*** bonzay has quit IRC | 00:01 | |
*** bonzay has joined #openstack | 00:02 | |
*** dh has quit IRC | 00:02 | |
*** dh_ has joined #openstack | 00:02 | |
*** bcwaldon has quit IRC | 00:12 | |
*** ton_katsu has quit IRC | 00:14 | |
*** clauden has quit IRC | 00:14 | |
*** derrick_ has quit IRC | 00:20 | |
*** heckj has quit IRC | 00:21 | |
*** sophiap has joined #openstack | 00:21 | |
*** mszilagyi has quit IRC | 00:22 | |
*** ton_katsu has joined #openstack | 00:26 | |
*** grapex has quit IRC | 00:38 | |
*** ahs3 has joined #openstack | 00:38 | |
*** Binbin has joined #openstack | 00:43 | |
*** cloudgroups has joined #openstack | 00:43 | |
*** bonzay is now known as zz_bonzay | 00:44 | |
*** msinhore has joined #openstack | 00:44 | |
*** yui has joined #openstack | 00:45 | |
*** RobertLaptop has quit IRC | 00:49 | |
*** msinhore has quit IRC | 00:49 | |
*** shentonfreude has quit IRC | 00:50 | |
*** shentonfreude has joined #openstack | 00:51 | |
*** RobertLaptop has joined #openstack | 00:53 | |
*** cloudgroups has quit IRC | 00:55 | |
*** sophiap has quit IRC | 00:56 | |
*** cloudgroups has joined #openstack | 00:59 | |
*** imsplitbit has quit IRC | 01:05 | |
*** adjohn has joined #openstack | 01:05 | |
*** AimanA has quit IRC | 01:13 | |
*** jdurgin has quit IRC | 01:14 | |
*** miclorb has quit IRC | 01:18 | |
*** jaypipes has quit IRC | 01:21 | |
*** Jamey has quit IRC | 01:22 | |
*** gaveen has joined #openstack | 01:22 | |
*** obino has quit IRC | 01:33 | |
*** joearnold has joined #openstack | 01:44 | |
*** joearnold has quit IRC | 01:45 | |
*** user3425235 has joined #openstack | 01:50 | |
*** shang_ has joined #openstack | 01:54 | |
*** Ryan_Lane has quit IRC | 01:54 | |
*** user3425235 has quit IRC | 01:58 | |
*** maplebed has quit IRC | 01:58 | |
*** nati has joined #openstack | 01:59 | |
*** mancdaz1203 has quit IRC | 02:00 | |
*** mancdaz1203 has joined #openstack | 02:02 | |
*** jkoelker has quit IRC | 02:05 | |
*** sophiap has joined #openstack | 02:13 | |
*** bluetux has joined #openstack | 02:14 | |
*** Binbin has quit IRC | 02:15 | |
*** matiu has quit IRC | 02:18 | |
*** matiu has joined #openstack | 02:19 | |
*** matiu has joined #openstack | 02:19 | |
*** nati has quit IRC | 02:27 | |
*** dragondm has quit IRC | 02:28 | |
*** grapex has joined #openstack | 02:29 | |
*** matiu has quit IRC | 02:33 | |
*** osier has joined #openstack | 02:35 | |
*** sophiap has quit IRC | 02:42 | |
*** carlp has quit IRC | 02:42 | |
*** carlp has joined #openstack | 02:47 | |
*** sophiap has joined #openstack | 02:48 | |
*** sophiap has quit IRC | 02:48 | |
*** wariola has joined #openstack | 02:49 | |
*** sophiap has joined #openstack | 02:49 | |
*** cloudgroups has left #openstack | 02:49 | |
*** Ephur has joined #openstack | 02:50 | |
*** sophiap has quit IRC | 02:56 | |
*** msinhore has joined #openstack | 03:01 | |
*** rchavik has joined #openstack | 03:07 | |
*** rchavik has joined #openstack | 03:07 | |
*** victor_lowther has quit IRC | 03:17 | |
*** rchavik has quit IRC | 03:17 | |
*** libspeed has joined #openstack | 03:27 | |
*** PeteDaGuru has quit IRC | 03:30 | |
*** joearnold has joined #openstack | 03:32 | |
*** msinhore has quit IRC | 03:37 | |
*** msinhore has joined #openstack | 03:40 | |
*** alex_openstack has joined #openstack | 03:50 | |
*** miclorb_ has joined #openstack | 03:50 | |
alex_openstack | hi, i use two pc(as server) for install muilt-node nova | 03:51 |
---|---|---|
alex_openstack | what's docs is the best can i follow? | 03:51 |
annegentle | You can start here http://wiki.openstack.org/InstallInstructions/Nova to choose your instructions. | 03:56 |
*** bcwaldon has joined #openstack | 03:59 | |
*** koolhead17 has joined #openstack | 04:02 | |
*** bcwaldon has quit IRC | 04:12 | |
alex_openstack | thanks annegentle | 04:13 |
*** GeoDud has quit IRC | 04:14 | |
*** naehring has joined #openstack | 04:14 | |
*** kashyap has joined #openstack | 04:27 | |
*** Binbin has joined #openstack | 04:27 | |
*** obino has joined #openstack | 04:27 | |
*** koolhead17 has quit IRC | 04:29 | |
*** stewart has quit IRC | 04:29 | |
*** cbeck has quit IRC | 04:30 | |
*** cbeck has joined #openstack | 04:30 | |
*** taihen has quit IRC | 04:32 | |
*** aaronAUS has joined #openstack | 04:40 | |
aaronAUS | Hi All | 04:42 |
*** victor_lowther has joined #openstack | 04:42 | |
aaronAUS | does anyone have a custom SAIO (Swift all in one) script that they use to quickly deploy a new virtual machine??? | 04:42 |
*** hadrian has quit IRC | 04:45 | |
annegentle | aaronAUS: may I email you one? I know, I know, old school. | 04:46 |
annegentle | it'll come from anne@openstack.org | 04:46 |
aaronAUS | annegentle: that would be fantastic | 04:46 |
*** rchavik has joined #openstack | 04:48 | |
*** victor_lowther has quit IRC | 04:51 | |
*** msinhore1 has joined #openstack | 04:55 | |
*** msinhore has quit IRC | 04:57 | |
*** aaronAUS has quit IRC | 05:04 | |
*** et_ has quit IRC | 05:04 | |
*** MarkusWPHRorg has quit IRC | 05:04 | |
*** nRy has quit IRC | 05:09 | |
*** taihen has joined #openstack | 05:13 | |
*** stewart has joined #openstack | 05:14 | |
*** wariola has quit IRC | 05:15 | |
*** wariola has joined #openstack | 05:17 | |
*** jpgeek1 has joined #openstack | 05:17 | |
*** naehring has quit IRC | 05:19 | |
*** nati has joined #openstack | 05:19 | |
*** naehring has joined #openstack | 05:19 | |
*** stewart has quit IRC | 05:19 | |
*** stewart has joined #openstack | 05:20 | |
*** _william_ has joined #openstack | 05:20 | |
*** naehring1 has joined #openstack | 05:22 | |
*** magglass1 has quit IRC | 05:23 | |
*** magglass1 has joined #openstack | 05:23 | |
*** naehring has quit IRC | 05:24 | |
*** johnpur has quit IRC | 05:25 | |
*** wariola has quit IRC | 05:28 | |
*** koolhead17 has joined #openstack | 05:29 | |
*** koolhead17 has quit IRC | 05:31 | |
*** grapex has quit IRC | 05:32 | |
*** koolhead17 has joined #openstack | 05:33 | |
*** PW__ has joined #openstack | 05:34 | |
*** msinhore1 has quit IRC | 05:36 | |
*** joearnold has quit IRC | 05:38 | |
*** stewart has quit IRC | 05:39 | |
*** stewart has joined #openstack | 05:39 | |
*** guigui1 has joined #openstack | 05:52 | |
*** freeflyi1g is now known as freeflying | 05:57 | |
*** freeflying has quit IRC | 05:57 | |
*** freeflying has joined #openstack | 05:57 | |
*** stewart has quit IRC | 06:00 | |
*** Binbin has quit IRC | 06:00 | |
*** stewart has joined #openstack | 06:00 | |
*** Ephur has left #openstack | 06:01 | |
*** bluetux has quit IRC | 06:07 | |
*** cloudgroups has joined #openstack | 06:11 | |
*** Binbin has joined #openstack | 06:13 | |
koolhead17 | HugoKuo_:hey | 06:13 |
HugoKuo_ | hello | 06:16 |
*** PW__ has quit IRC | 06:20 | |
*** PW__ has joined #openstack | 06:21 | |
*** glenc_ has quit IRC | 06:25 | |
*** glenc has joined #openstack | 06:25 | |
*** ironcamel2 has quit IRC | 06:26 | |
*** ironcamel2 has joined #openstack | 06:26 | |
*** bmullan_ has quit IRC | 06:27 | |
*** _william_ has quit IRC | 06:27 | |
*** _william_ has joined #openstack | 06:27 | |
*** stewart has quit IRC | 06:27 | |
*** cloudgroups has left #openstack | 06:35 | |
HugoKuo_ | koolheadi7 , how's going | 06:44 |
*** PW__ has quit IRC | 06:45 | |
*** koolhead17 has quit IRC | 06:52 | |
*** stewart has joined #openstack | 06:58 | |
*** reidrac has joined #openstack | 07:02 | |
*** gelbuhos has quit IRC | 07:09 | |
*** nati has quit IRC | 07:10 | |
HugoKuo_ | How to add an additional gateway in instance's routing table ? | 07:12 |
HugoKuo_ | during fire up an instance ? | 07:12 |
*** gregm has joined #openstack | 07:14 | |
*** dh_ has quit IRC | 07:19 | |
*** nati has joined #openstack | 07:23 | |
*** CloudChris has joined #openstack | 07:25 | |
*** CloudChris has left #openstack | 07:26 | |
*** nacx has joined #openstack | 07:28 | |
*** erik-s has quit IRC | 07:32 | |
*** _william_ has quit IRC | 07:32 | |
*** erik-s has joined #openstack | 07:33 | |
*** bkkrw has joined #openstack | 07:33 | |
*** cw has quit IRC | 07:46 | |
*** po has joined #openstack | 07:57 | |
*** nati has quit IRC | 08:00 | |
*** gregm has quit IRC | 08:01 | |
*** nati has joined #openstack | 08:02 | |
*** nerens has joined #openstack | 08:07 | |
*** bonzay has joined #openstack | 08:09 | |
*** bonzay has quit IRC | 08:12 | |
*** zz_bonzay is now known as bonzay | 08:13 | |
*** Funnnny has joined #openstack | 08:13 | |
Funnnny | Hello everyone, is someup up there, I have a question | 08:17 |
*** dh has joined #openstack | 08:21 | |
*** Razique has joined #openstack | 08:28 | |
*** uksysadmin has joined #openstack | 08:33 | |
*** cruciform has quit IRC | 08:36 | |
*** nati has quit IRC | 08:36 | |
zykes- | is there any deb build of keystone available ? | 08:37 |
*** Razique has quit IRC | 08:38 | |
*** Razique has joined #openstack | 08:41 | |
*** cw has joined #openstack | 08:51 | |
*** miclorb_ has quit IRC | 08:57 | |
*** nsokolov has joined #openstack | 09:02 | |
*** allsystemsarego has joined #openstack | 09:03 | |
*** allsystemsarego has joined #openstack | 09:03 | |
*** dobber has joined #openstack | 09:12 | |
*** daysmen has joined #openstack | 09:14 | |
*** jedi4ever has joined #openstack | 09:15 | |
*** nsokolov has quit IRC | 09:15 | |
*** nsokolov has joined #openstack | 09:15 | |
*** nsokolov is now known as chemikadze | 09:15 | |
*** ton_katsu has quit IRC | 09:16 | |
*** kazuhiko has joined #openstack | 09:20 | |
*** irahgel has joined #openstack | 09:23 | |
*** bonzay has quit IRC | 09:27 | |
*** bonzay has joined #openstack | 09:28 | |
*** alex_openstack has quit IRC | 09:30 | |
Linex | Is openstack an application or is it a framework where others do an implementation of it ? | 09:32 |
*** Binbin has quit IRC | 09:39 | |
chemikadze | Is replay for yesterday webinar available? | 09:44 |
*** jpgeek1 has quit IRC | 09:45 | |
*** magglass1 has quit IRC | 09:52 | |
*** shrii has joined #openstack | 09:54 | |
shrii | hey All | 09:55 |
