova.conf" diablo doc !!! | 14:37 |
| praefect | that's just unbelievable | 14:37 |
| *** javiF has quit IRC | 14:37 | |
| gnu111 | vidd: is this it? update_dhcp_on_disassociate ? | 14:37 |
| praefect | livemoon: auto_assignment of floating ips and auto_release (looking for a switch to make this one last work) | 14:37 |
| livemoon | I know those two switches | 14:38 |
| livemoon | but I don't use | 14:38 |
| gnu111 | praefect: do a nova-manage --help | 14:38 |
| praefect | thanks | 14:38 |
| *** dgags has joined #openstack | 14:39 | |
| gnu111 | notmyname: do you recommend any benchmarking tool for swift such as iozone? | 14:39 |
| *** osier has quit IRC | 14:39 | |
| praefect | grepping release from "nova-manage --help" I get force_dhcp_release : if true, send a dhcp release on instance termination | 14:40 |
| praefect | could that be it | 14:40 |
| notmyname | gnu111: there is an included swift-bench tool that will allow the closes apples-to-apples comparison between swift clusters | 14:40 |
| vidd | --fixed_ip_disassociate_timeout | 14:40 |
| praefect | vidd: but it's not fixed it's floating I'm looking for | 14:40 |
| vidd | Integer: Number of seconds after which a deallocated ip is disassociated. | 14:41 |
| gnu111 | notmyname: thanks, will check it out. | 14:41 |
| vidd | praefect, so if you say"dont use this ip anymore, this flag says how long before it realocatable again | 14:42 |
| vidd | the default is 10 minutes | 14:42 |
| praefect | vidd:yes but is says fixed_ip which is the private network not the floating ip | 14:42 |
| praefect | do you think it would work for floating too | 14:43 |
| vidd | not true | 14:43 |
| praefect | vidd: sorry? | 14:43 |
| vidd | you cant assign a fixed ip to an instance | 14:43 |
| livemoon | does anyone use other iscsi as volume | 14:43 |
| praefect | vidd: that's right, that's what I was wondering | 14:43 |
| vidd | you can only assign floating IPs | 14:44 |
| praefect | but still, we are way passed the 10 minutes and I still cannot release the addr, even manually with euca, there's a bug there for sure | 14:44 |
| vidd | you kill a VM, that fixed ip is immediately avaiable | 14:44 |
| praefect | vidd: that's right, especially with flatdhcp where DHCP handles this and not openstack | 14:45 |
| vidd | any other vms running? | 14:45 |
| livemoon | bye。vidd | 14:46 |
| vidd | its either 10 minutes or 10 hours....not sure if the "600" is seconds or minutes | 14:46 |
| livemoon | hi, vidd ,what's time now your area? | 14:47 |
| vidd | seconds | 14:47 |
| praefect | oh... well I could set it to 2 and test it, I'll do that | 14:47 |
| vidd | live 10:47 am | 14:47 |
| vidd | livemoon, 10:47 am | 14:47 |
| vidd | praefect, do you have other vm's running? | 14:48 |
| livemoon | bye | 14:48 |
| vidd | cuzz you have to kill nova to reload the conf | 14:48 |
| praefect | vidd: ok I just restarted nova-network.. | 14:49 |
| praefect | I do have other VMs running | 14:49 |
| *** f-yang has quit IRC | 14:49 | |
| *** livemoon has left #openstack | 14:49 | |
| vidd | its nova not nova-network you have to restart | 14:50 |
| praefect | but restarting nova-network, the nova-network.log now states "fixed_ip_diassociate_timeout = 2" I would think this is now in effect... | 14:50 |
| praefect | vidd: what do you mean nova (I'm on centos, running services like nova-api, nova-network, nova-scheduler, etc.) | 14:50 |
| vidd | nova-api | 14:50 |
| praefect | ok I can restart it, no problem it does not affect the running VMs | 14:51 |
| praefect | I'm on a multi-node config anyway | 14:51 |
| *** Ryan_Lane has joined #openstack | 14:51 | |
| vidd | so the other vm's are on other machines? | 14:51 |
| praefect | yes they are | 14:51 |
| praefect | nova-api restarted | 14:51 |
| vidd | good | 14:51 |
| vidd | lets test it | 14:51 |
| *** alexn6 has joined #openstack | 14:52 | |
| praefect | just a side-note, I've spent last week on a problem where I would launch a bunch of VM (rendering) overnight and in the morning they've lost all connectivity, nothing in the logs but no machine would ping, nothing... every night I would change something and in the morning, lost connectivity | 14:52 |
| praefect | it turned out my ntpdate wasn't being executed properly in the crontab.. once the machines kept their timer in sync, all works well... | 14:53 |
| *** code_franco has joined #openstack | 14:53 | |
| FabriceB | ntpdate in crontab??? | 14:53 |
| praefect | yeah I know I should use ntpd or something.. | 14:53 |
| praefect | but now at least it works aith ntpdate in cron.hourly | 14:54 |
| FabriceB | the both use the same configuration file | 14:54 |
| FabriceB | so if ntpdate is working, ntpd too | 14:54 |
| alexn6 | Hi all! who can say, is it ok to use last glance version from git clone with previous openstack release(swift, nova, swauth)? Will I need to recreate or modify stored images and how if yes? | 14:54 |
| praefect | basically what I'm saying is that without proper timer sync (ntp or whatever you want) this openstack cluster is falling apart in a matter of hours | 14:54 |
| *** nikolaij has quit IRC | 14:55 | |
| praefect | alexn6: I don't know for sure but I'd bet $2 it works right away, I have a a mixed env here (cactus compute nodes, with diablo comptue nodes and diablo controller) it works fine.. | 14:55 |
| *** reed__ has joined #openstack | 14:56 | |
| *** rnirmal has joined #openstack | 14:56 | |
| praefect | even though the nova-compute doesn't even call qemu-kvm the same way (the newer one uses the resize function) | 14:56 |
| praefect | still works | 14:56 |
| alexn6 | :) cool | 14:56 |
| vidd | thats why most ppl recommend you set your own internal ntpd and have all the machines in your cloud check there, and have that machine check the outside time every so often to update itsself | 14:57 |
| vidd | praefect, ^^^ | 14:57 |
| *** kbringard has quit IRC | 14:58 | |
| *** kbringard has joined #openstack | 14:58 | |
| vidd | so praefect did that setting do it for you? | 14:58 |
| praefect | vidd: actually we already have a ntp server running internally, I just tell my nodes to ntpdate there every hour | 14:58 |
| vidd | im telling you, have your nodes check your compute every 5 minutes | 14:59 |
| vidd | have your compute node check every 10 hours | 14:59 |
| vidd | you can probably have rabbitQM set up to push the updates | 15:00 |
| vidd | s/w compute/controller | 15:00 |
| kbringard | why not just apt-get install ntp and be done with it | 15:01 |
| *** mrjazzcat has joined #openstack | 15:01 | |
| kbringard | that's what I do | 15:01 |
| FabriceB | that's all we do | 15:02 |
| *** mrjazzcat has left #openstack | 15:02 | |
| kbringard | it checks against time.ubuntu.com or whatever and I've never had a problem | 15:02 |
| *** TheOsprey has quit IRC | 15:02 | |
| FabriceB | ntpdate does none monotonous time modification | 15:02 |
| vidd | kbringard, his machines get out of sync rather quickly | 15:02 |
| FabriceB | so the same time can be seen twice (or never) | 15:02 |
| *** dragondm has joined #openstack | 15:02 | |
| FabriceB | the faster the machines get out of sync, the more risks you take | 15:03 |
| kbringard | I think ntpd does continuous updates | 15:03 |
| FabriceB | yes | 15:03 |
| FabriceB | but not ntpdate | 15:03 |
| vidd | ntp server == ntpd | 15:03 |
| alexn6 | praefect: but you didn`t use git clone for that mix, right? just apt-get with corrected repos? | 15:05 |
| * vidd scrolls back.... | 15:05 | |
| vidd | i did say ntpd | 15:05 |
| *** kaigan_ has quit IRC | 15:08 | |
| *** AhmedSoliman has joined #openstack | 15:09 | |
| vidd | im confussed.... | 15:09 |
| *** freeflyi1g has quit IRC | 15:10 | |
| *** freeflying has joined #openstack | 15:10 | |
| *** dubsquared has joined #openstack | 15:10 | |
| praefect | alexn6: I'm on centos and I only use packages (yum.griddynamics.net) so basically I can install a compute node with a cactus repo and another one with a diablo repo... all packages | 15:11 |
| *** Slav has quit IRC | 15:11 | |
| praefect | vidd:I understand ntpd is the proper way to set it up, I took the lazy route and setup ntpdate in a cron.hourly | 15:11 |
| *** vladimir3p has joined #openstack | 15:12 | |
| *** Oneiroi has joined #openstack | 15:12 | |
| praefect | But I'm just warning the new ones, a time difference of less than a minute is enought to loose the smile in "nova-manage service list" and probably mess things up down the line.. be careful | 15:13 |
| vidd | the database structure for keystone for table "user" is ID (auto-increment) name password email enabled tenant_ID but if i try to do the keystone-manage user add for all these fields if throws an error | 15:13 |
| vidd | 28 second for my two machines | 15:14 |
| *** rnorwood has joined #openstack | 15:14 | |
| vidd | praefect, while your in there, ya might as well get the ntpd set up proper =] | 15:15 |
| *** bencherian has joined #openstack | 15:15 | |
| vidd | i have old machines...they get out of sync rather quickly | 15:16 |
| *** salv-orlando has joined #openstack | 15:16 | |
| *** javiF has joined #openstack | 15:17 | |
| *** g2 has joined #openstack | 15:18 | |
| *** Shentonfreude has joined #openstack | 15:21 | |
| *** HowardRoark has joined #openstack | 15:23 | |
| *** michaelkre has joined #openstack | 15:23 | |
| *** maplebed has joined #openstack | 15:24 | |
| *** reed__ is now known as reed | 15:27 | |
| *** maplebed has quit IRC | 15:33 | |
| *** cmagina has quit IRC | 15:33 | |
| *** fmancinelli has quit IRC | 15:36 | |
| *** chasd00 has joined #openstack | 15:37 | |
| *** chasd00 has left #openstack | 15:38 | |
| *** fmancinelli has joined #openstack | 15:39 | |
| *** cmagina has joined #openstack | 15:40 | |
| vidd | what listens on port 8774? | 15:42 |
| *** dobber has quit IRC | 15:42 | |
| agy | vidd: nova-api for the openstack api | 15:44 |
| vidd | ty | 15:44 |
| vidd | setting up the keystone database | 15:44 |
| vidd | i want to know whats going into is so I can write the plugins for the billing software we are going to use [groan] | 15:45 |
| *** fmancinelli has quit IRC | 15:46 | |
| agy | nova-api listens on 8773 for the EC2 api (for completeness) | 15:46 |