shrii | Hello | 09:55 |
shrii | Hello All, | 09:55 |
shrii | How are you All.! | 09:56 |
shrii | I am setting up stackops on our new server. | 09:56 |
*** shang has quit IRC | 09:56 | |
shrii | i am getting some issue with openstack-dashboard, | 09:56 |
shrii | not connecting to database. | 09:57 |
*** shang has joined #openstack | 09:57 | |
Razique | Hi | 09:57 |
Razique | I just filled a new bug | 09:57 |
shrii | are any on facing this type of issue. | 09:57 |
shrii | Bug * | 09:58 |
shrii | what that ? | 09:58 |
*** ton_katsu has joined #openstack | 09:58 | |
*** ton_katsu has joined #openstack | 10:02 | |
*** magglass1 has joined #openstack | 10:05 | |
*** wilmoore has joined #openstack | 10:07 | |
Funnnny | is openstack support xenserver | 10:08 |
Funnnny | and if it's not, is there any plan to support ? | 10:08 |
uvirtbot | New bug: #800591 in nova "Unable to attache more than one volume to a compute-node" [Undecided,New] https://launchpad.net/bugs/800591 | 10:11 |
uvirtbot | New bug: #800593 in nova "Flash websocket emulation doesn't work with nova-vncproxy" [Undecided,New] https://launchpad.net/bugs/800593 | 10:11 |
ccooke | Funnnny: openstack does support xenserver as a hypervisor | 10:15 |
Funnnny | ccooke: thanks, is there any document on it, I can't find any | 10:16 |
ccooke | Funnnny: http://wiki.openstack.org/XenServerDevelopment | 10:17 |
*** kashyap has quit IRC | 10:17 | |
*** kashyap has joined #openstack | 10:17 | |
Funnnny | ccooke: is document there seems to be wrong, XenServer use yum not apt-get | 10:18 |
Funnnny | I'm confusing | 10:18 |
bonzay | Funnnny: thats a "documentation error" | 10:18 |
ccooke | Funnnny: it wants you to do that in a VM on the hypervisor | 10:18 |
bonzay | you need to do these commands _inside_ the compute-node in a ubuntu.-vm | 10:19 |
*** adjohn has quit IRC | 10:19 | |
ccooke | Basically, in order for OpenStack to inject VM images into the system it needs access to the hypervisor internals. There's currently no clean way to do that on the hypervisor dom0, so it's donein domU | 10:19 |
bonzay | ccooke: but this is very confusing for a beginner, I had the same problems understanding as Funnnny | 10:20 |
ccooke | (I have a half-finished package that does it for dom0, but that was for my employer and I'm on a different project for the moment) | 10:20 |
Funnnny | so we must create a domU and follow the prepare nova part | 10:20 |
Funnnny | and then do the rest with XenServer | 10:21 |
ccooke | bonzay: Myself also. At least the document now *says* you need to do it inside a VM. When I started working with this, it didn't say that much :-) | 10:21 |
bonzay | ccooke: at least :) | 10:22 |
* ccooke is slightly frustrated; I have an alternate method for getting this to work which is much nicer, but I don't actually have time to do any work on it right now | 10:22 | |
Funnnny | I'm taking a research to choose OpenNebula or OpenStack | 10:23 |
ccooke | (note that I am not affiliated with openstack myself; I'm just someone using it) | 10:23 |
bonzay | ccooke: will you share your method? | 10:23 |
Funnnny | but seems that those two both are more nicely fit to a "workload", rather than a virtual infrastructure | 10:23 |
ccooke | bonzay: absolutely | 10:23 |
ccooke | bonzay: it's really a managed, packaged chroot image. | 10:24 |
ccooke | it's nice and clean and comes as a single file | 10:24 |
ccooke | there are, however, a couple of issues I need to deal with to make it work cleanly | 10:24 |
bonzay | ccooke: is it a .deb package you need to install in domU? | 10:25 |
ccooke | an rpm for the xenserver dom0, actually | 10:26 |
ccooke | much cleaner | 10:26 |
bonzay | i agree | 10:26 |
bonzay | so you dont need the domU then? | 10:26 |
ccooke | You currently do | 10:26 |
*** Fusion[SSP] has joined #openstack | 10:26 | |
ccooke | because the dom0 option requires me to document and fix a couple of bugs in openstack | 10:27 |
shrii | [15:26] <shrii> i am getting some issue with openstack-dashboard, [15:27] <shrii> not connecting to database. | 10:27 |
bonzay | but with your script you wont need it, right? | 10:27 |
bonzay | s/script/package | 10:27 |
ccooke | Aye. I expect I'll get time to work on that in a few weeks. | 10:27 |
ccooke | However, that is just my option - it might turn out a bad one for openstack as a whole | 10:27 |
ccooke | But hey, it will make me *much* happier with the xenserver support | 10:28 |
ccooke | a domU is... messy | 10:28 |
bonzay | yep, totally agreed | 10:28 |
ccooke | And involves violation of levels of access/control | 10:29 |
bonzay | I am very interested in a clean version of xen-server support | 10:29 |
bonzay | so keep us informed :) | 10:29 |
ccooke | Will do. | 10:30 |
bonzay | thx | 10:30 |
ccooke | I *have* had permission to share the implementation once it's actually usable, which is nice (since I'm doing it on work time) | 10:30 |
Funnnny | ccooke: did you make xenserver usable with openstack dashboard | 10:31 |
*** jtimberman has quit IRC | 10:32 | |
ccooke | Funnnny: Yes... with the proviso that creating VMs doesn't actually *work* yet | 10:32 |
ccooke | (All the issues are with the compute node, though: it was working absolutely fine with the dashboard) | 10:32 |
Funnnny | ccooke: so it's the problem with compute node right now ? can we solve it yet ? | 10:33 |
ccooke | Funnnny: the problem is specific to my implementation | 10:34 |
ccooke | Xenserver is fine with openstack if you follow the (not very clear, admittedly) instructions in the XenServerDevelopment page | 10:34 |
Funnnny | ccooke: Thanks for your help | 10:35 |
ccooke | No problem. I wish it had been in a few weeks time, when I could offer something that's easier ;-) | 10:35 |
Funnnny | another things, can I use XenCenter to manage the VM | 10:38 |
Funnnny | and then migrate to openstack when I setup and get the feature run nicely | 10:39 |
Funnnny | ccooke: or I have to convert and register the VM to openstack | 10:39 |
*** katkee has joined #openstack | 10:40 | |
ccooke | It would technically be possible to migrate the VMs to openstack, but I'm unaware of any way to do so at the moment. It could certainly be scripted, although there are fairly complex details to work out | 10:41 |
Funnnny | ccooke: btw openstack is more like a workload cloud, than a virtual infrastructure management right ? | 10:43 |
Funnnny | I'm not sure about the lifecycle of the VM in openstack | 10:43 |
*** bonzay_ has joined #openstack | 10:46 | |
ccooke | I'm not sure what you mean | 10:47 |
Funnnny | ccooke: I'm not so sure, as with my experience, I'm just testing both opennebula and openstack, when a VM restart or shutdown, it return to terminated state and cannot be start again | 10:49 |
*** bonzay_ has quit IRC | 10:49 | |
*** ziyadb has joined #openstack | 10:56 | |
*** adjohn has joined #openstack | 10:57 | |
*** lionel has quit IRC | 10:58 | |
*** lionel has joined #openstack | 10:59 | |
*** adjohn has quit IRC | 10:59 | |
*** markvoelker has quit IRC | 11:02 | |
*** markvoelker has joined #openstack | 11:05 | |
*** katkee has quit IRC | 11:20 | |
Razique | shrii: https://bugs.launchpad.net/nova/+bug/800591 | 11:20 |
shrii | ok Razique | 11:21 |
shrii | thanks | 11:21 |
*** Razique has quit IRC | 11:21 | |
*** Razique has joined #openstack | 11:21 | |
*** adjohn has joined #openstack | 11:23 | |
*** jedi4ever has quit IRC | 11:24 | |
*** mgoldmann has joined #openstack | 11:28 | |
*** ctennis has quit IRC | 11:30 | |
*** ctennis has joined #openstack | 11:43 | |
*** naehring1 has quit IRC | 11:44 | |
*** naehring has joined #openstack | 11:45 | |
*** gaveen has quit IRC | 11:46 | |
*** magglass1 has quit IRC | 11:53 | |
*** magglass1 has joined #openstack | 12:00 | |
*** zul has quit IRC | 12:01 | |
*** guigui1 has quit IRC | 12:01 | |
*** Fusion[SSP] has left #openstack | 12:01 | |
*** guigui3 has joined #openstack | 12:02 | |
*** aliguori has joined #openstack | 12:07 | |
*** jedi4ever has joined #openstack | 12:08 | |
*** victor_lowther has joined #openstack | 12:09 | |
*** onlany has joined #openstack | 12:12 | |
onlany | hi | 12:12 |
onlany | i have a question about kvm | 12:12 |
*** aaronAUS has joined #openstack | 12:12 | |
onlany | my cpu supports hardware virt | 12:13 |
onlany | and it is enabled in bios | 12:13 |
onlany | i have set the libvit type flag in conf | 12:13 |
onlany | restarted all the services | 12:13 |
onlany | but in mysql in the compute node table still qemu is the hypervisor type | 12:13 |
onlany | could u help me why? | 12:14 |
*** katkee has joined #openstack | 12:15 | |
*** dprince has joined #openstack | 12:15 | |
*** ryker has joined #openstack | 12:16 | |
Razique | onlany: it doesn't matter, qemu means qemu/kvm | 12:18 |
Razique | I've the same here, that does not prevent instances to work under kvm :) | 12:19 |
Razique | (Kvm is a fork of Qemu btw :) ) | 12:19 |
*** westmaas_ is now known as westmaas | 12:25 | |
*** dprince has quit IRC | 12:25 | |
*** victor_lowther has quit IRC | 12:28 | |
*** ewindisch has joined #openstack | 12:28 | |
onlany | thanks | 12:28 |
*** dprince has joined #openstack | 12:29 | |
xxtjaxx | How do you create that nova.zip again? | 12:33 |
*** PeteDaGuru has joined #openstack | 12:35 | |
Razique | xxtjaxx: nova-manage project zipfile $project $project_user /zip/destination | 12:37 |
xxtjaxx | Razique: Thx found it :) http://paste.debian.net/120635/ << I can haz explanation? :] | 12:39 |
doude | Hi all, someone use the inject file with OSAPI with XenSever driver ? | 12:39 |
Razique | it means you don't have the right to connect to nova-api server | 12:40 |
Razique | weird :/ | 12:40 |
Razique | do other nova-manage commands work ? like nova-manage flavor list ? | 12:40 |
doude | I get try this command 'nova boot --image 65 --flavor 1 --file /tmp=tmp.sh "Test VHD image"' but I get this error: http://paste.openstack.org/show/1734/ | 12:41 |
*** zul has joined #openstack | 12:42 | |
*** nati has joined #openstack | 12:43 | |
xxtjaxx | Razique: http://paste.debian.net/120636/ Looks kinda working | 12:43 |
Razique | xxtjaxx: ok, try,mmm a euca-add-group -d test test | 12:44 |
xxtjaxx | Razique: instead of doing all the mojo with manifest.xml files and things like that can I instead use those flavours and just tell it to start an instance? | 12:44 |
xxtjaxx | Razique: same result | 12:45 |
Razique | xxtjaxx: of course, you can run a euca-run-instance -k da_key -t $flavor $image-id | 12:45 |
Razique | xxtjaxx: which one, the auth error ? | 12:45 |
*** dh has quit IRC | 12:45 | |
xxtjaxx | The auth error is the same yes | 12:45 |
*** kaz__ has joined #openstack | 12:45 | |
*** dh has joined #openstack | 12:45 | |
Razique | ok | 12:45 |
Razique | mmm | 12:45 |
xxtjaxx | Razique: gee why have I been doing/trying all this mojo then | 12:45 |
* Razique thinking | 12:45 | |
xxtjaxx | da_key is my rsa key? | 12:46 |
xxtjaxx | there maybe an issue with the key. I didnt have any ssh keys inserted into it yet :> | 12:46 |
*** kaz_ has quit IRC | 12:47 | |
xxtjaxx | Do I need both private and pubkey on it? | 12:47 |
*** ziyadb_ has joined #openstack | 12:48 | |
*** tudamp has joined #openstack | 12:49 | |
*** nphase has quit IRC | 12:50 | |
*** FallenPegasus has joined #openstack | 12:50 | |
*** FallenPegasus is now known as MarkAtwood | 12:50 | |
*** nphase has joined #openstack | 12:50 | |
*** ziyadb has quit IRC | 12:50 | |
Razique | xxtjaxx: it uses the private key that u'll use | 12:51 |
Razique | xxtjaxx: in the process order, it's weird that you need the key at that point | 12:51 |
Razique | siince the project is new | 12:51 |
Razique | thus, you don't have the keys to it yet | 12:52 |
aaronAUS | can someone please review the following error message I get when i trry to start swift | 12:52 |
aaronAUS | http://paste.openstack.org/show/1735/ | 12:52 |
*** Razique_ has joined #openstack | 12:55 | |
*** imsplitbit has joined #openstack | 12:55 | |
*** msivanes has joined #openstack | 12:55 | |
*** ameade has joined #openstack | 12:56 | |
xxtjaxx | Razique: how should I add this key in order to get to it? | 12:58 |
*** Razique has quit IRC | 12:58 | |
*** Razique_ is now known as Razique | 12:58 | |
xxtjaxx | Razique: everything fine? | 12:59 |
Razique | xxtjaxx: try to create another project quickly and see if you have the same error | 13:00 |
zykes- | is there any meassures for antispam @ forums ? | 13:01 |
*** tudamp has quit IRC | 13:03 | |
xxtjaxx | Razique: done. + downloaded zipfile and sourced novarrc | 13:05 |
Razique | ok cool | 13:05 |
Razique | it means it'a a project issue then | 13:05 |
Razique | but, at my level of knowledge, could not tell why :/ sorry mate | 13:05 |
Razique | maybe you could see into the database, the project definition (into the "project" table) | 13:06 |
*** rchavik has quit IRC | 13:06 | |
xxtjaxx | hmm db sync didn't happen yet btw | 13:06 |
Razique | xxtjaxx: try to run it manually | 13:09 |
Razique | nova-manage db sync | 13:09 |
notmyname | aaronAUS: I'd guess that your proxy server config has "swauth" in the pipeline | 13:10 |
*** Shentonfreude1 has joined #openstack | 13:10 | |
*** shentonfreude has quit IRC | 13:10 | |
aaronAUS | notmyname: thats what I was trying to achieve.... damn this is impossible... I tried following a different article | 13:10 |
aaronAUS | http://www.slideshare.net/joshuamckenty/openstack-swift-msst-2011-tutorial-day | 13:11 |
aaronAUS | it uses swauth (and SSL) | 13:11 |
*** jedi4ever has quit IRC | 13:11 | |
aaronAUS | so I was hoping it would work with cyberduck | 13:11 |
notmyname | aaronAUS: you can use swauth, but swauth was just taken out of swift proper (it's its own project now). so that tutorial is slightly out of date. swauth can be found at http://github.com/gholt/swauth | 13:12 |
*** jedi4ever has joined #openstack | 13:12 | |
notmyname | there are also instructions there for setting it up | 13:12 |
aaronAUS | hmmm time to start over again | 13:13 |
xxtjaxx | Razique: db sync worked | 13:13 |
*** rchavik has joined #openstack | 13:14 | |
notmyname | aaronAUS: install swauth from that project and you should be good to go. no need to start over | 13:14 |
aaronAUS | I have stuffed around with too much stuff... its ok... just revert back to blank ubuntu with snapshot | 13:15 |
*** jedi4ever has quit IRC | 13:16 | |
*** jedi4ever has joined #openstack | 13:16 | |
*** po has quit IRC | 13:22 | |
*** grapex has joined #openstack | 13:23 | |
*** grapex has left #openstack | 13:23 | |
xxtjaxx | Razique: okay whats next.... | 13:24 |
Razique | xxtjaxx: mmm so you told me it worked with your newly created project, but not with your current project right ? | 13:24 |
Razique | if yours ain't critical, you should scrub it and delete it | 13:24 |
Razique | :/ | 13:24 |
xxtjaxx | Razique: where do I take the Image-ID from? I'd have to create images for that first right? | 13:24 |
Razique | xxtjaxx: yup, what kind of images u looking for ? | 13:25 |
Razique | a good start : http://uec-images.ubuntu.com/ | 13:25 |
xxtjaxx | Razique: some smallish web servery things for a start. I'm just evalling it | 13:25 |
Razique | from it, you download the tar.gz and use the uec-publish-tarball command to upload it to your store (local or glance one) | 13:26 |
xxtjaxx | Razique: shame is we are a debian shop mostli | 13:26 |
xxtjaxx | eh mostly | 13:26 |
xxtjaxx | I am a bit picky about this... :) | 13:27 |
*** ziyadb_ has quit IRC | 13:27 | |
*** ziyadb has joined #openstack | 13:27 | |
*** ziyadb has joined #openstack | 13:27 | |
*** _william_ has joined #openstack | 13:28 | |
Razique | xxtjaxx: http://www.eucalyptussoftware.com/downloads/eucalyptus-images/list.php | 13:28 |
Razique | I downladed a debian lenny image which works into niva | 13:29 |
Razique | nova | 13:29 |
Razique | I you manage to create a debian squeeze AMI,'ill thanks you a ton ; i tried for a week, but with no success | 13:29 |
Razique | xxtjaxx: the original link : http://open.eucalyptus.com/wiki/EucalyptusUserImageCreatorGuide_v1.6 | 13:31 |
*** katkee has quit IRC | 13:31 | |
*** throughn1thing has quit IRC | 13:31 | |
xxtjaxx | Razique: fascinating how tiny they are | 13:32 |
Razique | hehe, yup :) | 13:32 |
xxtjaxx | Razique: whats key concern about the Squeeze ones? What didn't work? | 13:32 |
*** Cromulent has joined #openstack | 13:32 | |
Razique | xxtjaxx: well after an intensive work on it, I concluded that it was an issue with Grub 2 | 13:33 |
*** guigui3 has quit IRC | 13:33 | |
xxtjaxx | Razique: specific please? What did grub say? | 13:33 |
Razique | well, there were several cases 1 - the instance refused to work (some key injection issue) | 13:34 |
xxtjaxx | Razique: we had a few'o'those lately in the lug so I'm kinda prepped for it ;) | 13:34 |
Razique | 2- the boot worked, but the the ramdisk failed to work, it complained about non existent directories : leme check my logs, so you'll have the boot sequence | 13:34 |
*** throughnothing has joined #openstack | 13:35 | |
*** rchavik has quit IRC | 13:35 | |
*** throughnothing has quit IRC | 13:36 | |
*** _william_ has left #openstack | 13:36 | |
xxtjaxx | Razique: tried fixing it by chrooting and update initrd? | 13:36 |
*** throughnothing has joined #openstack | 13:37 | |
*** rchavik has joined #openstack | 13:37 | |
Razique | ah get it! http://paste.openstack.org/show/1641/ | 13:38 |
Razique | xxtjaxx: well it pissed me off, and I lost so many hours on it, so i've suspended this task atm ^^ | 13:38 |
*** naehring has quit IRC | 13:39 | |
*** throughnothing has quit IRC | 13:39 | |
xxtjaxx | Razique: I feel the same here with nova for some reason | 13:39 |
*** naehring has joined #openstack | 13:40 | |
btorch | aaronAUS: here | 13:40 |
*** mdomsch has joined #openstack | 13:41 | |
*** throughnothing has joined #openstack | 13:41 | |
xxtjaxx | Razique: 1. g \/dev on \/root\/dev fa << WTF is this escaping? and \/root\/dev ? Thats a bit off isnit? | 13:41 |
*** throughnothing has quit IRC | 13:41 | |
Razique | xxtjaxx: it's libvirt which escapes, | 13:42 |
Razique | it 's nothing, /root/dev -> initramfs booting sequence | 13:42 |
xxtjaxx | odd... haven't seen this | 13:42 |
Razique | I opened the script, temporarly, initramfs creates /root/proc, /root/dev and /root/sys | 13:42 |
xxtjaxx | Razique: no init command? tf? | 13:43 |
Razique | then deletes them after the driver loading & all cumbersomes ramfs processes ^ | 13:43 |
Razique | tf ? | 13:43 |
xxtjaxx | wtf - w | 13:43 |
*** aliguori has quit IRC | 13:43 | |
*** bcwaldon has joined #openstack | 13:44 | |
*** med_out is now known as medberry | 13:44 | |
*** guigui has joined #openstack | 13:45 | |
*** jedi4ever has quit IRC | 13:45 | |
*** aliguori has joined #openstack | 13:45 | |
*** jedi4ever has joined #openstack | 13:45 | |
xxtjaxx | Razique: tried fixing the Script? | 13:46 |
Razique | xxtjaxx: the ramfs script is good, it's grub2 which messes up | 13:46 |
Razique | but can't tell precisely how | 13:46 |
Razique | trust me it burned my brain out for a whole week ^^ | 13:47 |
xxtjaxx | hm | 13:48 |
Razique | i'm quite sure that if you remove grub2 from your squeeze image and use either grub1 or lilo, it'll work | 13:49 |