| *** Razique has joined #openstack | 15:47 | |
| Razique | Back | 15:47 |
| vidd | hello Razique | 15:47 |
| vidd | welcome back | 15:47 |
| Razique | since I've integrated keystone, I can't launch anymore instances :( | 15:47 |
| Razique | http://paste.openstack.org/show/2794/ | 15:47 |
| *** fmancinelli has joined #openstack | 15:47 | |
| Razique | :) | 15:47 |
| *** vernhart has quit IRC | 15:47 | |
| Razique | RemoteError: NoMoreNetwork | 15:47 |
| *** Ryan_Lane has quit IRC | 15:47 | |
| Razique | weird :/ | 15:47 |
| Razique | the network is here, linked to the project | 15:48 |
| Razique | and 254 addresses are available | 15:48 |
| Razique | everything went perfect before I integrated keystone, any hint ? | 15:48 |
| vidd | how out-of-sync are your machines time-wise? | 15:48 |
| Razique | vidd: I don't understand what you are asking | 15:49 |
| Razique | :p | 15:49 |
| Razique | ahhh | 15:49 |
| Razique | exact same time | 15:49 |
| Razique | they both use a ntp server | 15:50 |
| vidd | we just had a discussion about how important 30 seconds is to thes things | 15:50 |
| vidd | how many vm's are up? | 15:50 |
| Razique | there is not time offset at all | 15:50 |
| Razique | I had two, but since I messed up the network in order to try to fix that | 15:50 |
| Razique | I terminated all the instances | 15:50 |
| *** nati2 has joined #openstack | 15:51 | |
| Razique | now not any instance is running | 15:51 |
| vidd | so currently...none | 15:51 |
| Razique | yup | 15:51 |
| vidd | restart all your -networks | 15:51 |
| Razique | I'm so pissed I was nearly able to validate the migration process cactus -> diablo/ keystone but now I can't guarantee the network will be still up | 15:52 |
| Razique | like service nova-network restart ? | 15:52 |
| vidd | right | 15:52 |
| Razique | done | 15:52 |
| *** fmancinelli has quit IRC | 15:52 | |
| vidd | what process are you using to bring up an instance? | 15:53 |
| vidd | the old nova way or the keystone nova way? | 15:53 |
| Razique | depends.. I use to use euca2ools or nova | 15:53 |
| vidd | thats the old nova way | 15:53 |
| Razique | typically nova --debug boot --flavor 1 --image 94 --availability_zone razique_lab --security_groups default novaclient_instance | 15:53 |
| vidd | now you need to do it the keystone-nova way | 15:54 |
| Razique | oh, there is a new way to boot up instances with keystone ahah ? | 15:54 |
| Razique | sick :p | 15:54 |
| *** AhmedSoliman has quit IRC | 15:54 | |
| vidd | yes....dont ask me what it is....i havent gotten that far yet | 15:54 |
| Razique | do you have any leads for me ? | 15:55 |
| *** reidrac has left #openstack | 15:55 | |
| vidd | im still setting up the database in keystone | 15:55 |
| uvirtbot | New bug: #878925 in openstack-ci "glance builder has user_xattr mount option" [Undecided,New] https://launchpad.net/bugs/878925 | 15:56 |
| *** halfss has quit IRC | 15:56 | |
| vidd | in glance you send the request to keystone and keystone tells nova to do it | 15:56 |
| vidd | *tells glance to do it | 15:56 |
| vidd | so i would imagine nova works pretty much the same way | 15:56 |
| *** ewindisch has quit IRC | 15:56 | |
| Razique | oh my good, that is so not the role of Keystone :/ | 15:57 |
| Razique | I'm trying to google that | 15:57 |
| *** ewindisch has joined #openstack | 15:57 | |
| vidd | what do you mean? of course its the role of keystone | 15:57 |
| Razique | Keystone is just here to manage auth. | 15:58 |
| Razique | keystone is an entrypoint to endpoints | 15:58 |
| Razique | since the user is authenticated ; why does keystone manage instances launching ? | 15:58 |
| vidd | keystone is there to take the call "do this" check to make sure the account isnt suspended, then either say"go to hell" or ok, go ahead | 15:58 |
| vidd | its also an excelent billing interupt | 15:59 |
| *** rsampaio has joined #openstack | 16:00 | |
| Razique | vidd: there is not any LINE at all on the web about that new instances launching | 16:00 |
| vidd | its the reason its taking me hous to enter a single user into the keystone database | 16:00 |
| Razique | it's so scary =D | 16:00 |
| vidd | im going to have to write the billing api | 16:00 |
| uvirtbot | New bug: #878927 in glance "Glance + keystone Exception: Unexpected response: 400" [Undecided,New] https://launchpad.net/bugs/878927 | 16:01 |
| vidd | its NOT NEW | 16:01 |
| vidd | its one of the thigs keystone is built around | 16:01 |
| Razique | what do you mean ? | 16:02 |
| vidd | this why, some guy in billing doesnt have to call some guy in engineering to call some guy in the datcenter to suspend an account | 16:02 |
| vidd | s/w why/way | 16:03 |
| Razique | yah I know | 16:03 |
| Razique | but keystone don't have to DIRECTLY interfere with a service usage | 16:03 |
| Razique | it makes sure a user is granted to consume a service | 16:04 |
| Razique | and provide the right token for him | 16:04 |
| *** worstadmin has joined #openstack | 16:04 | |
| vidd | if he isnt 6 months behind in his payment | 16:04 |
| Razique | we are talking about keystone that would manage instance launching here ? | 16:04 |
| *** TheOsprey has joined #openstack | 16:06 | |
| vidd | we keystone doesnt "manage" it | 16:06 |
| vidd | keystone just checks to make sure those credentials are still valid | 16:06 |
| *** nid0 has quit IRC | 16:07 | |
| vidd | and then lets the call go thru | 16:07 |
| *** alekibango_ is now known as alekibango | 16:11 | |
| *** catintheroof has joined #openstack | 16:11 | |
| *** obino has quit IRC | 16:11 | |
| *** nid0 has joined #openstack | 16:12 | |
| alexn6 | is it very complex to migrate cactus->diabolo? the most complex subjects is keystone? and how about image conversion? Thanks for answer! | 16:12 |
| FabriceB | does'n the image format depends of you're virtualizer ? | 16:13 |
| *** dendrobates is now known as dendro-afk | 16:14 | |
| vidd | port 9292 is glance api or registry? | 16:15 |
| *** df1 has joined #openstack | 16:15 | |
| alexn6 | I just don`t know what that means -"then perhaps your images if you were already running images in the nova-objectstore" from diablo manual migrating | 16:15 |
| *** crayon has quit IRC | 16:15 | |
| *** justinlw has joined #openstack | 16:16 | |
| vidd | i think nova-objectstore is being replaced with glance | 16:16 |
| vidd | im not 100% on that though | 16:16 |
| *** lotrpy has quit IRC | 16:16 | |
| *** adjohn has joined #openstack | 16:16 | |
| *** crayon_z has joined #openstack | 16:16 | |
| *** lotrpy has joined #openstack | 16:17 | |
| *** katkee has quit IRC | 16:17 | |
| *** kbringard has quit IRC | 16:18 | |
| agy | vidd: 9292 glance-api | 16:20 |
| *** nacx has quit IRC | 16:20 | |
| vidd | ty | 16:20 |
| *** nelson has joined #openstack | 16:20 | |
| agy | vidd: 9191 is glance-registry | 16:20 |
| vidd | that should be the last one =] | 16:21 |
| vidd | almost done with the database =] | 16:21 |
| *** vdo has quit IRC | 16:25 | |
| *** kbringard has joined #openstack | 16:26 | |
| *** dendro-afk is now known as dendrobates | 16:29 | |
| Razique | vidd: but why i need to use the "new" nova-keystone method to launch instances then :s ? | 16:29 |
| *** maplebed has joined #openstack | 16:30 | |
| Razique | viid | 16:30 |
| Razique | vidd: It's like there is some mixes about user/id with keystone and nova | 16:30 |
| Razique | for instance my user id for nova is "razique" | 16:30 |
| Razique | while it's my name for keystone, and my id is "1" | 16:31 |
| Razique | and i think it's where that's messy :p | 16:31 |
| *** orclev has quit IRC | 16:32 | |
| Razique | djee if I don't figure it out quickyl, I won't use keystone in production :/ can't afford to break my network just because of that | 16:32 |
| Razique | the production openstack I mean :) | 16:32 |
| *** tillmo has joined #openstack | 16:35 | |
| Razique | gtg bye all | 16:36 |
| Razique | thanks vidd | 16:36 |
| *** Razique has left #openstack | 16:36 | |
| *** lvaughn_ has quit IRC | 16:39 | |
| *** lvaughn has joined #openstack | 16:42 | |
| *** jdurgin has joined #openstack | 16:43 | |
| *** javiF has quit IRC | 16:45 | |
| *** aliguori has quit IRC | 16:48 | |
| *** darraghb has joined #openstack | 16:48 | |
| *** llang629 has joined #openstack | 16:53 | |
| *** nerens has joined #openstack | 16:57 | |
| *** mdomsch has quit IRC | 16:59 | |
| *** vish1 is now known as vishy | 17:00 | |
| tillmo | I have a dashboard problem. I get the main page, but when loggin in, I get "DoesNotExist at /accounts/login/" | 17:01 |
| *** g2 has quit IRC | 17:01 | |
| *** bencherian has quit IRC | 17:01 | |
| vidd | did you set up the keystone database with dashboadr endpoints? | 17:02 |
| tillmo | probably not... with keystone, I get another problem: "keystone-manage token add 999888777666 user user 2015-02-05T00:00" gives "ERROR: 'NoneType' object has no attribute 'id'" | 17:03 |
| tillmo | I use the latest keystone version from git | 17:03 |
| mjfork | does user "user" exist? | 17:04 |
| *** rods has joined #openstack | 17:04 | |
| mjfork | if you paste the stack trace I can probably point you to what is missing | 17:04 |
| tillmo | I added "user" with "keystone-manage user add user test". How can I check whether this succeeded? | 17:05 |
| *** llang629 has left #openstack | 17:05 | |
| *** hggdh has quit IRC | 17:05 | |
| vidd | keystone-manager user list | 17:06 |
| *** OutBackDingo has quit IRC | 17:06 | |
| vidd | also, when you do "keystone-manage token list" look at the order displayed | 17:07 |
| tillmo | keystone-manager user list gives "1user1None" | 17:07 |
| *** Oneiroi has quit IRC | 17:07 | |
| vidd | mine shows "token user expiration tenant" | 17:08 |
| tillmo | yes, I have omitted that line | 17:08 |
| *** lorin11 has joined #openstack | 17:08 | |
| tillmo | ah, you mean keystone-manage token list | 17:08 |
| vidd | what does "keystone-manage token list" look at the order displayed | 17:08 |
| tillmo | tokenuserexpirationtenant | 17:08 |
| tillmo | aha, I see | 17:08 |
| mjfork | can you get the stack trace? | 17:09 |
| *** cp16net has joined #openstack | 17:09 | |
| vidd | i'm going to have to write the billing API for this so I have been paying VERY close attention to what does what in the keystone database | 17:09 |
| vidd | mjfork, dont need the stack trace | 17:10 |
| tillmo | keystone-manage token add 999888777666 user user 2015-02-05T00:00 | 17:10 |
| tillmo | /usr/lib/pymodules/python2.6/paste/deploy/loadwsgi.py:8: UserWarning: Module keystone was already imported from /root/keystone2/keystone/__init__.pyc, but /usr/lib/python2.6/dist-packages is being added to sys.path | 17:10 |
| tillmo | import pkg_resources | 17:10 |
| tillmo | ERROR: 'NoneType' object has no attribute 'id' | 17:10 |
| tillmo | Traceback (most recent call last): | 17:10 |
| tillmo | File "bin/keystone-manage", line 16, in <module> | 17:10 |
| tillmo | keystone.manage.main() | 17:10 |
| tillmo | File "/root/keystone2/keystone/manage/__init__.py", line 288, in main | 17:10 |
| tillmo | raise exc | 17:10 |
| tillmo | AttributeError: 'NoneType' object has no attribute 'id' | 17:10 |
| tillmo | root@nova-controller:~/keystone2# bin/keystone-manage token list | 17:10 |
| tillmo | /usr/lib/pymodules/python2.6/paste/deploy/loadwsgi.py:8: UserWarning: Module keystone was already imported from /root/keystone2/keystone/__init__.pyc, but /usr/lib/python2.6/dist-packages is being added to sys.path | 17:10 |
| tillmo | import pkg_resources | 17:10 |
| tillmo | tokenuserexpirationtenant | 17:10 |
| vidd | the issue the documentation has not been updated to match the latest version | 17:10 |
| *** hggdh has joined #openstack | 17:11 | |
| tillmo | I have changed the order "keystone-manage token add 999888777666 user 2015-02-05T00:00 user", but I get the same error | 17:11 |
| *** lorin1 has quit IRC | 17:11 | |
| *** lorin11 has quit IRC | 17:11 | |
| mjfork | thats good.that stack trave doesn't look right | 17:11 |
| mjfork | can you paste the log to pastebin? | 17:11 |
| *** lorin1 has joined #openstack | 17:11 | |
| vidd | is "user" part of the Admin role? | 17:11 |
| mjfork | whoops - ignore "thats good" | 17:11 |
| *** lorin1 has quit IRC | 17:11 | |
| mjfork | thats not the right stacktrace | 17:12 |
| *** obino has joined #openstack | 17:13 | |
| tillmo | http://pastebin.com/p0nkKMWR | 17:13 |
| mjfork | sorry - from the keystone log | 17:14 |
| mjfork | not the client | 17:14 |
| vidd | ive been working on this database for 3 days...i have a very good handle on what needs to be done to what] | 17:14 |
| vidd | i dont need the stacktrace....i know the issue | 17:15 |
| tillmo | in /var/log I cannot see anything... | 17:15 |
| tillmo | where is this log? | 17:15 |
| vidd | tillmo, you want the answer or do you want to hunt for log taces? | 17:15 |
| mjfork | not sure - look at keystone.conf and see where it poitns | 17:15 |
| vidd | what does tenant list say | 17:16 |
| tillmo | idnameenabled | 17:17 |
| tillmo | and nothing listed... | 17:17 |
| vidd | so you missed the first field | 17:17 |
| tillmo | sorry, I do not understand | 17:17 |
| mjfork | keystone-manage tenant add user | 17:18 |
| vidd | the database needs to be filled out in order | 17:18 |
| vidd | tenant add name_of_tenant | 17:18 |
| *** negronjl has quit IRC | 17:18 | |
| tillmo | I see | 17:18 |
| vidd | user add name password tenant_name | 17:19 |
| tillmo | ah, now I hace successfully created the token. Thanks! | 17:19 |
| vidd | you created the token but your keystone database is most likely fizzled | 17:20 |
| *** amiranda has quit IRC | 17:20 | |
| tillmo | how should I proceed? | 17:20 |
| vidd | delete the database and start ove | 17:20 |
| tillmo | OK, will do so and come back later | 17:21 |
| *** adrianmir has quit IRC | 17:21 | |
| vidd | but if this is not for production, then it can wait | 17:21 |
| vidd | i'll be doing some in-depth documentation | 17:21 |
| tillmo | good! | 17:21 |
| tillmo | it is not for production | 17:21 |
| *** marrusl has quit IRC | 17:21 | |
| vidd | ive been pounding away at this database for a few days so i know it pretty well =] | 17:22 |
| *** aliguori has joined #openstack | 17:22 | |
| *** negronjl has joined #openstack | 17:24 | |
| tillmo | I have now set OPENSTACK_ADMIN_TOKEN and OPENSTACK_KEYSTONE_URL in local/local_settings.py, started keystone, but still get "DoesNotExist at /accounts/login/" when logging into dashboard | 17:24 |
| vidd | thats because you have no endpoint enabled for dashboard , you have no admin role | 17:25 |
| *** darraghb has quit IRC | 17:25 | |
| vidd | so dashboard has no clue who you are | 17:26 |
| tillmo | I have an admin role, but how do I assign it to user "user"? | 17:26 |
| mjfork | vidd: is it "Admin" or "admin"? | 17:26 |
| tillmo | admin | 17:27 |
| vidd | keystone-manage role grant Admin user | 17:27 |
| mjfork | he first needs to keystone-manage role create Admin right? | 17:27 |
| mjfork | admin != Admin | 17:27 |
| vidd | he said he had Admin | 17:28 |
| tillmo | I have grant both admin and Admin to user, problem remains | 17:29 |
| *** vernhart has joined #openstack | 17:29 | |
| vidd | really, its not that critical....the endpoint is what matters | 17:29 |
| tillmo | my OPENSTACK_KEYSTONE_URL is http://192.168.178.224:5000/v2.0/, and under this address, I get an XML document | 17:30 |
| vidd | what do you have listed under "service list" | 17:30 |
| tillmo | idnametype | 17:31 |
| *** guigui1 has quit IRC | 17:32 | |
| vidd | "service add dash Dash OpenStackDashboard" or something similar [the first section is what's important] | 17:33 |
| vidd | now i need to get the endpoint_template info | 17:34 |
| *** Ryan_Lane has joined #openstack | 17:35 | |
| tillmo | how do I get that? | 17:35 |
| vidd | im launching my devstack vm | 17:36 |
| vidd | give it a minute to load so i can get the info | 17:36 |
| *** adjohn has quit IRC | 17:36 | |
| *** bengrue has joined #openstack | 17:38 | |
| *** hub_cap has joined #openstack | 17:40 | |
| vidd | this devstack has NO database discription -\ | 17:40 |
| tillmo | :-( | 17:41 |
| *** robix has joined #openstack | 17:44 | |
| vidd | try user and the password for user | 17:45 |
| tillmo | I have found https://github.com/4P/deployscripts/blob/master/nova.sh I will try to use this | 17:45 |
| tillmo | I have tried user and the passwd, same problem | 17:46 |
| vidd | did you manually add dashboard or did you run a master script that set everything up | 17:46 |
| tillmo | I did it manually | 17:46 |
| vidd | ok...what is the endpoint info for dashboard? | 17:47 |
| tillmo | OPENSTACK_KEYSTONE_URL = "http://192.168.178.224:5000/v2.0/" | 17:48 |
| tillmo | or do you mean the dashboard endpoint? | 17:48 |
| vidd | the dashboard endpoint | 17:48 |
| tillmo | NOVA_DEFAULT_ENDPOINT = 'http://localhost:8773/services/Cloud' | 17:48 |
| vidd | that is nova | 17:48 |
| vidd | what is the ip of dash + the port | 17:49 |
| tillmo | 192.168.178.224:8001 | 17:49 |
| vidd | there we go.... | 17:50 |
| tillmo | no, this is a local IP | 17:50 |
| tillmo | I will make a public one... | 17:50 |
| *** adjohn has joined #openstack | 17:51 | |
| tillmo | http://95.33.105.52:8001/ | 17:51 |
| tillmo | user: user, password: test | 17:52 |
| vidd | endpointTemplates add RegionOne dash http://95.33.105.52:8001 http://192.168.178.224:8001 http://192.168.178.224:8001 1 1 | 17:54 |
| *** bencherian has joined #openstack | 17:54 | |
| tillmo | ERROR: (OperationalError) table endpoint_templates has no column named version_id | 17:55 |
| vidd | endpointTemplates add RegionOne dash http://95.33.105.52:8001 http://192.168.178.224:8001 http://192.168.178.224:8001 1 1 ver | 17:55 |
| tillmo | same error | 17:56 |
| *** mattray has joined #openstack | 17:56 | |
| *** praefect has quit IRC | 17:57 | |
| *** koolhead17 has quit IRC | 17:57 | |
| tillmo | although the arguments seem to be OK. I leave out one of them, I get "Missing arguments: endpointTemplates add 'region' 'service' 'publicURL' 'adminURL' 'internalURL' 'enabled' 'global'" | 17:57 |
| *** bcwaldon has quit IRC | 17:59 | |
| vidd | endpointTemplates add RegionOne dash http://95.33.105.52:8001/ver/%tenant_id% http://192.168.178.224:8001ver/%tenant_id% http://192.168.178.224:8001ver/%tenant_id% 1 1 | 18:00 |
| tillmo | sorry, same error | 18:01 |
| vidd | make sure it says 8001/ver/%tenant_id% after each URL | 18:01 |
| tillmo | do you mean to type in "ver" literally? | 18:01 |
| vidd | yes | 18:02 |
| vidd | /ver/%tenant_id% | 18:02 |
| tillmo | but %tenant_id% replaced with the real id of the tenant | 18:02 |
| vidd | no | 18:02 |
| vidd | with %tenant_id% | 18:03 |
| tillmo | still the same error | 18:03 |
| vidd | /v1.1/%tenant_id% | 18:03 |
| vidd | or find out what version of dashboard you have | 18:03 |
| * vidd hasnt gotten around to dashboard yet =\ | 18:05 | |
| *** FabriceB has quit IRC | 18:05 | |
| tillmo | keystone-manage endpointTemplates add RegionOne dash http://95.33.105.52:8001/v1.1/%tenant_id% http://192.168.178.224:8001/v1.1/%tenant_id% http://192.168.178.224:8001/v1.1/%tenant_id% 1 1 | 18:06 |
| tillmo | did not work | 18:06 |
| *** tyska has joined #openstack | 18:06 | |
| vidd | whats the eror? | 18:06 |
| *** timr1 has quit IRC | 18:06 | |
| tyska | hi guys | 18:06 |
| tillmo | ERROR: (OperationalError) table endpoint_templates has no column named version_id u'INSERT INTO endpoint_templates (region, service_id, public_url, admin_url, internal_url, enabled, is_global, version_id, version_list, version_info) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?)' ('RegionOne', 1, 'http://95.33.105.52:8001/v1.1/%tenant_id%', 'http://192.168.178.224:8001/v1.1/%tenant_id%', 'http://192.168.178.224:8001/v1.1/%tenant_id%', '1 | 18:06 |
| tillmo | ', '1', None, None, None) | 18:06 |
| tyska | i need help to install a test-environment with stackops distro | 18:06 |
| vidd | there is a 1 between regionOne and the first URL.... | 18:08 |
| vidd | remove that and try again | 18:08 |
| tillmo | hm, there is no 1, see my command above | 18:09 |
| tyska | i have basically two doubts: 1) i have 2 HDD's in my 2 servers? How i must partition they? 2) How i must install and configure the servers using the SERVER_IP:8888 URL, i installed the node controller and entered the URL, but dont know what i have to do next. | 18:09 |