*** osier has quit IRC | 13:49 | |
*** throughnothing has joined #openstack | 13:50 | |
*** throughnothing has quit IRC | 13:51 | |
*** dmshelton has joined #openstack | 13:53 | |
*** vernhart has joined #openstack | 13:53 | |
*** uksysadmin has quit IRC | 13:54 | |
*** jkoelker has joined #openstack | 13:55 | |
*** msinhore has joined #openstack | 13:55 | |
*** tahoe_ has joined #openstack | 13:56 | |
*** tahoe has quit IRC | 13:56 | |
*** vernhart has quit IRC | 13:58 | |
*** katkee has joined #openstack | 13:58 | |
*** exlt_ has joined #openstack | 13:58 | |
*** amccabe has joined #openstack | 13:58 | |
*** throughnothing has joined #openstack | 13:59 | |
*** throughnothing has quit IRC | 13:59 | |
*** exlt has quit IRC | 14:00 | |
*** throughnothing has joined #openstack | 14:01 | |
*** throughnothing has quit IRC | 14:01 | |
*** throughnothing has joined #openstack | 14:01 | |
*** throughnothing has quit IRC | 14:01 | |
*** nati has quit IRC | 14:02 | |
*** nacx has quit IRC | 14:04 | |
*** throughnothing has joined #openstack | 14:04 | |
*** throughnothing has quit IRC | 14:04 | |
*** throughnothing has joined #openstack | 14:05 | |
*** msinhore has quit IRC | 14:06 | |
*** msinhore1 has joined #openstack | 14:06 | |
*** dkim1987 has joined #openstack | 14:07 | |
xxtjaxx | Razique: have yoz had a look at virt-install/debootstrap? | 14:12 |
Razique | xxtjaxx: yup I did, but same symptoms | 14:12 |
xxtjaxx | feels odd | 14:13 |
xxtjaxx | Razique: You are employed or Free lancing this? | 14:16 |
Razique | xxtjaxx: we created a company which offers cloud services | 14:16 |
*** aliguori has quit IRC | 14:17 | |
Razique | I'm actually the all technical-aspect manager | 14:17 |
Razique | and tech | 14:17 |
Razique | what about you ? | 14:17 |
*** ton_katsu has quit IRC | 14:18 | |
*** nati has joined #openstack | 14:19 | |
NashTrash | Good morning Openstack'ers! | 14:20 |
*** shrii has quit IRC | 14:21 | |
*** zobel has joined #openstack | 14:21 | |
NashTrash | Question for the group on nova-network and iptables | 14:22 |
Razique | NashTrash: hey from France where it's 4;30 pm ^^ | 14:22 |
NashTrash | Razique: Always with the attitude :) | 14:22 |
NashTrash | Potato — pomme de terre | 14:23 |
NashTrash | same thing | 14:23 |
NashTrash | Ok, my question... | 14:23 |
Razique | NashTrash: ahah, everytime I'll have my frenchie word coming from you then :p | 14:23 |
NashTrash | I have an instance with a public IP address. If I try to ping that address from another instance within the cloud I get a time out. If I ping the private IP it works just fine. | 14:24 |
NashTrash | Clearly when I use the public IP address it routes through the nova-network machine. | 14:25 |
*** markvoelker has quit IRC | 14:25 | |
NashTrash | I looked at the IPtables setup on nova-network and it seemed like it should work, but my iptables skillz are poor | 14:25 |
NashTrash | Should one instance be able to connect to another instance using a public IP address? | 14:26 |
*** dgags has joined #openstack | 14:31 | |
xxtjaxx | Razique: I'm working for an open source support company and official debian partner called credativ. We employ a good amount of the german DebianDevelopers. | 14:32 |
Razique | xxtjaxx: ah cool | 14:33 |
Razique | xxtjaxx: how old are you btw ? ^^ | 14:34 |
Razique | NashTrash: is the public ip routable ? | 14:34 |
*** jmckenty_ has joined #openstack | 14:34 | |
*** joearnold has joined #openstack | 14:36 | |
xxtjaxx | Razique: I'm 19 . Btw whats the underlying device you tried to boot this debian image from? | 14:37 |
*** coggy has joined #openstack | 14:37 | |
NashTrash | Razique: How do you mean? The public IP was done using euca-allocate-address and euca-associate-address | 14:37 |
NashTrash | Razique: P.S. I am using FlatDHCP networking, if that matters | 14:38 |
*** joearnold has quit IRC | 14:39 | |
coggy | hi, will the current dashboard code from git work with the nova release from http://ppa.launchpad.net/nova-core/release/ubuntu? | 14:39 |
*** deshantm has quit IRC | 14:40 | |
Razique | coggy: the nova-dashboard works fine here | 14:41 |
*** devicenull has joined #openstack | 14:42 | |
*** joearnold has joined #openstack | 14:42 | |
Razique | NashTrash: I mean, is the class of the "public" ip a public or a private one ? | 14:42 |
Razique | here my public ones are in fact non-routable ones, I've a frontend-firewall | 14:42 |
*** rchavik has quit IRC | 14:43 | |
coggy | Razique: alright. I'm asking because the template for local_settings.py seems to be out of sync with the documentation in the wiki. Thanks, I'll experiment... | 14:43 |
*** aliguori has joined #openstack | 14:43 | |
*** deshantm has joined #openstack | 14:45 | |
Razique | coggy: I followed that guide : http://wiki.openstack.org/OpenStackDashboard and it worked perfectly | 14:45 |
*** rchavik has joined #openstack | 14:45 | |
*** jmckenty_ has quit IRC | 14:46 | |
coggy | Razique: that's what I'm referring to. None of the NOVA_* settings mentioned in the wiki are found in the local_settings.py template. I'll just add them and see what happens | 14:46 |
Razique | coggy: what settings ? | 14:49 |
Razique | i'll look at my files | 14:49 |
*** joearnold has quit IRC | 14:50 | |
coggy | Razique: NOVA_DEFAULT_ENDPOINT, for instance | 14:51 |
*** imsplitbit has quit IRC | 14:52 | |
*** vernhart has joined #openstack | 14:52 | |
Razique | coggy: in root/src/openstack-dashboard/trunk/openstack-dashboard/local/local_settings.py | 14:53 |
Razique | I've | 14:53 |
Razique | NOVA_DEFAULT_ENDPOINT | 14:53 |
Razique | and the reste | 14:53 |
Razique | rest | 14:53 |
coggy | Razique: is this current code from git? | 14:53 |
xxtjaxx | Razique: Looking at the paste again it may be that the boot failed because it was trying to use the device(vda) and not a partition on the device like vda1. This explains the "unknown partitiontable" failure reported in this log. | 14:54 |
zobel | xxtjaxx, Razique: what are you talking about? url to pastbin? | 14:55 |
Razique | zobel: na, I' explained him I was unable to create a squeeze ami | 14:56 |
xxtjaxx | zobel: Razique tried to boot a Debian image with grub 2. This log http://paste.openstack.org/show/1641/ entails his misfortune at this endeavour. | 14:56 |
Razique | I had an error during the instanciation process | 14:56 |
zobel | xxtjaxx: unknown partitiontable may mean that there is no partition table at all. | 14:58 |
zobel | on the other hand... | 14:58 |
zobel | Razique: it reads like the mount-all.sh tries to mount a device that is non-existing | 14:58 |
zobel | line 238 reads as there is an error while mounting. | 15:00 |
zobel | are you sure your "root=" in the grub-config is okay? | 15:00 |
*** alandman has joined #openstack | 15:01 | |
Razique | zobel: yup I used kvm to create a stand-alone image | 15:03 |
Razique | when I reboot it, it reboots fine under kvm | 15:04 |
Razique | but when I put it into my nova-repo, I've the error ^^ | 15:04 |
Razique | coggy: I installed it few weeks ago | 15:05 |
Razique | please paste your local_settings.py file | 15:05 |
zobel | Razique: may it be that vda devices are mapped to sda in nova? | 15:06 |
Razique | zobel: already tried :) | 15:06 |
Razique | by modifying the /etc/fstab file | 15:06 |
zobel | Razique: no, in your grub | 15:06 |
zobel | its grub that passes root= to the kernel/initrd | 15:07 |
Razique | zobel: ah, no I did not tried that | 15:07 |
coggy | Razique: it's this version: https://github.com/4P/openstack-dashboard/blob/master/openstack-dashboard/local/local_settings.py.example | 15:07 |
zobel | Razique: if that works, you should be able to start your vm, but will need to adjust /etc/fstab afterwards, and then rebuild your initrd | 15:08 |
coggy | Razique: never mind. I added the missing configuration manually and it seems to work just fine. Thank you | 15:09 |
Razique | coggy: ok nice, you should look at the file's history to see if these settings were removed | 15:10 |
Razique | it's weird though Oo | 15:10 |
*** guigui has quit IRC | 15:11 | |
aaronAUS | so I have tried to install SWAUTH, but still it won't start | 15:12 |
aaronAUS | http://paste.openstack.org/show/1736/ | 15:12 |
aaronAUS | does it matter where I downloaded swauth to???? I downloaded the tgz to root. and unpacked it there. and ran "python setup.py install" from there | 15:13 |
aaronAUS | root = /root | 15:13 |
coggy | Razique: it was changed 5 days ago, apparently the authentication method is being changed to keystone | 15:13 |
*** coggy has quit IRC | 15:13 | |
Razique | colinnich: https://github.com/4P/openstack-dashboard/commit/945cf9ff7c2d2aee27695828656389e7116143c4#openstack-dashboard/local/local_settings.py.example | 15:14 |
Razique | ah | 15:14 |
*** bwidmer has joined #openstack | 15:14 | |
Razique | here we are, exactly, so the wiki has not been updated meanwhile ; maybe you could report that ? :) | 15:14 |
btorch | aaronAUS: sorry here now | 15:15 |
gholt | aaronAUS: "Entry point 'swauth' not found in egg 'swift'" that means you need to change the egg:swift#swauth to egg:swauth#swauth | 15:15 |
*** Cromulent has quit IRC | 15:15 | |
btorch | gholt: aaronAUS I thought I added that to that modified paste http://paste.openstack.org/show/1728/ | 15:16 |
*** naehring has quit IRC | 15:19 | |
*** heckj has joined #openstack | 15:20 | |
*** nati has quit IRC | 15:24 | |
*** ziyadb_ has joined #openstack | 15:24 | |
bpaluch | Does the current code in trunk support the keystone code in the latest dashboard then? | 15:24 |
bpaluch | its funny I was trying to set up dashboard again this morning, and then I looked at what coggy posted in here having the same issue | 15:25 |
*** ziyadb has quit IRC | 15:25 | |
*** ziyadb has joined #openstack | 15:26 | |
*** ziyadb has joined #openstack | 15:26 | |
heckj | bpaluch: nova yes, glance not yet, swift damn near but not quite | 15:28 |
*** ziyadb_ has quit IRC | 15:28 | |
bpaluch | do I need to add anything into nova.conf for support then? | 15:28 |
*** Jamey has joined #openstack | 15:29 | |
heckj | bpaluch: there's a nova.sh install that makes putting together the nova set quite easy: See the auto.sh that cloudbuilders has at https://github.com/cloudbuilders/deploy.sh/blob/master/auto.sh | 15:30 |
heckj | It uses an extended nova.sh install script in that same repo | 15:31 |
bpaluch | got it thanks | 15:31 |
*** primeministerp has quit IRC | 15:31 | |
aaronAUS | btorch: http://paste.openstack.org/show/1737/ | 15:32 |
aaronAUS | btorch: error trying to add users | 15:32 |