| tillmo | the problem seems to be more fundamental: the keystone code speaks about a column version_id, but the database table does not have such a column | 18:10 |
| *** mencken has joined #openstack | 18:10 | |
| *** nelson has quit IRC | 18:11 | |
| tyska | can anyone help me? | 18:11 |
| tillmo | need to go offline for 5 minutes... | 18:12 |
| vidd | ok tillmo | 18:12 |
| vidd | tyska, you have two hard drives | 18:13 |
| tyska | vidd: yeah | 18:13 |
| vidd | when you get to partitioning, select "manual" | 18:13 |
| tyska | vidd: i created 1 ext4 root(32GB) 1 swap (8.1GB) 1 LVM 460GB | 18:14 |
| tyska | vidd: and created 1 LVM 500GB in the other hard drive | 18:14 |
| vidd | good | 18:14 |
| tyska | vidd: can be done with this scheme? | 18:14 |
| vidd | yes | 18:14 |
| tyska | vidd: ok, then i entered the URL to configure the server1 - the controller node | 18:15 |
| vidd | is the install complete? | 18:15 |
| vidd | tyska, did you complete the cd install | 18:16 |
| tyska | vidd: yeah, i selected the /dev/sdb1 (the 2nd hd) as the lvm volume to put the volumes of the volumes controller | 18:16 |
| *** tillmo has quit IRC | 18:16 | |
| vidd | then you are good to go | 18:16 |
| tyska | vidd: yeah, i entered the installation url after the cd instalation has completed | 18:16 |
| tyska | vidd: my doubt now is how i set the compute node to this node controller | 18:17 |
| tyska | vidd: and how i configure the interfaces of both server | 18:17 |
| vidd | did you register with stackops? | 18:17 |
| tyska | vidd: yeah | 18:17 |
| vidd | did you select the multi-node config? | 18:17 |
| tyska | vidd: no | 18:18 |
| tyska | vidd: i want to do dual node | 18:18 |
| *** mgius-away is now known as mgius | 18:18 | |
| vidd | the site is pretty self explainitory | 18:18 |
| *** adjohn has quit IRC | 18:18 | |
| tyska | vidd: yeah, but some points are still 'cloud' =) | 18:18 |
| tyska | vidd: for example, eth1, that is the if i will use to comunicate with compute node, does not have any configuration | 18:19 |
| *** tillmo has joined #openstack | 18:19 | |
| vidd | answer the questions on the screen while you put the disk (or whatever) into the other server and choose compute node install] | 18:19 |
| tyska | vidd: eth1=null is showed | 18:19 |
| vidd | does your machine have 1 nic or 2? | 18:19 |
| tyska | 2 | 18:19 |
| tyska | i think it could be 10.0.0.0/24 to the public if and 192.168.0.0/24 to the private if | 18:20 |
| vidd | then eth1 should not be showing =null | 18:20 |
| tyska | vidd: yeah, i thought it was weird | 18:21 |
| *** koolhead17 has joined #openstack | 18:21 | |
| tyska | vidd: then i decide get some help here | 18:21 |
| *** Horroreyes has joined #openstack | 18:21 | |
| tyska | vidd: because im stucked in this part | 18:21 |
| vidd | are both nics plugged in? | 18:21 |
| tyska | tyska: yeah | 18:22 |
| vidd | here isnt really where you should be going....you should be going to stackops | 18:22 |
| tyska | vidd: hmm, you're right | 18:22 |
| *** zul has quit IRC | 18:22 | |
| vidd | they are a private company | 18:22 |
| vidd | there should be some info on thier web site | 18:23 |
| tyska | vidd: in fact, im migrating from eucalyptus to openStack | 18:23 |
| tyska | vidd: then im just starting | 18:23 |
| vidd | actually, sounds like your migrating to stackops | 18:24 |
| vidd | =] | 18:24 |
| tyska | vidd: but StackOps and openstack does not are the same? | 18:24 |
| tyska | vidd: i thought stackops was just a distro with stackops embbed | 18:24 |
| vidd | dunno...never looked under thier hood | 18:24 |
| tyska | hmm | 18:25 |
| tyska | vidd: then, whats the best way to use stack ops with ubuntu? | 18:25 |
| vidd | i know thier configuration utility is proprietary...so i stay away | 18:25 |
| tyska | vidd: i must use 10.04? or 11.10? | 18:25 |
| vidd | this isnt stackops | 18:25 |
| vidd | this is open stack | 18:25 |
| *** zul has joined #openstack | 18:26 | |
| tillmo | Stackops comes with its own cd that include Ubuntu | 18:26 |
| vidd | i would recommend using 10.04 | 18:26 |
| vidd | and follow the info on http://docs.openstack.org/diablo/openstack-compute/admin | 18:27 |
| vidd | i recommed the LTS because you need to add the PPA's to the repo's and the built-in packages bundled with 11.10 are not what i would call thier finest work | 18:29 |
| vidd | cant really blame the *buntu guys for that....the newest openstack was released less than a month befor 11.10 was released | 18:29 |
| WormMan | of course, if you use 10.04 with KVM your performance is going to be rather sad. | 18:31 |
| *** HowardRoark has quit IRC | 18:31 | |
| WormMan | but plenty for testing and getting to know openstack | 18:31 |
| *** tyska has quit IRC | 18:32 | |
| vidd | WormMan, once openstack is configured from ppa's, and running, do-release-upgrade will NOT break it | 18:32 |
| *** thingee has joined #openstack | 18:33 | |
| vidd | [right away] | 18:33 |
| crussell | Oh boy, here goes... cluster successfully setup. | 18:34 |
| crussell | gluster fs that is.. | 18:34 |
| crussell | time for nova install...again | 18:34 |
| vidd | what? you didnt export your /etc ? | 18:34 |
| vidd | or did you istall from all those scripts? | 18:35 |
| *** HowardRoark has joined #openstack | 18:36 | |
| uvirtbot | New bug: #879005 in openstack-ci "Documentation should switch "hostname" for actual hostname" [Undecided,New] https://launchpad.net/bugs/879005 | 18:38 |
| *** imsplitbit has quit IRC | 18:42 | |
| *** ewindisch_ has joined #openstack | 18:42 | |
| *** ewindisch has quit IRC | 18:45 | |
| *** ewindisch_ is now known as ewindisch | 18:45 | |
| *** mjfork has quit IRC | 18:49 | |
| *** mjfork has joined #openstack | 18:49 | |
| *** mjfork has quit IRC | 18:51 | |
| *** tyska has joined #openstack | 18:52 | |
| tyska | vidd: im now installing ubuntu 10.04 to put OpenStack =) | 18:52 |
| tyska | vidd: i need some advice | 18:52 |
| tyska | vidd: now i configured the 2 hd's with RAID1 | 18:53 |
| *** timr has joined #openstack | 18:53 | |
| vidd | sorry...cant help yopu there | 18:53 |
| *** Susanne-Balle has joined #openstack | 18:53 | |
| tyska | vidd: i already did it | 18:53 |
| tyska | vidd: i just wanna know, if i can put the volumes on a logical volume of the LVM | 18:54 |
| vidd | i dont touch raid installs | 18:54 |
| tyska | vidd: or it's required a volume group of LVM | 18:54 |
| *** mjfork has joined #openstack | 18:54 | |
| *** dendrobates is now known as dendro-afk | 18:54 | |
| tyska | vidd: logical volume or group volume? | 18:55 |
| vidd | i do not know | 18:55 |
| *** pixelbeat has quit IRC | 18:55 | |
| vidd | i am not the person to answer that question | 18:55 |
| *** GeoDud has joined #openstack | 18:55 | |
| tyska | vidd: ok, i will figure out here then share with you =) | 18:56 |
| vidd | personally, i only use LVMs | 18:56 |
| *** xmltok has joined #openstack | 18:56 | |
| *** markwash has joined #openstack | 18:57 | |
| *** dendro-afk is now known as dendrobates | 18:58 | |
| *** mjfork has quit IRC | 18:59 | |
| *** mjfork has joined #openstack | 19:00 | |
| *** bcwaldon has joined #openstack | 19:02 | |
| *** Susanne-Balle has quit IRC | 19:02 | |
| tillmo | tyska: logical volume should work | 19:03 |
| vidd | tillmo, they want to use raid1 | 19:04 |
| vidd | i aint touching it | 19:04 |
| tillmo | I am using LVM with raid10 | 19:05 |
| vidd | whatever....im not touching it | 19:06 |
| vidd | im like the rouches...i dont like raid | 19:06 |
| vidd | =] | 19:06 |
| tyska | vidd: the RAID1 is implemented by hardware | 19:07 |
| tyska | vidd: linux just see 1 hd now | 19:08 |
| vidd | doesnt matter | 19:08 |
| tyska | tillmo: i created several partitions, being just 1 of them a LVM group to the volumes | 19:08 |
| *** vladimir3p has quit IRC | 19:08 | |
| *** justinlw has quit IRC | 19:09 | |
| *** tacitone has joined #openstack | 19:09 | |
| tacitone | Hi all, I have a dumb question: I just installed Swift and used st to upload a file to a container | 19:09 |
| tacitone | Do I need to install Nova or Glance in order to share out the resources of Swift, or can I set up a samba/nfs mount and start sharing out node resources right away? | 19:10 |
| notmyname | tacitone: there is no samba/nfs interface included with swift | 19:10 |
| Ryan_Lane | is there any way to use the UEC images in such a way that I can upgrade the kernel inside of the instance? | 19:10 |
| notmyname | swift isn't a file system | 19:10 |
| Ryan_Lane | instead of having to rebuild all my instances every time I need to upgrade a kernel? | 19:11 |
| notmyname | tacitone: you can grant permissions to other users so they can access the file (depending on the auth system you are using) | 19:11 |
| tacitone | Are VMs (LXC, KVM, etc.) able to access Swift nodes as a file system? | 19:11 |
| alexn6 | tacitone: better try to use cyberduck - if it is for users | 19:12 |
| notmyname | tacitone: not without some other 3rd party software (like a fuse layer). and I wouldn't recommend it. performance would be bad | 19:12 |
| Ryan_Lane | tacitone: unless there is some fuse filesystem, then no | 19:12 |
| vidd | Ryan_Lane, do you have auto-updates turnrd on? | 19:12 |
| tacitone | I had it in my head that the whole point of using Swift was to set up a redundant file system pool. | 19:12 |
| Ryan_Lane | tacitone: what you are likely looking for is something like ceph or gluster | 19:12 |
| Ryan_Lane | vidd: in the instance? | 19:12 |
| vidd | yes | 19:12 |
| Ryan_Lane | vidd: I can install new kernels, but the instance will always reboot into the old one | 19:12 |
| notmyname | tacitone: swift is for unstructured data that can grow without bound. like backups or static web files | 19:12 |
| notmyname | tacitone: swift is directly comparable to s3 | 19:13 |
| alexn6 | Ryan_Lane: why don`t you use image with external kernel defined with eki? | 19:13 |
| Ryan_Lane | alexn6: that sounds good. how do I go about that? | 19:13 |
| vidd | alexn6, you dont HAVE to seperate out the kernel?!? | 19:14 |
| Ryan_Lane | I know that openstack has support for it | 19:15 |
| Ryan_Lane | I just don't know how to use it | 19:15 |
| Ryan_Lane | docs or it doesn't exist is the new pics or it didn't happen :) | 19:15 |