*** primeministerp has joined #openstack | 15:32 | |
aaronAUS | I successfully did swauth-prep command | 15:32 |
*** jedi4ever has quit IRC | 15:32 | |
aaronAUS | and my proxy-config reflects those changes you made in past 1728 | 15:33 |
btorch | aaronAUS: hmm is everything running ? what does syslog show ? | 15:34 |
CatKiller | Hi There, does anyone know what I should do if I wanted to extend a table within the Nova DB with a new field? | 15:34 |
heckj | CatKiller it's all set up with SQLAlchemy - is your question about the process to contribute, or how to technically do it? | 15:35 |
CatKiller | Well, it's both actually | 15:35 |
CatKiller | How to technically do it so I can contribute ;) | 15:35 |
*** Shentonfreude1 has quit IRC | 15:36 | |
heckj | CatKiller - in the nova codebase, there's a directory nova/db/ - all the guts of SQLAlchemy are in there. The "tables" are based on the models.py in nova/db/sqlalchemy/models.py and generated by that system | 15:36 |
CatKiller | heckj: That's the thing, I modified the db.models.Volume to add a new field | 15:37 |
heckj | SQLAlchemy itself has pretty good docs on how to set up those models and what optionsa re available | 15:37 |
CatKiller | heckj: but I don't see the results in the sqlite3 db, the field doesn't show up | 15:37 |
*** jaypipes has joined #openstack | 15:37 | |
CatKiller | heckj: Thanks, I'll also look at that, I wasn't sure if this was standard or if this was specific to nvoa. | 15:38 |
aaronAUS | btorch: actually no it isn't starting correctly | 15:40 |
aaronAUS | http://paste.openstack.org/show/1738/ | 15:40 |
*** dobber has quit IRC | 15:41 | |
*** reidrac has quit IRC | 15:42 | |
btorch | aaronAUS: ok try to stop everything " sudo swift-init all stop" , then check to make sure nothing is already listening on 6011 port .. "sudo netstat -tnpl" | 15:42 |
aaronAUS | python is listening on that port | 15:43 |
*** Jamey has quit IRC | 15:43 | |
aaronAUS | can I kill a PID ?? | 15:43 |
btorch | aaronAUS: also double check your configs to make sure you don't have something trying to bind on the same port twice | 15:43 |
*** Jamey has joined #openstack | 15:43 | |
heckj | CatKiller: nova-manage db sync should create/recreate the database after your updates. I know it works on versions to manage forward and reverse migrations, but I'm not 100% clear on how that's all configured | 15:43 |
btorch | aaronAUS: find out what that is first | 15:43 |
btorch | aaronAUS: "sudo ps -ef |grep PID" | 15:44 |
*** mattray has joined #openstack | 15:45 | |
aaronAUS | btorch: checked all configs... only one container-server conf is set to use port 6011 | 15:46 |
btorch | aaronAUS: ok cool, what is attached to that port right now after you stopped everything ? | 15:46 |
aaronAUS | btorch: container-server 1 | 15:47 |
btorch | aaronAUS: ok kill it | 15:47 |
btorch | aaronAUS: then try to start things up again | 15:48 |
aaronAUS | much better | 15:48 |
aaronAUS | btorch: so shall I try to swauth-add-user ???? | 15:48 |
btorch | sure | 15:49 |
*** vernhart has quit IRC | 15:49 | |
aaronAUS | btorch: yay! | 15:49 |
btorch | aaronAUS: a swauth-list will give you info back to verify things | 15:49 |
*** adjohn has quit IRC | 15:50 | |
*** nRy has joined #openstack | 15:51 | |
uvirtbot | New bug: #800759 in nova "Zone Manager does not "expire" offline hosts" [Undecided,New] https://launchpad.net/bugs/800759 | 15:52 |
CatKiller | heckj: Thanks, I figured out | 15:55 |
heckj | CatKiller glad to help | 15:55 |
CatKiller | heckj: The DB is created with sqlachemy.migrate using a migration repo, and the models.py just defines the object model to match the db | 15:55 |
CatKiller | heckj: The models is used internally in nova while the migrate defines the tables. Technically it could take the table definition from the models.py classes but then you'd lose the versioning capability. | 15:56 |
*** mgoldmann has quit IRC | 15:57 | |
*** iRTermite has quit IRC | 15:58 | |
*** vernhart has joined #openstack | 15:58 | |
*** jkoelker has quit IRC | 15:58 | |
aaronAUS | btorch: once again you have come to the rescue | 16:01 |
aaronAUS | btorch: it is working when I curl from within ubuntu | 16:02 |
*** bodepd has joined #openstack | 16:02 | |
bodepd | hi all, I have an installation question | 16:02 |
*** llang629 has joined #openstack | 16:02 | |
aaronAUS | btorch: but unfortunately still not from cyberduck | 16:02 |
*** llang629 has left #openstack | 16:02 | |
bodepd | is there a preferred multi-node installation method? I am looking at http://nova.openstack.org/adminguide/multi.node.install.html | 16:02 |
bonzay | I've got an installation question as well | 16:04 |
heckj | bodepd: nothing definite right now - there's chef recipes to it, an install script (like nova.sh) - that's referenced in that document, and the puppet guys are layout out puppet modules to do the same. It's still a fairly manual process right now | 16:04 |
bodepd | I am that Puppet guy :) | 16:04 |
*** bkkrw has quit IRC | 16:04 | |
bonzay | is there a working xen installation instruction somewhere besides http://wiki.openstack.org/XenServerDevelopment | 16:04 |
heckj | bodepd: Ah, excellent! I've been watching your code updates as they flow in | 16:04 |
bodepd | cool! | 16:05 |
heckj | bonzay: that's really it as far as I know. | 16:05 |
bodepd | if the db is split from the controller | 16:05 |
bodepd | who should run db sync? | 16:05 |
bodepd | the controller? | 16:05 |
heckj | bodepd: controller is probably the most appropriate component if you want/need to assign "responsibility" | 16:06 |
*** mdomsch has quit IRC | 16:06 | |
bonzay | heckj: do i need the part with "upload a SSH key to LP before it will allow you to proceed with the branching. " if i don't want to develop (i just want to install nova) | 16:07 |
bodepd | heckj: thanks! | 16:07 |
*** obino has quit IRC | 16:07 | |
heckj | bonzay: Nope. BZR will let you check out without any ID | 16:07 |
bonzay | heckj: ah ok, thx | 16:07 |
heckj | bonzay: the ssh keys and bzr credentials are only needed if you want to make a patch and submit it | 16:09 |
bonzay | heckj: i thought it is something like that, but i wasnt | 16:09 |
bonzay | sure | 16:09 |
*** Ryan_Lane has joined #openstack | 16:10 | |
*** ziyadb_ has joined #openstack | 16:10 | |
bonzay | can someone have a look at http://paste.openstack.org/show/1739/ | 16:10 |
bonzay | thats what i get when i run sudo pip install -r nova/tools/pip-requires | 16:11 |
*** johnpur has joined #openstack | 16:11 | |
*** ChanServ sets mode: +v johnpur | 16:11 | |
*** jkoelker has joined #openstack | 16:11 | |
bonzay | apparenty the urls are outdated | 16:11 |
*** ziyadb has quit IRC | 16:13 | |
bodepd | adam_g and I were just discussing what node the db sync should be done on. | 16:13 |
bodepd | the easist would be on the DB itself. Could this be problematic (b/c potentially a DB could have multiple instances per controller?) | 16:13 |
bodepd | and just install nova-common on the db | 16:13 |
*** vernhart has quit IRC | 16:14 | |
*** jkoelker has quit IRC | 16:15 | |
adam_g | as i understand it, the first 'nova-manage db sync' is necessary to run the required migrations on the specificied database. i dont believe its bound to any one nova project/controller/instance | 16:15 |
*** dmshelton has quit IRC | 16:16 | |
*** jkoelker has joined #openstack | 16:16 | |
*** chemikadze has quit IRC | 16:16 | |
heckj | adam_g: correct | 16:17 |
heckj | adam_g: I've been using the rough outlines of the nova.sh install script from https://github.com/cloudbuilders/deploy.sh/blob/master/nova.sh | 16:18 |
bonzay | can someone help me with my problem? please? | 16:19 |
bonzay | as i can see there are two major problems | 16:20 |
bonzay | 1. ImportError: No module named greenlet | 16:20 |
*** rnirmal has joined #openstack | 16:20 | |
heckj | bonzay: what are you installing this on? Ubuntu? CentOS? | 16:20 |
bonzay | heckj: ubuntu 10.04.2 LTS | 16:21 |
adam_g | bodepd: if thats the case, i think the best would be bind database migration/sync to some component of the controller since installations might be depending on an external db instance outside the scope of puppet | 16:21 |
heckj | You might riff off the nova.sh script I just mentioned and install the base packages into the OS from that - it'll be reasonably quick | 16:21 |
heckj | bonzay: you could also just "sudo pip install greenlet" - I think it's pip installable | 16:22 |
adam_g | bonzay: install python-greenlet | 16:22 |
bodepd | if we run db sync from teh controller, which subcomponent makes the most sense? | 16:23 |
*** jkoelker has quit IRC | 16:23 | |
heckj | bonzay: here's the list of packages from nova.sh | 16:24 |
heckj | apt-get install -y python-mox python-ipy python-paste python-migrate \ | 16:24 |
heckj | python-gflags python-greenlet python-libvirt python-libxml2 python-routes \ | 16:24 |
heckj | python-netaddr python-pastedeploy python-eventlet python-novaclient \ | 16:24 |
heckj | python-glance python-cheetah python-carrot python-tempita \ | 16:24 |
heckj | python-sqlalchemy python-suds python-lockfile python-m2crypto python-boto | 16:24 |
*** Razique has quit IRC | 16:24 | |
adam_g | bodepd: nova-api seems logical to me, for some reason, since without the proper nova-manage setup the api is useless to users | 16:25 |
bodepd | I was thinking that or the scheduler. | 16:26 |
bonzay | ok looks as if the first problem is solved by installing python-greenlet | 16:26 |
bonzay | thx for that | 16:27 |
aaronAUS | btorch: is there a log i can vieww to see the actual http headers which were passed to swift??? | 16:28 |
*** mies has joined #openstack | 16:28 | |
*** johnpur has quit IRC | 16:28 | |
*** ohnoimdead has joined #openstack | 16:29 | |
*** Jamey has quit IRC | 16:29 | |
bonzay | but now i'm getting this error http://paste.openstack.org/show/1740/ | 16:31 |