| vidd | It is officiall....the new keystone database id a [(*&)&^%$&*(_(*&(*^&^&%$^&%$#^&%&*(^*(&^ | 19:15 |
| mjfork | vidd is it possibel to set up an admin in keystone that has no tenant? | 19:16 |
| alexn6 | Ryan_Lane: as I understand - you just need publish your image aka ami bundled with eki and eri (kernel and ramdisk). It`s rather complicated when using ec2-like API(AWS style). But with glance you can do it just with a single "click" :) | 19:16 |
| Ryan_Lane | I'm using glance | 19:16 |
| alexn6 | vidd: didn`t understand you | 19:16 |
| alexn6 | so with glance image can be altered with glance update cmd | 19:16 |
| vidd | alexn6, i thought you had to seperate the kernel from the ramdisk | 19:17 |
| alexn6 | just load another kernel in glance and update to a new eki | 19:17 |
| Glacee | In the swift 1.4.3 documentation it says to use UBUNTU 10.4 LTS(Lucid) and run apt-gets to get the packages.. but with that disto you get swift 1.20 | 19:17 |
| vidd | mjfork, yes' | 19:17 |
| Glacee | just saying | 19:17 |
| *** hurrycane has joined #openstack | 19:17 | |
| alexn6 | vidd: yes - like it always done - kernel and initrd | 19:18 |
| vidd | Glacee, you want to add the ppa's to 10.4-11.4 | 19:18 |
| vidd | get all the latest | 19:18 |
| hurrycane | Hello @all, it is safe to use the ubuntu 11.10 packages for openstack. Are the keystone/dashboard pkgs working? Is there a tutorial/example ? | 19:18 |
| *** justinlw has joined #openstack | 19:18 | |
| Ryan_Lane | alexn6: so, I just change the kernel-name property? | 19:19 |
| Ryan_Lane | to a new one? | 19:19 |
| Ryan_Lane | then reboot the instances? | 19:19 |
| vidd | hurrycane, no to keystone and dashboard packages | 19:19 |
| tillmo | tyska: I have installed openstack using two logical volumes, one for the file system root, one for the swap. That is also what stackops does | 19:19 |
| hurrycane | @vidd :( | 19:19 |
| larissa | hurrycane: Error: "vidd" is not a valid command. | 19:19 |
| alexn6 | Ryan_Lane: I think this would work | 19:19 |
| Ryan_Lane | ah. cool | 19:19 |
| Glacee | cool vidd | 19:19 |
| tyska | tillmo: i abandoned stackops | 19:19 |
| tyska | tillmo: know im using ubuntu 10.04 =) | 19:20 |
| tyska | tillmo: but im having the following problem now | 19:20 |
| Ryan_Lane | we'll see if that works soon enough. heh | 19:20 |
| tillmo | yes, but it should also work for Ubuntu 10.04 | 19:20 |
| tyska | tillmo: when i run sudo add-apt-repository ppa:openstack-release/2011.3 | 19:20 |
| vidd | hurrycane, and the database for this new keystone in nasty | 19:20 |
| tyska | tillmo: i get a HTTP fetch error 7: couldn't connect to host | 19:20 |
| notmyname | Glacee: use either the official openstack ppas or you can use the packaging we use at rackspace (http://crashsite.github.com/swift_debian/) | 19:20 |
| vidd | im going to have to write up some new documentations | 19:20 |
| tyska | how can i fix this? | 19:20 |
| hurrycane | @vidd I've managed to run keystone & dashboard on ubuntu 11.04 but in the end I found that Glance was failing too | 19:20 |
| larissa | hurrycane: Error: "vidd" is not a valid command. | 19:20 |
| alexn6 | who can help with Glance >5G files? I`m using cactus release and think that migration all system would be complicated -can I somehow upgrade only Glance? | 19:20 |
| tyska | vidd: glance is missing in ubuntu 10.04, apt-get install glance does not work | 19:21 |
| alexn6 | Ryan_Lane: say please if it works. | 19:22 |
| vidd | tyska, did you add the ppa's? | 19:22 |
| *** imsplitbit has joined #openstack | 19:22 | |
| tillmo | tyska: first correct the other problem | 19:22 |
| hurrycane | is there a safe way to get keystone and dashboard and diablo working ? | 19:22 |
| Glacee | thanks notmyname | 19:22 |
| tillmo | tyksa, could you paste /etc/apt/sources.list.d/openstack-release-2011.2-lucid.list | 19:22 |
| tyska | vidd: i tried | 19:22 |
| tillmo | sorry /etc/apt/sources.list.d/openstack-release-2011.3-lucid.list | 19:22 |
| tyska | vidd: but i get that error | 19:22 |
| vidd | huh? | 19:23 |
| gnu111 | how can i change user password in swift? | 19:23 |
| tyska | vidd: when i run sudo add-apt-repository ppa:openstack-release/2011.3 | 19:23 |
| notmyname | gnu111: passwords are handled by an auth system. changing it is dependent on the auth system | 19:23 |
| tyska | vidd: i get a HTTP fetch error 7: couldn't connect to host | 19:23 |
| tyska | tillmo: deb http://ppa.launchpad.net/openstack-release/2011.3/ubuntu lucid main | 19:24 |
| tyska | tillmo: there is just this line | 19:24 |
| tillmo | looks fine | 19:25 |
| vidd | the command to add the ppa's is | 19:25 |
| vidd | sudo add-apt-repository ppa:openstack-release/2011.3 | 19:25 |
| gnu111 | notmyname: I followed the guide here: http://swift.openstack.org/howto_installmultinode.html so not sure what is the auth system by default. | 19:25 |
| tyska | vidd: that was i did, no? | 19:25 |
| notmyname | gnu111: do you see "tempauth" in your proxy config pipeline? | 19:25 |
| tillmo | tyksa, try: ping ppa.launchpad.net | 19:26 |
| gnu111 | notmynames: ah, yes. /etc/swift/proxy-server.conf | 19:26 |
| tyska | tillmo: success | 19:26 |
| vidd | tyska, my comment was to tillmo | 19:26 |
| tyska | vidd: ah, sorry | 19:26 |
| uvirtbot | New bug: #879044 in nova "single xenapi session causes concurrency issues" [High,New] https://launchpad.net/bugs/879044 | 19:26 |
| *** dendrobates is now known as dendro-afk | 19:26 | |
| notmyname | gnu111: if you are using tempauth, it's really easy to change a password. the password is stored in the proxy config in the filter:tempauth section | 19:26 |
| gnu111 | notmyname: what other option do I have for auth? I haven't connected nova to swift yet, can I use the nova authorization in swift? | 19:27 |
| notmyname | gnu111: honestly, I'm not sure what "nova authorization" means | 19:27 |
| notmyname | gnu111: check out http://swift.openstack.org/overview_auth.html for a start on what the auth system can do | 19:28 |
| notmyname | gnu111: using tempauth will not allow you to use the same credentials between nova and swift | 19:28 |
| tyska | tillmo: vidd: im following this guide | 19:28 |
| tyska | tillmo: vidd: but when it fail to install glance, i went to the openstack website | 19:28 |
| gnu111 | notmyname: ok. that's what I wanted to confirm if I can use the same cred between nova and swift. | 19:29 |
| tillmo | tyska: you have to solve your HTTP ftach error 7 before you can install glance | 19:29 |
| tillmo | try wget http://ppa.launchpad.net/openstack-release/2011.3/ubuntu/dists/lucid/main/binary-amd64/Packages | 19:30 |
| tyska | tillmo: yeah, i realize that, but how? | 19:30 |
| notmyname | gnu111: you may be able to use keystone for that, but keystone (IIRC) doesn't yet support all the authorization functionality for swift that tempauth does | 19:30 |
| tyska | tillmo: where i run this? | 19:30 |
| tyska | tillmo: to what directory? | 19:30 |
| tillmo | in a terminal, say in /tmp | 19:30 |
| tyska | tillmo: ok, success, what now? | 19:31 |
| tillmo | which command leads to fetch error 7 ? | 19:31 |
| tyska | tillmo: sudo add-apt-repository ppa:openstack-release/2011.3 | 19:32 |
| gnu111 | notmynameL regarding swift-bench. I tried this: swift-bench -A https://$PROXY_LOCAL_NET_IP:8080/auth/v1.0 -U | 19:32 |
| gnu111 | system:root -K pass -c 10 -s 2 -n 100 -g 10 | 19:32 |
| gnu111 | and got Container PUT failed error. It seems to created an empty container. | 19:32 |
| Ryan_Lane | alexn6: that didn't seem to work | 19:33 |
| Ryan_Lane | alexn6: old image is now pointing to new kernel, but when rebooting the instance, it still uses the old kernel | 19:34 |
| tyska | tillmo: the complete error is | 19:34 |
| tyska | gpg: requesting key 3D1B4472 from hkp server keyserver.ubuntu.com | 19:34 |
| tyska | gpgkeys: HTTP fetch error 7: couldn't connect to host | 19:34 |
| tyska | gpg: no valid OpenPGP data found. | 19:34 |
| tyska | gpg: Total number processed: 0 | 19:34 |
| vidd | mjfork, to add credentials to a user that has no tenent, set tenant to 0 | 19:34 |
| alexn6 | Ryan_Lane: hmm rebooting - just by "reset" or restart? | 19:34 |
| tillmo | tyska, try: apt-key adv --recv-keys --keyserver keyserver.ubuntu.com 3D1B4472 | 19:34 |
| Ryan_Lane | alexn6: by using ec2 reboot command | 19:34 |
| Ryan_Lane | I'm pretty sure calling reboot in the instance will make it non-operational | 19:35 |
| alexn6 | Ryan_Lane: now, reboot of course not lead us to final goal :) Only restart instance | 19:35 |
| tyska | tillmo: same problem | 19:35 |
| Ryan_Lane | alexn6: you mean inside of the instance? | 19:35 |
| Ryan_Lane | last time I tried this, the instance didn't come back | 19:36 |
| tillmo | apt-key adv --recv-keys --keyserver wwwkeys.de.pgp.net 3D1B4472 | 19:36 |
| alexn6 | Ryan_Lane: no - with euca tool - stop then recreate | 19:36 |
| Ryan_Lane | ah | 19:36 |
| Ryan_Lane | that's exactly what I'm trying to avoid | 19:36 |
| alexn6 | oh then sorry | 19:36 |
| tyska | tillmo: same problem | 19:36 |
| alexn6 | but then It`s kinda impossible yet | 19:36 |
| Spirilis | whew thank god I'm backing up my .builder files to my other storage nodes, I just had a python traceback error while trying to rebalance the Swift rings after removing a device | 19:36 |
| Spirilis | restored my builder files from a backup copy and now I'm trying this again... | 19:37 |
| notmyname | gnu111: http://paste.openstack.org/show/2797/ | 19:37 |
| Ryan_Lane | there's really no way to upgrade a kernel in an instance? | 19:37 |
| tyska | tillmo: i can ping keyserver.ubuntu.com, but cant ping wwwkeys.de.pgp.net | 19:37 |
| tillmo | apt-key adv --recv-keys --keyserver gpg-keyserver.de 3D1B4472 | 19:37 |
| alexn6 | Ryan_Lane: nova use cache for images in local dir - it updates and reconfig all files and VM system only when start it. You may see the code or ask a question or for feature on launchpad | 19:38 |
| *** adjohn has joined #openstack | 19:39 | |
| *** PeteDaGuru has quit IRC | 19:39 | |
| Ryan_Lane | I could have sworn openstack made this possible in the cactus release | 19:39 |
| tyska | tillmo: error changed a little --- gpgkeys: HTTP fetch error 7: Failed to connect to 2a01:198:200:20a::2: Network is unreachable | 19:40 |
| tyska | tillmo: but i can ping this server | 19:40 |
| tillmo | tyska: you can also try without the key | 19:40 |
| gnu111 | notmyname: didn't work for me. something not set right. http://paste.openstack.org/show/2798/ | 19:40 |
| tyska | tillmo: how? | 19:40 |
| tillmo | just enter: apt-get update, and ignore the errors | 19:40 |
| tillmo | when you install glance, you need to confirm that you want to install unauthorized software | 19:41 |
| alexn6 | Ryan_Lane: in that case google definitely knows it :) | 19:41 |
| Ryan_Lane | it's not that I have a major issue re-creating instances, except I have a few with like 15 floating IPs assigned, and it is *really* annoying to have to reassociate all of those | 19:41 |
| notmyname | gnu111: do you have autocreate accounts turned on? (having it off may cause the 404) | 19:41 |
| tyska | tillmo: perfect! that's fit for me =) | 19:42 |
| *** mgoldmann has quit IRC | 19:42 | |
| Ryan_Lane | it just seems crazy to have to rebuild all my instances every time I need to do a security patch | 19:42 |
| Ryan_Lane | and re-do all the associations. | 19:42 |
| tyska | tillmo: are you think its better install following this http://cssoss.wordpress.com/2011/04/27/openstack-beginners-guide-for-ubuntu-11-04-installation-and-configuration/ | 19:42 |
| gnu111 | notmyname: in proxy-server.conf? | 19:42 |
| alexn6 | make a script :) | 19:42 |
| tyska | tillmo: or this: http://docs.openstack.org/diablo/openstack-compute/admin/content/installing-the-cloud-controller.html | 19:42 |
| Ryan_Lane | alexn6: some of these are being created by users | 19:42 |
| Ryan_Lane | and I can't script it | 19:43 |
| Ryan_Lane | because nova doesn't support DNS or updating mediawiki | 19:43 |
| Ryan_Lane | or updating puppet | 19:43 |
| notmyname | gnu111: under [app:proxy-server]: account_autocreate = true | 19:43 |
| Ryan_Lane | oh. good. reboot command inside of instances works now :) | 19:44 |
| Ryan_Lane | that's a plus | 19:44 |
| Ryan_Lane | still doesn't solve my problem. heh | 19:44 |
| gnu111 | notmyname: I only have allow_account_management = true | 19:45 |
| *** HowardRoark has quit IRC | 19:45 | |
| notmyname | ok, add the autocreate | 19:45 |
| notmyname | gnu111: ^ | 19:45 |
| gnu111 | notmyname: do i need to run swift-init in all the nodes? | 19:45 |
| notmyname | gnu111: just swift-init proxy reload | 19:46 |
| tyska | vidd: what do you think is better to guide me? the ubuntu beginners guide or the openstack instruction website? | 19:46 |
| notmyname | gnu111: (since just the proxy config changed) | 19:46 |
| alexn6 | Ryan_Lane: I see... Try to raise question on nova answers. I`ve said you already about how it is "inside" - information will be updated only at instance creation and even more - immediately cache it :) (just to not update anything more ;) ) | 19:46 |
| tillmo | tyska: I think the first one is easier because it is more comprehensive. Still the first one did not tell me how to get the dashboard running... the problem is keystone... | 19:46 |
| vidd | go with the beginners guide | 19:46 |
| Ryan_Lane | I thought I remembered that images can boot without kernel and initrd being external | 19:47 |
| gnu111 | notmyname: same error: Container PUT failed | 19:47 |
| notmyname | gnu111: ok, then look in syslog to see what's going on | 19:47 |
| tillmo | vidd: how can I create a fresh keystone db? | 19:47 |
| alexn6 | Ryan_Lane: oh yes, that possible | 19:47 |
| vidd | you can always come back to the openstack docs later | 19:47 |
| vidd | tillmo, what database manager do you use? | 19:48 |
| Ryan_Lane | I guess it would still just boot with whatever is inside of the image, then eh> | 19:48 |
| vidd | personally, I us phpmyadmin | 19:48 |
| Ryan_Lane | though it would be a locally cached image, and when it was updated would also update the kernel maybe? | 19:49 |
| gnu111 | notmyname: http://paste.openstack.org/show/2800/ there is this error: "python: abrt: detected unhandled Python exception in /usr/bin/swift-bench | 19:49 |
| gnu111 | " | 19:49 |
| alexn6 | Ryan_Lane: no, I think with single image if you update files inside it -they will store after reboot, but system can be booted from another inner kernel | 19:50 |
| tillmo | vidd: I use sqlitebrowser (just a test env with sqlite3) | 19:50 |
| alexn6 | it`s just a copy of cache with updated blocks | 19:50 |
| vidd | tillmo, then simply locate the file and delete it | 19:50 |
| tillmo | OK | 19:51 |
| vidd | when you run db_sync [and get the message] a fresh set of tables will be waiting for you to soil =] | 19:51 |
| *** timr has quit IRC | 19:52 | |
| Ryan_Lane | yeah. seems like this is an impossible task | 19:53 |
| Ryan_Lane | I get that instances are considered "non-persistent", but this is kind of a PITA | 19:54 |
| alexn6 | Ryan_Lane: do you have any image without external kernel/initrd? colud you try update and reboot inside instance? | 19:55 |
| Ryan_Lane | all uec images are external | 19:55 |
| Ryan_Lane | I haven't built an image from scratch | 19:55 |
| Ryan_Lane | I guess I can try that | 19:56 |
| alexn6 | i see. And looks like I didn`t manage to run any of that by myself. But there more simple thing - if KVM can it -than openstack can also(via libvirt) | 19:56 |
| Ryan_Lane | yep | 19:57 |
| Ryan_Lane | I guess I can wade through the code to see if it can upgrade kernels now. I doubt it though | 19:58 |
| alexn6 | I propose just do it with kvm instance :) | 19:59 |
| Ryan_Lane | it would be good to have a flag like "re-fetch kernel on instance reboot" | 19:59 |
| Toger | Does the VG for nova-volume *have* to be 'nova-volumes' or is that configurable? I don't see a config entry for it on http://docs.openstack.org/bexar/openstack-compute/admin/content/ch05s08.html# | 20:00 |
| alexn6 | :)) sound interesting | 20:00 |
| *** mx2323 has quit IRC | 20:00 | |
| Ryan_Lane | that would let you update images with new kernels | 20:00 |
| Ryan_Lane | and reboot into them | 20:00 |
| annegentle | Toger: you'll want the nova-volume (and nova-volumes) doc in docs.openstack.org/diablo/openstack-compute/admin/content/ for sure | 20:01 |
| notmyname | gnu111: that's odd. not sure what it means. you are using the latest version of the code? | 20:01 |
| *** Horroreyes has quit IRC | 20:03 | |
| annegentle | Toger: and I think the volume group does have to be named nova-volumes | 20:03 |
| *** dgags has quit IRC | 20:03 | |
| gnu111 | notmyname: i installed it with yum rpms. swift --version says 1.0 that's old right? | 20:03 |
| notmyname | gnu111: yes, probably (depends on the rpm packaging) | 20:03 |
| notmyname | gnu111: ah wait | 20:04 |
| notmyname | gnu111: that's the version of the swift CLI tool :-) | 20:04 |
| *** HowardRoark has joined #openstack | 20:04 | |
| tyska | tillmo: vidd: problem with the sudo pvcreate /dev/sda4 =( | 20:04 |
| gnu111 | notmyname: openstack-swift-1.4.3-b352.noarch | 20:04 |
| tyska | tillmo: vidd: the message was Device /dev/sda4 not found (or ignored by filtering). | 20:05 |
| notmyname | gnu111: ok, 1.4.3 is the current version | 20:05 |
| *** clauden has quit IRC | 20:05 | |
| tyska | tillmo: vidd: but fdisk -l shows me --- /dev/sda4 4870 60801 449265664 8e Linux LVM | 20:05 |
| *** clauden_ has joined #openstack | 20:05 | |
| vidd | tyska, please do not include me in your conversation....im working on documenting the las 4 days of working with this keystone databse | 20:05 |
| tyska | vidd: ok, sry | 20:06 |
| tyska | tillmo: i think it is because RAID | 20:06 |
| tyska | tillmo: because fdisk -l shows the sdb4 too | 20:06 |
| tyska | tillmo: but how i do a pvcreate in the raid volume? what is the right file? /dev/... ? | 20:07 |
| notmyname | gnu111: what's weird is that it does a HEAD on the account, gets a 204, then does a PUT to a container in that account and gets a 404. that should only happen if the account isn't found | 20:07 |
| gnu111 | notmyname: yes. and I see four new containers now with random text as the name. these are created each time I ran the bench tool. | 20:09 |
| tillmo | tyska: something in /dev/mapper | 20:09 |
| tyska | tillmo: there is no /dev/mapper | 20:09 |
| notmyname | gnu111: that makes sense. try create a container in that account with curl (so it's isolated and repeatable) and see what happens | 20:10 |
| tillmo | tyska: mdadm --detail --scan | 20:11 |
| gnu111 | notmyname: i used the root account. how do i create a new container: swift -A https://$PROXY_LOCAL_NET_IP:8080/auth/v1.0 -U system:root -K testpass upload myfiles bigfile1.tgz | 20:11 |
| gnu111 | ? | 20:11 |
| tillmo | or look at /dev/md??? | 20:12 |
| *** sandywalsh has quit IRC | 20:13 | |
| gnu111 | notmyname: "upload newcontainer filename" works. | 20:13 |
| *** adjohn has quit IRC | 20:15 | |
| tyska | tillmo: the command returns nothing | 20:15 |
| *** PiotrSikora has quit IRC | 20:15 | |
| *** PiotrSikora has joined #openstack | 20:16 | |
| *** bencherian has quit IRC | 20:16 | |
| tillmo | tyska: then you have a hardware raid? | 20:16 |
| tyska | tillmo: yeah | 20:16 |
| *** clauden__ has joined #openstack | 20:17 | |
| *** clauden_ has quit IRC | 20:17 | |
| *** hurrycane has quit IRC | 20:17 | |
| tillmo | tyska: but this should be transparent to the operating system, that is, you use /dev/sdx | 20:17 |
| tyska | tillmo: yeah, i think so too, but for some reason it is being showed in fdisk -l like 2 hd's | 20:18 |
| koolhead17 | openstack-dashboard is horizon now!! | 20:18 |
| tyska | tillmo: dont believe i will have to restart all the process =( | 20:18 |
| *** jdag has quit IRC | 20:19 | |
| *** sandywalsh has joined #openstack | 20:19 | |
| *** PotHix has quit IRC | 20:21 | |
| tillmo | grep filter /etc/lvm/lvm.conf | 20:22 |
| tyska | tillmo: i have to go | 20:23 |
| tyska | tillmo: see you later | 20:23 |
| *** dprince has quit IRC | 20:23 | |
| *** dendro-afk is now known as dendrobates | 20:23 | |
| *** tyska has quit IRC | 20:24 | |
| tillmo | see you. I think commenting out the filter in /etc/lvm/lvm.conf should work | 20:24 |
| *** debo_os has joined #openstack | 20:27 | |
| gnu111 | notmyname: i posted the issue in mailing list. I will add the log details as well. | 20:31 |