*** gaveen has joined #openstack | 16:33 | |
*** rchavik has quit IRC | 16:33 | |
*** kashyap has quit IRC | 16:34 | |
heckj | bonzay: apt-get install python-suds | 16:35 |
*** obino has joined #openstack | 16:35 | |
bonzay | "python-suds is already the newest version." | 16:36 |
*** jfluhmann has joined #openstack | 16:36 | |
bonzay | ah its suds 0.4.1 but the requirement is ==0.4 | 16:37 |
*** Guest3167 has joined #openstack | 16:38 | |
*** anneg has joined #openstack | 16:40 | |
bonzay | will it work if I just change the requirements or should I use ==0.4 | 16:41 |
heckj | bonzay: I would suspect it will just work | 16:41 |
*** bsa has joined #openstack | 16:41 | |
*** iRTermite has joined #openstack | 16:42 | |
bonzay | heckj: ok, thx, I'll try to change the requirements first | 16:42 |
*** markvoelker has joined #openstack | 16:43 | |
bonzay | now cheetah is outdated :) but I think I know now what to do | 16:44 |
heckj | bonzay: I suspect you'll find several outdated pieces there. You should mark up a bug and suggest a patch since you're getting all the details nailed down :-) | 16:44 |
bonzay | heckj: will do :) | 16:45 |
aaronAUS | btorch: OMFG!!!! i finally got it to work | 16:45 |
aaronAUS | I had to change a hidden configuration in cyberduck which sets the authentication.context (eg: /auth/v1.0) | 16:45 |
*** daysmen has quit IRC | 16:49 | |
*** irahgel has quit IRC | 16:49 | |
*** Guest3167 has left #openstack | 16:52 | |
*** kashyap has joined #openstack | 16:52 | |
*** mrjazzcat has joined #openstack | 16:54 | |
*** obino has quit IRC | 16:55 | |
*** mszilagyi has joined #openstack | 16:55 | |
*** obino has joined #openstack | 16:55 | |
*** jdurgin has joined #openstack | 16:57 | |
*** odyi has quit IRC | 17:00 | |
*** Ryan_Lane has quit IRC | 17:00 | |
*** imsplitbit has joined #openstack | 17:01 | |
*** jamiep has joined #openstack | 17:01 | |
aaronAUS | good night all! | 17:03 |
*** aaronAUS has quit IRC | 17:03 | |
*** maplebed has joined #openstack | 17:05 | |
*** katkee has quit IRC | 17:05 | |
*** koolhead17 has joined #openstack | 17:14 | |
*** ctennis has quit IRC | 17:15 | |
*** GeoDud has joined #openstack | 17:16 | |
*** ctennis has joined #openstack | 17:17 | |
*** llang629_ has joined #openstack | 17:18 | |
*** tomh has joined #openstack | 17:21 | |
*** ziyadb has joined #openstack | 17:23 | |
*** tomh has joined #openstack | 17:24 | |
*** imsplitbit has quit IRC | 17:25 | |
*** tomh has quit IRC | 17:25 | |
*** ziyadb_ has quit IRC | 17:25 | |
*** ziyadb_ has joined #openstack | 17:26 | |
*** ziyadb has quit IRC | 17:27 | |
*** ziyadb has joined #openstack | 17:28 | |
*** imsplitbit has joined #openstack | 17:29 | |
*** ziyadb_ has quit IRC | 17:31 | |
*** ziyadb_ has joined #openstack | 17:33 | |
*** ziyadb has quit IRC | 17:36 | |
*** bz has quit IRC | 17:38 | |
*** bz has joined #openstack | 17:39 | |
*** qhead has joined #openstack | 17:40 | |
*** jkoelker has joined #openstack | 17:41 | |
qhead | hey.. I'm researching distributed file storage systems and found Swift.. from the website I got the impression that Proxy Server is SPOF in the system? | 17:41 |
WormMan | only if you don't put in more than one of them | 17:42 |
NashTrash | Nope. You can have N proxy servers | 17:42 |
WormMan | (from what I've read) | 17:42 |
qhead | ah, ok | 17:42 |
qhead | I'm currently trying to decide whether to use Riak's Luwak or Swift | 17:42 |
NashTrash | Be aware that Swift is not a "file system". It just has buckets and blobs inside the buckets. If you need a true filesystem look at things like ceph, MooseFS, Gluster, Lustre, etc. | 17:43 |
qhead | Yeah, I knew that, poor choice of words from my part :/ | 17:44 |
qhead | I'll spread it over several clusters which are basically Ubuntu VPS instances | 17:44 |
qhead | Hmm, I think I'm going to do some benchmarking to see which one has the edge there | 17:45 |
*** huslage has joined #openstack | 17:48 | |
*** cp16net has joined #openstack | 17:50 | |
NashTrash | I have an instance with a public IP address. Â If I try to ping that address from another instance within the cloud I get a time out. Â If I ping the private IP it works just fine. | 17:50 |
NashTrash | Clearly when I use the public IP address it routes through the nova-network machine. | 17:50 |
*** mgoldmann has joined #openstack | 17:51 | |
NashTrash | How can I get one VM to be able to use the public IP address of another VM? | 17:51 |
NashTrash | Is this expected behavior? | 17:51 |
*** ahale has quit IRC | 17:51 | |
*** ahale has joined #openstack | 17:52 | |
*** aliguori has quit IRC | 17:52 | |
huslage | i would like my network (configured as FlatDHCP) to be able to just masquerade outbound traffic from any VM that doesn't have a public IP assigned to it. what's the magic? | 17:52 |
*** koolhead17 has quit IRC | 17:54 | |
*** woleium has joined #openstack | 17:54 | |
*** aliguori has joined #openstack | 17:55 | |
*** ksteward1 has quit IRC | 17:57 | |
*** ksteward has joined #openstack | 17:57 | |
*** obino has quit IRC | 17:57 | |
CatKiller | Hi there, | 17:57 |
*** obino has joined #openstack | 17:57 | |
CatKiller | Would anyone know of a good Python interface to the Nova API, similar to the Euca suite? | 17:57 |
*** anneg has quit IRC | 17:58 | |
*** shawndecuir has joined #openstack | 17:59 | |
heckj | qhead: you can load balance the proxy servers - and there's a general rule of thumb that you want 1 proxy server for every ~8 storage servers | 18:00 |
heckj | NashTrash: yeah, I'm afraid so | 18:01 |
*** ziyadb_ has quit IRC | 18:01 | |
heckj | NashTrash - nova-network manages that routing (and is the bottleneck) in the system today | 18:02 |
heckj | CatKiller - there's a nova-api client library - I believe Dashboard uses it | 18:02 |
NashTrash | heckj: Yeah, I know. But, right now if I try to ping from one VM to another using the pubic IP it fails. Ping with the private IP works. Is this expected? | 18:03 |
* heckj looks for it... | 18:03 | |
CatKiller | heckj: Thanks, I'll check it out. | 18:03 |
NashTrash | CatKiller: nova-adminclient is the name I think | 18:03 |
*** obino1 has joined #openstack | 18:03 | |
huslage | NashTrash: if the IPs are on the same host then it won't work to talk between the public IPs | 18:04 |
heckj | NashTrash - yeah, I believe that's expected. You'll need to fiddle IPTables on your nova-network instance to allow the traffic | 18:04 |
*** obino1 has quit IRC | 18:04 | |
heckj | CatKIller: This is what I was thinking of: http://pypi.python.org/pypi/python-novaclient | 18:04 |
*** obino1 has joined #openstack | 18:05 | |
*** obino has quit IRC | 18:05 | |
NashTrash | heckj: I took a look at the nova-network iptables, but my *pathetic* knowledge of iptables led me to not see where things were getting hung up. | 18:05 |
heckj | github link is broken on PyPi - source is here: https://github.com/rackspace/python-novaclient | 18:05 |
CatKiller | heckj: Thanks, that sounds ok. I guess the issue is that there are so many tools/api that I don't know which is the "best" or the most relevant | 18:07 |
CatKiller | heckj: For instance I had started using euca2ools in python too | 18:07 |
NashTrash | CatKiller: Euca is great. It makes use of Boto underneath. | 18:08 |
*** CloudChris has joined #openstack | 18:08 | |
*** CloudChris has left #openstack | 18:08 | |
*** blamar_ has joined #openstack | 18:08 | |
CatKiller | NashTrash, heckj: So maybe I should use euca2ools instead of python-novaclient | 18:08 |
CatKiller | heckj: Is python-novaclient what the dashboard uses? | 18:08 |
NashTrash | CatKiller: euca2ools is good for command line work, but from python use either boto (AWS API) or python-novaclient (OS API) | 18:09 |
*** wilmoore has quit IRC | 18:09 | |
heckj | CatKiller - historically it was very focused on the euca based APIs (EC2 API setup), but is quickly migrating to the OpenStack API sets | 18:10 |
*** rminnear has joined #openstack | 18:10 | |
*** dragondm has joined #openstack | 18:10 | |
CatKiller | NashTrash: Thanks, I'll look at boto too just to see what it's all about | 18:10 |
*** dragondm has quit IRC | 18:11 | |
CatKiller | heckj: Ok, I'd say I'll give the OS API a try then, thanks! | 18:11 |
*** dragondm has joined #openstack | 18:11 | |
NashTrash | CatKiller: NP | 18:11 |
*** PW__ has joined #openstack | 18:13 | |
bonzay | hi, me again :) | 18:13 |
NashTrash | hola | 18:14 |
bonzay | almost got the xen thing working, but now http://paste.openstack.org/show/1742/ | 18:14 |
*** Funnnny has quit IRC | 18:14 | |
bonzay | when i run_tests.sh it runs ok for 117 tests and then fails | 18:15 |
bonzay | sry for being such a noob regarding python and these errors... | 18:15 |
vishy | bonzay: are you running tests with a running nova? | 18:19 |
bonzay | vishy: no | 18:20 |
*** Ryan_Lane has joined #openstack | 18:22 | |
NashTrash | vishy: Do you have a moment for an iptables/nova-network question? | 18:22 |
NashTrash | vishy: I have two VMs (A and B). If B pings A using A's public IP it times out. If B pings A's private IP it works. What do I need to do to nova-network's iptables to allow B to use A's public IP address? | 18:23 |
*** mdomsch has joined #openstack | 18:24 | |
NashTrash | I have tried understanding iptables enough to get this working, but my kun-fo is not strong here | 18:24 |
*** gaveen has quit IRC | 18:24 | |
*** imsplitbit has quit IRC | 18:25 | |
vishy | euca-authorize | 18:25 |
vishy | NashTrash: euca-authorize -P icmp -t -1:-1 default | 18:25 |
NashTrash | vishy: That has been done. | 18:26 |
vishy | oh right | 18:26 |
vishy | yeah you can't hit via public instance to instance | 18:26 |
vishy | let me think if there is a way you can make it work | 18:26 |
NashTrash | vishy: Still doesn't work. It is only the difference between using the public vs. private IPs. | 18:26 |
vishy | i think if you add the forward to output it will work | 18:27 |