| notmyname | gnu111: sorry. was called away from my desk | 20:32 |
| Kiall | Heya, Trying to get keystone setup .. and it looks like I'm nearly there (I can get tokens etc with curl).. How do i then go about getting credentials for use with the `nova` command? | 20:36 |
| *** Shentonfreude1 has joined #openstack | 20:36 | |
| tillmo | vidd: I managed to solve the keystone problem by creating a new database | 20:39 |
| vidd | awesome | 20:40 |
| *** Shentonfreude has quit IRC | 20:40 | |
| vidd | any issues setting up the database? | 20:40 |
| vidd | or did you just populate the sample date | 20:40 |
| tillmo | I just populated some sample data | 20:40 |
| vidd | and you can get into keystone? | 20:41 |
| vidd | tillmo is a cheater =] | 20:41 |
| tillmo | I can run keystone, at least | 20:42 |
| vidd | how about dashboard? | 20:42 |
| tillmo | vidd: then I managed to solve the dashboard problem by inserting a data value into the table django_site | 20:42 |
| tillmo | now I get a nova error... | 20:42 |
| vidd | thats because you are using sample data in keystone and real data in nova | 20:43 |
| tillmo | probably also because of NOVA_DEFAULT_ENDPOINT = 'http://localhost:8773/services/Cloud' in the dashboard settings | 20:43 |
| tillmo | or is this OK? | 20:43 |
| vidd | that should be good | 20:44 |
| vidd | its the internal value | 20:44 |
| vidd | assuming dash is on the same server as nova | 20:45 |
| tillmo | which is the case | 20:45 |
| vidd | oh... tillmo can you pastebin me your database ? | 20:46 |
| tillmo | which one? | 20:46 |
| vidd | at least the info for dashboad ... keystone | 20:46 |
| tillmo | I mean, there is a dashboard database, and a keystone database. I will post both | 20:47 |
| *** nhubbard is now known as nhubbard_away | 20:48 | |
| tillmo | vidd: dashboard: http://pastebin.com/VPnmbA6K and keystone: http://pastebin.com/wuJui7qt | 20:52 |
| *** localhost has quit IRC | 20:52 | |
| tillmo | I just see that the user names are different... | 20:52 |
| *** localhost has joined #openstack | 20:53 | |
| *** miclorb_ has joined #openstack | 20:54 | |
| tillmo | I have now added a user "till" for keystone, same password "test" as for dashboard, but same error | 20:55 |
| *** salv-orlando has quit IRC | 20:56 | |
| tillmo | at least, I can sign in and sign out at the dashboard... | 20:56 |
| vidd | i was right about the infor for dashboard =] | 20:57 |
| tillmo | which one? | 20:57 |
| vidd | the the endpoint template | 20:58 |
| *** lts has quit IRC | 20:58 | |
| *** rnirmal has quit IRC | 20:58 | |
| *** miclorb_ has quit IRC | 20:58 | |
| *** debo_os has quit IRC | 20:59 | |
| *** bencherian has joined #openstack | 21:01 | |
| tillmo | sorry, I cannot confirm this. I just have deleted the keystone db, set it up again without the endpoint template: same behaviour as with it | 21:02 |
| tillmo | NovaServerError at / Bad Request | 21:03 |
| *** Razique has joined #openstack | 21:04 | |
| vidd | tillmo, its because you have to upload your nova credentials into keystone and actually configure the database properly | 21:04 |
| *** obino has quit IRC | 21:04 | |
| tillmo | aha | 21:04 |
| vidd | *wave Razique | 21:05 |
| vidd | and i havent figured that part out yet | 21:05 |
| tillmo | OK, I will try as well | 21:06 |
| gnu111 | anyone seen this note: "Balance of 58.33 indicates you should push this | 21:06 |
| gnu111 | ring, wait at least 1 hours, and rebalance/repush." | 21:06 |
| vidd | i want to document what I have so far so I can pass it on to annegentle | 21:06 |
| Razique | hi guys | 21:07 |
| Razique | back from a concert :) | 21:08 |
| *** nati2 has quit IRC | 21:08 | |
| Razique | so anyone could confirm the fact that nova + keystone require a new way to launch instances ? | 21:08 |
| *** aliguori has quit IRC | 21:08 | |
| *** tacitone has quit IRC | 21:08 | |
| vidd | i still havent been to bed...been here since 9am EST yesterday | 21:08 |
| *** hub_cap has quit IRC | 21:09 | |
| vidd | thats 32 hours | 21:09 |
| Razique | what time is it in ur contry ? | 21:09 |
| Razique | sick :o | 21:09 |
| *** Horroreyes has joined #openstack | 21:09 | |
| vidd | this databse will WORK before I go to bed | 21:10 |
| vidd | [probably sometime tommorrow] | 21:10 |
| Razique | vidd: ahah still on it ? :D | 21:10 |
| *** dopacki has joined #openstack | 21:10 | |
| Razique | anyway i'll continue tomorrow, bye guys | 21:11 |
| vidd | Im writing it all down while i still remmeber it | 21:11 |
| dopacki | nova/cactus question - trying to bind nova-api to a particular IP address. It seems to want to bind to 0.0.0.0 every time. Any suggestions? | 21:11 |
| Razique | that's why I hate scripted installation :p | 21:11 |
| Razique | when it doesn't work you dunno where too look :D | 21:11 |
| *** dendrobates is now known as dendro-afk | 21:11 | |
| Razique | dopacki: check conf files | 21:12 |
| Razique | glance-api.conf anf api-paste.ini | 21:12 |
| *** bencherian has quit IRC | 21:12 | |
| Razique | both have lines that define on what services must listen on | 21:12 |
| dopacki | Razique - I'm setting my_ip in the conf file, but it seems to ignore that and grab 0.0.0.0 | 21:12 |
| vidd | dopacki, its the controller right? | 21:13 |
| *** cp16net has quit IRC | 21:13 | |
| dopacki | just the api goods | 21:13 |
| dopacki | hmm.. perhaps osapi_listen? that sounds interesting | 21:14 |
| Razique | dopacki: maybe a bug... have u looked on launchpad ? | 21:14 |
| *** markvoelker has quit IRC | 21:14 | |
| vidd | its not a bug...its nova...i HAS to listen to all incoming requests or it cant do its job | 21:15 |
| *** jeh has quit IRC | 21:15 | |
| *** bencherian has joined #openstack | 21:15 | |
| dopacki | lol - "its not a bug...its nova" | 21:16 |
| vidd | you want to stop other traffic coming into it, us iptables...but dont be surprised when stuff strops working correctly | 21:16 |
| dopacki | looks like osapi_listen works, though not documented | 21:16 |
| dopacki | nah, vidd.. I just needed to shim nginx in front of the api | 21:16 |
| dopacki | so binding the api to 127.0.0.1, and terminating connections to 8774 on nginx before sending them to the api | 21:17 |
| *** df1_ has joined #openstack | 21:17 | |
| dopacki | working now. thx | 21:17 |
| vidd | osapi is depreciated | 21:18 |
| vidd | unless i am mistaken | 21:18 |
| *** debo_os has joined #openstack | 21:19 | |
| *** debo_os has quit IRC | 21:20 | |
| Razique | vidd: it nova nova ;) | 21:20 |
| Razique | you can specify the interface to listen on | 21:20 |
| Razique | 1- in conf there are entries for that | 21:20 |
| *** df1 has quit IRC | 21:21 | |
| *** df1_ is now known as df1 | 21:21 | |
| *** marrusl has joined #openstack | 21:21 | |
| Razique | 2- for every Linux I've ever worked on , you specify the interface to listen on | 21:21 |
| Razique | (specify -> have the ability to tweak that part) | 21:21 |
| *** dubsquared has quit IRC | 21:22 | |
| *** tacitone has joined #openstack | 21:22 | |
| *** pothos_ has joined #openstack | 21:24 | |
| *** Razique has left #openstack | 21:26 | |
| *** pothos has quit IRC | 21:26 | |
| *** pothos_ is now known as pothos | 21:26 | |
| *** gnu111 has quit IRC | 21:29 | |
| *** msivanes has quit IRC | 21:30 | |
| *** nati2 has joined #openstack | 21:30 | |
| *** reed has quit IRC | 21:37 | |
| *** aliguori has joined #openstack | 21:38 | |
| *** reed has joined #openstack | 21:38 | |
| uvirtbot | New bug: #879106 in nova "console/api assumes console host is compute host" [Medium,In progress] https://launchpad.net/bugs/879106 | 21:41 |
| *** robix has quit IRC | 21:44 | |
| *** robbiew has quit IRC | 21:44 | |
| *** dopacki has quit IRC | 21:45 | |
| *** mattray has quit IRC | 21:46 | |
| *** rnorwood has quit IRC | 21:47 | |
| *** marrusl has quit IRC | 21:49 | |
| *** mattray has joined #openstack | 21:50 | |
| uvirtbot | New bug: #879111 in horizon "tenant_id in url not enforced/checked" [Critical,Confirmed] https://launchpad.net/bugs/879111 | 21:52 |
| *** bencherian has quit IRC | 21:53 | |
| *** obino has joined #openstack | 21:56 | |
| *** bencherian has joined #openstack | 21:56 | |
| Ryan_Lane | anyone have an idea how I would load this kexec-loader using the glance commands? http://bazaar.launchpad.net/~smoser/+junk/kexec-loader/view/head:/doc/loader-ramdisks.txt | 21:59 |
| tillmo | vidd: I just managed to really login into dashboard! I can now see the images and launch instances. The next error is now "failed so spawn", but that shows up within a working dashboard! | 22:01 |
| uvirtbot | New bug: #862980 in horizon "Make keystone related status messages more usable (dup-of: 875231)" [Undecided,Invalid] https://launchpad.net/bugs/862980 | 22:01 |
| vidd | tillmo, you get your creds uploaded to credentials in keystone? | 22:02 |
| uvirtbot | New bug: #862667 in horizon "install_venv fails behind the proxy " [Low,Confirmed] https://launchpad.net/bugs/862667 | 22:02 |
| tillmo | vidd: no | 22:02 |
| tillmo | possibly that's why I cannot launch the instance | 22:03 |
| vidd | try this: | 22:05 |
| *** catintheroof has quit IRC | 22:06 | |
| vidd | keystone-manage endpointTemplates add RegionOne nova http://192.168.15.200:8774/v1.1/ http://192.168.15.200:8774/v1.1 http://192.168.15.200:8774/v1.1 1 1 | 22:06 |
| *** jdag has joined #openstack | 22:06 | |
| *** Horroreyes has quit IRC | 22:07 | |
| vidd | keystone-manage endpoint add [insert your username in keystone] [whtever number nova shows up as in your database | 22:09 |
| *** mattray has quit IRC | 22:09 | |
| tillmo | in the template, cmd I probably need to replace the ips | 22:09 |
| *** kbringard has quit IRC | 22:10 | |
| tillmo | IndexError: Service nova not found | 22:10 |
| vidd | keystone-manage credentials add nova [your admin key from your novarc] [your secret key from novarc] 1 | 22:11 |
| vidd | tillmo, you have to add it first | 22:11 |
| tillmo | ERROR: 'NoneType' object has no attribute 'id' | 22:13 |
| *** obino has quit IRC | 22:13 | |
| tillmo | I need to add the service nova first. How? | 22:14 |
| *** obino has joined #openstack | 22:14 | |
| vidd | keystone-manage endpointTemplates add RegionOne nova http://192.168.15.200:8774/v1.1/ http://192.168.15.200:8774/v1.1 http://192.168.15.200:8774/v1.1 1 1 | 22:14 |
| vidd | wait | 22:14 |
| vidd | keystone-manage service add nova Nova_API_v1.1 Open-Stack-Nova-v1.1 | 22:14 |
| *** patelna has quit IRC | 22:15 | |
| vidd | are you not glad im setting up all this lovely documentation =] | 22:15 |
| tillmo | yep | 22:15 |
| vidd | [cuzz yule never do it?] | 22:15 |
| tillmo | once I have managed this setup, I will also write something about it... | 22:17 |
| vidd | dont tell ppl im a pain in the ...... oh...like a how-to? | 22:17 |
| tillmo | the endpointTemplate addition works, but the credentials not... | 22:17 |
| vidd | =] | 22:17 |
| tillmo | perhaps even like a how-to... | 22:17 |
| tillmo | I still get "ERROR: 'NoneType' object has no attribute 'id'" | 22:18 |
| vidd | do you have a tenant id linked to your user account? | 22:19 |
| tillmo | in keystone, yes | 22:19 |
| vidd | what is the tenant # | 22:19 |
| tillmo | I just have managed to shut down a VM via dashboard (I had started it via console) | 22:20 |
| tillmo | the tenant # is 1 | 22:20 |
| *** lborda has quit IRC | 22:21 | |
| vidd | try the tenant name instead of the number | 22:22 |
| tillmo | ha, I just have successfully launched a VM! | 22:22 |
| tillmo | via dashboard, I mean | 22:22 |
| vidd | with my assistance? | 22:22 |
| tillmo | yes, of course | 22:22 |
| vidd | i want to know if the instructions i gave you are correct | 22:23 |
| vidd | i dont want to put bogus stuff up for pple | 22:23 |
| vidd | do i look like i work for a company in redmaond wa? | 22:24 |
| tillmo | yes, your instructions have been of great help | 22:24 |
| tillmo | although I have the impression that some things were not needed, but I am not sure | 22:24 |
| vidd | its better to be right than fast | 22:25 |
| *** code_franco has quit IRC | 22:25 | |
| *** pixelbeat has joined #openstack | 22:26 | |
| tillmo | yes. The only thing I am not sure about is the endpoint stuff in keystone | 22:26 |
| Glacee | http://pastebin.com/2X7A1URi Getting a Put 404 when trying to upload file after install.. any idea? | 22:26 |
| Glacee | 1.4.3 and tempauth | 22:26 |
| vidd | tillmo, ill pastebin my doc for you to grab =] | 22:27 |
| tillmo | oh, that is great | 22:27 |
| vidd | tillmo, youhave the ability to save images in dashboard? | 22:28 |
| *** bencherian has quit IRC | 22:28 | |
| vidd | i should say does dashboard have the ability | 22:28 |
| tillmo | no, I cannot. However, first I would like to view the console. That does not work either | 22:29 |
| vidd | need all the endpoits -] | 22:30 |
| vidd | let me get cracking on this doc | 22:30 |
| tillmo | I have all those that you told me :-) | 22:30 |
| *** nati2 has quit IRC | 22:31 | |
| *** dragondm has quit IRC | 22:32 | |
| tillmo | another point is that my launched instance has no public IP, only a private one | 22:32 |
| *** HowardRoark has quit IRC | 22:32 | |
| tillmo | with the console, I used euca-allocate-address and euca-associate-address | 22:33 |
| vidd | you need to add your ec2 creds | 22:34 |
| vidd | keystone-manage credentials add EC2 [your EC2 key from your novarc] [your EC2 secret key from novarc] 1 | 22:35 |
| vidd | keystone-manage credentials add EC2 [your EC2 key from your novarc] [your EC2 secret key from novarc] [tenant name] | 22:35 |
| uvirtbot | New bug: #879136 in glance "keyerror: 'image' when doing nova image-list" [Undecided,New] https://launchpad.net/bugs/879136 | 22:36 |
| vidd | sorry...the 1 at the end is incorrect | 22:36 |
| tillmo | this gives: ERROR: 'NoneType' object has no attribute 'id' | 22:36 |
| vidd | so we need to create an endpoint for ec2 | 22:37 |
| tillmo | in the same way as for nova? | 22:38 |
| vidd | i dont know how to do that [yet] | 22:38 |
| vidd | cant hurt...make sure you use the right port | 22:38 |
| *** Shentonfreude1 has quit IRC | 22:39 | |
| vidd | endpoint template, then endpoint, then credentials | 22:39 |
| vidd | BTW you should be able to do all this from ***not the host*** | 22:39 |
| *** TheOsprey has quit IRC | 22:40 | |
| tillmo | ? | 22:40 |
| vidd | at least the dashboard stuff | 22:40 |
| vidd | my server machines are in the other room....cant even see them from here....i ssh in =] | 22:40 |
| tillmo | yes, I do the same | 22:41 |
| tillmo | I am working from my laptop | 22:41 |
| * `g ssh in vidd | 22:42 | |
| *** miclorb_ has joined #openstack | 22:43 | |
| *** nati2 has joined #openstack | 22:44 | |
| vidd | too bad razique isnt here...he'd love this documentation | 22:45 |
| vidd | getting dashboard and keystone to work is the only thing holding up his documentation | 22:47 |
| *** rsampaio has quit IRC | 22:54 | |
| tillmo | OK, I am too tired now, going to bed | 22:57 |
| tillmo | If you have some doc, please let me know: till (at) communtu.org | 22:58 |
| tillmo | and good luck with your tries! | 23:02 |
| tillmo | bye | 23:02 |
| vidd | bye | 23:02 |
| *** Ryan_Lane has quit IRC | 23:02 | |
| *** tillmo has quit IRC | 23:07 | |
| *** mattray has joined #openstack | 23:07 | |
| *** ziyadb has quit IRC | 23:10 | |
| *** ziyadb has joined #openstack | 23:12 | |
| *** vernhart has quit IRC | 23:13 | |
| *** magg has joined #openstack | 23:18 | |
| magg | im gettin this error trying to run openstack dashboard: http://pastebin.com/YjKTE847 | 23:19 |
| *** clauden__ has quit IRC | 23:20 | |
| magg | helpz | 23:20 |
| *** clauden has joined #openstack | 23:20 | |
| vidd | magg you dont have openstackx installed...just like the error says | 23:21 |
| vidd | you downloaded this from git...right? | 23:21 |
| magg | vidd: i installed the dashboard from the ubuntu 11.10 repos | 23:22 |
| magg | vidd: im pretty sure it downloaded openstackx | 23:22 |
| vidd | ok...that explains alot =] | 23:22 |
| magg | vidd: but maybe i need to add openstackx path or something | 23:22 |
| *** doorlock has joined #openstack | 23:23 | |
| magg | :( | 23:23 |
| vidd | downloading keystone from from repos was a PITA too | 23:23 |
| *** koolhead17 has quit IRC | 23:24 | |
| magg | yeah i installed keystone also | 23:24 |
| magg | i have already configured | 23:24 |
| vidd | magg does it work? | 23:25 |
| magg | yeah | 23:25 |
| magg | i configured it | 23:25 |
| magg | added an admin, tenant and token | 23:25 |
| magg | and did the curl tests | 23:26 |
| vidd | did you add the endpoints? | 23:26 |
| magg | where? | 23:26 |
| vidd | so that when dashboard is set up it has something to do? | 23:26 |
| vidd | in keystone | 23:26 |
| magg | mmmm i guess | 23:26 |
| magg | how do i check | 23:26 |
| magg | heh | 23:26 |
| vidd | what database did you select? | 23:27 |
| magg | slqlite | 23:27 |
| magg | OPENSTACK_KEYSTONE_URL = "http://localhost:5000/v2.0/" | 23:27 |
| magg | i added that to the local_settings.py | 23:28 |
| magg | in the dashboard | 23:28 |
| vidd | did you add a user, tenant etc? | 23:29 |
| magg | yeah | 23:29 |
| magg | and a token | 23:29 |
| vidd | did you add all the endpointTemplates | 23:29 |
| vidd | there was like 5 of them on the examples page | 23:30 |
| magg | maybe not | 23:31 |
| magg | how do i add them | 23:31 |
| vidd | you had to replace %HOST_IP% witht the actual ips, and then bind them to your tenants | 23:31 |
| magg | curl -H "X-Auth-Token:999888777666" http://localhost:5000/v2.0/tenants | 23:31 |
| magg | {"tenants": {"values": [{"enabled": 1, "id": "MyTenant", "description": null}], "links": []} | 23:31 |
| vidd | im still in the middle of updating the documentations ... | 23:32 |
| magg | oh | 23:32 |
| vidd | i need someone to test them | 23:32 |
| magg | should i? | 23:32 |
| *** vernhart has joined #openstack | 23:33 | |
| vidd | do you have nova working ? | 23:33 |
| magg | yeah | 23:33 |
| vidd | yeah...i would like you to test them for me then | 23:33 |
| vidd | im nearly finished | 23:34 |
| magg | ok | 23:34 |
| *** doorlock has quit IRC | 23:34 | |
| vidd | but for your issue.... | 23:34 |
| magg | yeah :( | 23:34 |
| vidd | "sudo apt-get install python-pip | 23:35 |
| *** ziyadb has quit IRC | 23:35 | |
| vidd | sudo pip install [insert missing python requirements here] | 23:35 |
| magg | sudo pip install openstackx? | 23:36 |
| vidd | yes | 23:36 |
| magg | ook | 23:36 |
| *** nerens has quit IRC | 23:36 | |
| *** ziyadb has joined #openstack | 23:36 | |
| *** vernhart has quit IRC | 23:37 | |
| vidd | im sure you'll come across other missing stuff as well | 23:37 |
| magg | could not find any download that satisfy the requirement openstackx | 23:37 |
| vidd | openstack released diablo WAY to close to ubuntu's October release so some things got missed | 23:37 |
| magg | damn | 23:38 |
| vidd | try sudo apt-get install --reinstall openstackx | 23:38 |
| *** osier has joined #openstack | 23:39 | |
| *** misheska has joined #openstack | 23:40 | |
| magg | when i do a python -c "import openstackx" i get a No module named openstackx | 23:41 |
| magg | after a reinstall | 23:41 |
| magg | i think its a path problem | 23:41 |
| vidd | do you see openstackx in your /etc folder? | 23:41 |
| *** vernhart has joined #openstack | 23:41 | |
| *** misheska has quit IRC | 23:41 | |
| vidd | I dunno....i havent gotten to dashboard yet | 23:42 |
| *** vernhart1 has joined #openstack | 23:42 | |
| magg | nope | 23:42 |
| magg | damn | 23:42 |
| magg | :( | 23:42 |
| vidd | try getting dashboard from github | 23:43 |
| vidd | anybody got a link? | 23:43 |
| vidd | !dashboard | 23:43 |
| openstack | vidd: Error: "dashboard" is not a valid command. | 23:43 |
| vidd | do a google serch | 23:44 |
| vidd | they are calling it something else for the time being....i dont remeber offhand | 23:44 |
| magg | ook thanks | 23:45 |
| *** vernhart has quit IRC | 23:46 | |
| vidd | `bug dashboard | 23:46 |
| vidd | !bug dashboard | 23:46 |
| openstack | vidd: Error: "bug" is not a valid command. | 23:46 |
| *** `g has quit IRC | 23:51 | |
| *** vernhart1 has quit IRC | 23:54 | |
| *** nerens has joined #openstack | 23:58 | |
| vidd | magg, https://github.com/4P/horizon is the link for dashboard | 23:59 |
| magg | vidd: thanks downloading it now | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