NashTrash | vishy: Sorry, I am not sure I understand. Add which forward to which output? | 18:28 |
*** bonzay is now known as zz_bonzay | 18:31 | |
vishy | one sec | 18:31 |
NashTrash | vishy: No problem. | 18:33 |
adam_g | are the nova.conf options glance_host and glance_port needed if glance_api_servers points to the same? | 18:35 |
*** zz_bonzay is now known as bonzay | 18:35 | |
vishy | adam_g: no glance_api_servers replaces the other options | 18:35 |
*** odyi has joined #openstack | 18:36 | |
*** odyi has joined #openstack | 18:36 | |
vishy | NashTrash: so there is a prerouting rule that nats for floating ips | 18:36 |
vishy | i think if you add the same rule to the output chain it will allow vm traffic to work | 18:36 |
NashTrash | vishy: Hmm…The same rule is already in the nova-network-OUTPUT chain. | 18:38 |
vishy | yeah i was just noticing that | 18:38 |
vishy | :( | 18:39 |
vishy | are you in flatdhcp? | 18:39 |
NashTrash | :) | 18:39 |
NashTrash | vishy: Yes, flatDHCP | 18:39 |
*** gregp76 has joined #openstack | 18:39 | |
vishy | is br100 in promisc mode? | 18:40 |
NashTrash | vishy: Not certain. How do I check? | 18:41 |
vishy | try this on the network host: sudo ip link set promisc on dev br100 | 18:42 |
NashTrash | vishy: Ok, it was not before and now it is. Testing... | 18:42 |
NashTrash | vishy: Still does not work | 18:43 |
NashTrash | :( | 18:43 |
vishy | :( well it was worth a shot :) | 18:43 |
vishy | so i'm thinking that the issue is actually the snat | 18:44 |
NashTrash | the nova-network-snat? It seems to me that it makes everything look like it is coming from the nova-network machine | 18:44 |
vishy | right which is fine for outbound traffic | 18:45 |
NashTrash | but no good for VM to VM | 18:45 |
vishy | can youy try this | 18:45 |
NashTrash | I also thought that this could a problem behind why my 169.254.169.254 metadata stuff never worked. I had to add a NAT entry on each nova-compute | 18:46 |
vishy | iptables -t nat -A nova-network-POSTROUTING -s 10.0.0.0/24 -d <public_ip>/32 -j ACCEPT | 18:46 |
vishy | assuming your private ips are in the 10.0.0.0/24 range | 18:46 |
vishy | NashTrash: the metadata stuff in flat mode is usually due to br100 not being promisc | 18:47 |
*** jesse_ has joined #openstack | 18:47 | |
*** sdadh01 has quit IRC | 18:47 | |
vishy | * flatdhcp | 18:47 |
NashTrash | Still not working. Grr... | 18:49 |
*** katkee has joined #openstack | 18:50 | |
*** wilmoore has joined #openstack | 18:51 | |
*** arun_ has joined #openstack | 18:51 | |
*** arun_ has joined #openstack | 18:51 | |
*** clauden_ has joined #openstack | 18:52 | |
*** jesse_ has quit IRC | 18:52 | |
*** dgags has quit IRC | 18:55 | |
vishy | rhar was my best guess | 18:57 |
vishy | er that | 18:57 |
vishy | going to have to do some tcpdumping to find out what is happenning to the packets | 18:57 |
NashTrash | Ok. Strangely, that rule gets wiped out after about 30 seconds. It is like nova-network clears everything periodically. Kind of strange | 18:58 |
vishy | yes it does | 18:58 |
vishy | the nova chains get recreated from memory | 18:59 |
*** johnpur has joined #openstack | 18:59 | |
*** ChanServ sets mode: +v johnpur | 18:59 | |
vishy | so you can't manually add stuff and have it stick | 18:59 |
vishy | but you can throw it in for testing :) | 18:59 |
NashTrash | Ah. Ok. | 18:59 |
CatKiller | Hi again, I have a quick question about nova zones. I can't find where to add a new zone within nova. Would anyone know where should I look? | 19:01 |
*** alandman has quit IRC | 19:02 | |
*** ctennis has quit IRC | 19:04 | |
*** pvo has joined #openstack | 19:04 | |
NashTrash | vishy: I see a bunch of ARP traffic happen between the two VMs when packets have to go through the nova-network (using public IPs), but no ARP traffic when they go directly between the VMs (using privateIPs) | 19:04 |
uvirtbot | New bug: #800843 in nova "OSAPI: extra 'checksum' property in image metadata" [Undecided,New] https://launchpad.net/bugs/800843 | 19:06 |
*** perestrelka has quit IRC | 19:07 | |
*** irahgel has joined #openstack | 19:08 | |
*** ewindisch has quit IRC | 19:09 | |
*** woleium_ has joined #openstack | 19:09 | |
*** Xenith has quit IRC | 19:09 | |
vishy | NashTrash: does the arp exchange work? Is it followed by actual data packets being sent? | 19:09 |
*** woleium has quit IRC | 19:10 | |
*** woleium has joined #openstack | 19:10 | |
uvirtbot | New bug: #800842 in nova "nova-manage displaying incorrect error msgs on flavor create" [Undecided,In progress] https://launchpad.net/bugs/800842 | 19:12 |
*** irahgel has quit IRC | 19:12 | |
*** woleium_ has quit IRC | 19:13 | |
*** perestrelka has joined #openstack | 19:15 | |
CatKiller | I guess I should probably understand this first: | 19:15 |
CatKiller | What is the concept of availability zones? | 19:15 |
CatKiller | I can't find much on it even after googling a while | 19:16 |
*** PW__ has quit IRC | 19:19 | |
*** Xenith has joined #openstack | 19:19 | |
*** dprince has quit IRC | 19:20 | |
*** PW__ has joined #openstack | 19:20 | |
*** gregp76 has quit IRC | 19:23 | |
*** vasichkin has joined #openstack | 19:23 | |
smoser | is this a known issue with trunk at the moment: | 19:24 |
smoser | http://paste.ubuntu.com/630945/ ? | 19:24 |
bpaluch | smoser you are using an older version of glance with trunk nova | 19:26 |
bpaluch | try using trunk glance | 19:26 |
bpaluch | the older version of glance doesn't support filter | 19:27 |
smoser | thank you. | 19:29 |
bpaluch | also if you are using newer glance the nova.conf attributes have changed. | 19:30 |
vasichkin | Hello. I'm trying to install openstack, diablo-1 on RHEL 6.1. I've done all needed configuration, but when i try to start nova-compute i get trouble with libvrt: (nova.exception): TRACE: libvirtError: internal error Cannot find suitable emulator for x86_64 | 19:30 |
vasichkin | can anybody help? | 19:30 |
*** vasichkin has quit IRC | 19:37 | |
NashTrash | vishy: Sorry, I was AFK. No. The ARP exchange does not seem to work. | 19:55 |
NashTrash | vishy: It appears to me that no traffic that goes from VM to nova-manage to VM works. I very well could be wrong though. | 19:56 |
*** vernhart has joined #openstack | 19:57 | |
*** imsplitbit has joined #openstack | 20:02 | |
*** pvo has quit IRC | 20:03 | |
*** dubs_ is now known as dubs | 20:03 | |
vishy | hmm | 20:04 |
*** mgoldmann has quit IRC | 20:04 | |
*** ctennis has joined #openstack | 20:09 | |
*** ctennis has joined #openstack | 20:09 | |
*** ohnoimdead has quit IRC | 20:11 | |
*** gregp76 has joined #openstack | 20:12 | |
NashTrash | vishy: Should I open a but to track this? I really do feel that this is the same problem I had with 169.254.169.254. | 20:12 |
vishy | NashTrash: 169.254 works fine for me | 20:13 |
NashTrash | vishy: It should be the nova-compute instances that do the iptables magic and not nova-network | 20:13 |
vishy | so i don't think it is the same issue | 20:13 |
NashTrash | vishy: Well then maybe my nova-manage is just wrong somehow | 20:14 |
NashTrash | vishy: If I try to have 169.254 traffic go through nova-manage it does the same ARP-then-die dance | 20:15 |
vishy | arp should work for 169 | 20:15 |
vishy | nova-network adds the ip to loopback | 20:15 |
*** bz has quit IRC | 20:16 | |
vishy | perhaps you are missing a setting somewhere | 20:17 |
NashTrash | I see VM GET to 169.254, it gets nat at nova-manage, and then I see nova-api trying to arp the VM. Nothing happens after tat | 20:17 |
*** woleium has quit IRC | 20:18 | |
vishy | is nova-manage running nova-network and nova-api? | 20:21 |
NashTrash | nova-manage service list shows everything running nicely | 20:22 |
vishy | no i mean is the host that you are referring to nova-manage | 20:22 |
vishy | running the network and api? | 20:22 |
vishy | or are they all different hosts? | 20:22 |
*** AimanA has joined #openstack | 20:22 | |
NashTrash | they are on different machines | 20:22 |
*** dmshelton has joined #openstack | 20:23 | |
vishy | so nova-api is trying to arp for the private address? | 20:23 |
*** rminnear has quit IRC | 20:23 | |
vishy | you need a route on nova-api to private range | 20:23 |
NashTrash | yes | 20:23 |
vishy | route add -net <private_range> gw <ip_of_nova_network> | 20:23 |
vishy | if they are on separate hosts like that | 20:24 |
NashTrash | Makes sense. But I have the same issue if I try to do a GET from a nova-compute machine to the nova-api (via nova-network using 169.254) | 20:24 |
vishy | so if you curl 169.254.169.254 from compute it doesn't work? | 20:25 |
vishy | is nova-network the default gateway for your compute machines? | 20:25 |
*** bz has joined #openstack | 20:26 | |
vishy | it works for vms because it is their gateway | 20:26 |
vishy | it won't necessarily work for the hosts | 20:26 |
*** huslage has quit IRC | 20:26 | |
NashTrash | Wait wait wait…My nova-compute network is 192.168.50.x. If I run route on nova-api it only shows routes for "localnet" and "default". Default points to the nova-network machine. I really should have a route to 192.168.50.0 right?!?! | 20:27 |
NashTrash | nova-compute do have the nova-network as their gateway | 20:27 |
vishy | you should have a route for guests | 20:27 |
vishy | i don't know if you need one for hosts | 20:27 |
vishy | what private range are you using? | 20:27 |
vishy | * fixed_range | 20:27 |
NashTrash | 192.168.51.0/24 | 20:28 |
vishy | and you are using 50.0/24 for hosts? | 20:29 |
*** bkkrw has joined #openstack | 20:29 | |
NashTrash | yes | 20:29 |
vishy | and your guests are getting ips properly? | 20:30 |
*** vernhart has quit IRC | 20:30 | |
NashTrash | yes | 20:30 |
*** jaypipes is now known as jaypipes-afk | 20:31 | |
vishy | might need to add a route for 51.0/24 | 20:31 |
*** vernhart has joined #openstack | 20:32 | |
vishy | also you should make sure that there are no weird iptables rules using the 10.0 range or something weird like that | 20:32 |
NashTrash | Ok. I will try that and keep tinkering. I added an iptables entry on each nova-compute that translates 169.254 to my nova-api and that works. | 20:33 |
NashTrash | Will do | 20:33 |
NashTrash | my nova-api machine has nothing (I mean nothing) in its iptables | 20:33 |
vishy | i'm thining that you need to set --dmz_cidr as well | 20:33 |
NashTrash | what is that used for? | 20:34 |
vishy | to a cidr that includes nova-api | 20:34 |
vishy | there is an accept rule that is added | 20:34 |
vishy | so it doesn't snat traffic to that range | 20:34 |
*** pvo has joined #openstack | 20:35 | |
vishy | otherwise it snats traffic going to nova-api to the network_host ip | 20:35 |
vishy | and the response won't get there | 20:35 |
NashTrash | hmm. Could be | 20:35 |
*** katkee has quit IRC | 20:35 | |
uvirtbot | New bug: #800889 in nova "OSAPI: image metadata item improperly presented" [Undecided,New] https://launchpad.net/bugs/800889 | 20:41 |
*** cbeck has quit IRC | 20:43 | |
*** Xenith has quit IRC | 20:47 | |
*** sdadh01 has joined #openstack | 20:47 | |
*** cbeck has joined #openstack | 20:47 | |
cowmix | I'm setting up a single node to host a few VMs running KVM as the hypervisor.. does using OpenStack still make sense (if it never grows beyond the single node) | 20:48 |
vishy | depends on whether you like to control via cloudy apis | 20:51 |
*** katkee has joined #openstack | 20:53 | |
*** pvo has quit IRC | 20:57 | |
*** nerens has quit IRC | 20:59 | |
*** katkee has quit IRC | 20:59 | |
*** katkee has joined #openstack | 21:00 | |
*** flat has joined #openstack | 21:08 | |
heckj | Reading through the docs - does each compute node in a multi-node installation need to have nova-network installed and running? | 21:09 |
heckj | I thought it was a single instance that everything routed through (at least at this stage) | 21:09 |
NashTrash | heckj: You are right. Single instance with all traffic through it (for now) | 21:11 |
heckj | NashTrash: thanks | 21:11 |
NashTrash | heckj: NP | 21:12 |
cowmix | vishy: the overhead of OpenStack shouldn't be a big deal on one node then? | 21:14 |
vishy | wouldn't thik so | 21:19 |
*** ameade has quit IRC | 21:21 | |
*** msivanes has quit IRC | 21:22 | |
*** qhead has quit IRC | 21:30 | |
*** Davor` has joined #openstack | 21:31 | |
*** exlt_ is now known as exlt | 21:37 | |
*** pvo has joined #openstack | 21:38 | |
*** vernhart has quit IRC | 21:38 | |
*** vernhart has joined #openstack | 21:40 | |
*** flat has quit IRC | 21:43 | |
*** mdomsch has quit IRC | 21:48 | |
*** vernhart1 has joined #openstack | 21:52 | |
*** vernhart has quit IRC | 21:54 | |
*** pguth66 has joined #openstack | 21:55 | |
*** arun_ has quit IRC | 21:58 | |
*** cereal_bars has joined #openstack | 22:01 | |
*** vernhart has joined #openstack | 22:06 | |
*** vernhart1 has quit IRC | 22:08 | |
*** imsplitbit has quit IRC | 22:08 | |
Davor` | Is this the right place to follow-up on my unanswered Launchpad questions? | 22:12 |
*** clauden_ has quit IRC | 22:12 | |
*** qhead has joined #openstack | 22:13 | |
*** cp16net has quit IRC | 22:15 | |
*** vernhart1 has joined #openstack | 22:15 | |
*** vernhart has quit IRC | 22:17 | |
*** PW__ has quit IRC | 22:18 | |
cereal_bars | Davor`: maybe.. tell us about your question :) | 22:19 |
Davor` | I have two machines (controller + compute), but I can't even ping from the controller its own VMs | 22:21 |
*** vernhart1 has quit IRC | 22:21 | |
Davor` | Each host has two NICs, one on the corporate LAN and the other on the private management subnet; plus another subnet (10.x.x.x) for the VMs | 22:21 |
Davor` | Using FlatDHCP manager... | 22:22 |
*** vernhart has joined #openstack | 22:22 | |
heckj | Davor: just checking - you can't ping from the CloudController to a VM running on the compute node? | 22:24 |
*** llang629_ has quit IRC | 22:24 | |
Davor` | No, the VM is running on the cloud controller | 22:24 |
*** markvoelker has quit IRC | 22:26 | |
*** Cromulent has joined #openstack | 22:27 | |
*** bcwaldon has quit IRC | 22:35 | |
*** cereal_bars has quit IRC | 22:38 | |
*** bsa has quit IRC | 22:39 | |
*** amccabe has quit IRC | 22:39 | |
*** victor_lowther has joined #openstack | 22:40 | |
heckj | Davor: sorry, distracted. nova-network is running on that cloud controller? How did you set this up? (I'm wondering what your ifconfig -a and iptables --list looks like on that controller, which should be enabling this) | 22:41 |
*** Cromulent has quit IRC | 22:45 | |
*** mattray has quit IRC | 22:47 | |
Davor` | heckj: Can I paste them here or do you prefer that I use pastebin? | 22:48 |
*** Xenith has joined #openstack | 22:48 | |
heckj | Davor:pastebin please | 22:49 |
*** MotoMilind has quit IRC | 22:51 | |
Davor` | heckj: ifconfig is at http://pastebin.com/7GYPWBa9; iptables at http://pastebin.com/9eqLhCJC | 22:52 |
*** katkee has quit IRC | 22:53 | |
*** bkkrw has quit IRC | 22:56 | |
*** MarkAtwood has quit IRC | 22:57 | |
heckj | Davor: looks reasonable - you have eth0 unplugged? And when the VM comes up, what IP address does it have? | 22:57 |
*** PW__ has joined #openstack | 22:57 | |
*** pvo has quit IRC | 22:58 | |
Davor` | heckj: eth0 is not unplugged, I'm using both NICs on the machine (so that it is on two "real" networks -- corporate and openstack-management) | 22:58 |
Davor` | heckj: and the VM is shown by euca-describe-instances as 10.0.1.2 | 22:58 |
heckj | Davor: eth0 is bound to the br100 then? | 22:59 |
heckj | Davor: the IP tables are all set up for this system to be able to ping 10.0.1.2 | 23:00 |
*** MotoMilind has joined #openstack | 23:00 | |
heckj | Davor: are you getting no route to host when you try? | 23:00 |
*** gregp76 has left #openstack | 23:00 | |
Davor` | heckj: I'm pretty sure eth0 is bound to br100. I'm able to ping its "management network" IP from the compute node | 23:02 |
Davor` | heckj: ah right, that's how it looks from ip addr show: http://pastebin.com/G7Hasenm | 23:04 |
*** rnirmal has quit IRC | 23:05 | |
Davor` | heckj: br100 is bound to eth0's MAC and shows IPs on both the management (192.168.11.1) and VM (10.0.1.1) subnets | 23:05 |
*** jaypipes-afk is now known as jaypipes | 23:06 | |
*** soosfarm has quit IRC | 23:06 | |
heckj | Davor: I'm at a loss beyond here - my IPTables foo isn't cutting this, from what I can see, that node should be completely able to ping 10.0.1.2 | 23:07 |
*** soosfarm has joined #openstack | 23:08 | |
*** aliguori has quit IRC | 23:09 | |
*** miclorb has joined #openstack | 23:11 | |
*** allsystemsarego has quit IRC | 23:11 | |
Davor` | heckj: No worries, thanks for trying. Do you have any suggestions for how I can further debug this? Any logging I can turn on? | 23:13 |
Davor` | heckj: Oh, and one more thing-- how come the log file for the instance is empty? (length 0) | 23:13 |
Davor` | btw, I started LP question 159198 to figure this out | 23:14 |
*** dkim1987 has quit IRC | 23:14 | |
Davor` | heckj: and I noticed that I have two identical dnsmasq processes running | 23:15 |
Davor` | except for the PID, both have the same parameters in 'ps ax' output: 1495 ? S 0:02 dnsmasq --strict-order --bind-interfaces --conf-file= --domain=novalocal --pid-file=/var/lib/nova/networks/nova-br100.pid --listen-address=10.0.1.1 --except-interface=lo --dhcp-range=10.0.0.2,static,120s --dhcp-lease-max=32 --dhcp-hostsfile=/var/lib/nova/networks/nova-br100.conf --dhcp-script=/usr/bin/nova-dhcpbridge --lea | 23:15 |
Davor` | sefile-ro | 23:15 |
heckj | Davor: you probably want to kill one of those DNSMASQ processes, but that shouldn't stop the networking. | 23:15 |
heckj | Davor: not sure about the log file either - nova is mostly set to scroll output to the console from each process, but can be configured to log to a file. And I'm learning the debugging of the networking stuff right now - so I'm afraid I don't have much idea of where to go next except to ask others on this channel. | 23:17 |
Ryan_Lane | Davor`: one of them is the parent of the other | 23:17 |
Davor` | Ryan_Lane: so I shouldn't touch either? | 23:18 |
*** msinhore1 has quit IRC | 23:18 | |
Ryan_Lane | correct. unless they need to be restarted because your configuration changed and the dnsmasq process has the incorrect info | 23:18 |
Ryan_Lane | in that case you'd kill both and restart nova-network | 23:19 |
Ryan_Lane | Davor`: you are using the ubuntu packages? | 23:19 |
Ryan_Lane | the cactus release? | 23:19 |
Davor` | Ryan_Lane: and there is a third one, comming from libvirt, but listening on a different IP and for a different DHCP range | 23:19 |
*** MarkAtwood has joined #openstack | 23:19 | |
Davor` | Ryan_Lane: that's correct. Ubuntu Natty server, cactus release | 23:19 |
Ryan_Lane | *that* one needs to die | 23:19 |
Ryan_Lane | so... | 23:19 |
Ryan_Lane | /etc/init.d/dnsmasq stop | 23:20 |
Davor` | Ryan_Lane: there is no /etc/init.d/dnsmasq | 23:20 |
Ryan_Lane | eh? really? | 23:20 |
Ryan_Lane | ah. right | 23:21 |
Ryan_Lane | do this, then: stop dnsmasq | 23:21 |
Ryan_Lane | stupid upstart | 23:21 |
Ryan_Lane | you'll also need to disable dnsmasq in the upstart file | 23:21 |
Davor` | Ryan_Lane: no such luck -- "stop: Unknown job: dnsmasq" | 23:21 |
Ryan_Lane | that's odd. dnsmasq must be installed. that usually installs an init script too | 23:22 |
Ryan_Lane | somehow that 3rd process is getting started | 23:22 |
*** victor_lowther has quit IRC | 23:23 | |
Ryan_Lane | well. i gotta go. good luck :) | 23:23 |
*** Ryan_Lane has quit IRC | 23:23 | |
*** vernhart has quit IRC | 23:25 | |
Davor` | Me too... Thanks all for your help. To be continued... | 23:25 |
*** Davor` has quit IRC | 23:26 | |
*** adjohn has joined #openstack | 23:26 | |
*** adjohn has quit IRC | 23:26 | |
*** RoAkSoAx has quit IRC | 23:35 | |
*** rminnear has joined #openstack | 23:45 | |
*** pguth66 has quit IRC | 23:53 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!