vidd | jdg, pipeline = authtoken keystone_shim context registryapp is what should be set in there | 00:00 |
---|---|---|
* swill is going to call it a night... | 00:01 | |
jdg | vidd: ok, shall I try and run it again? | 00:01 |
vidd | no | 00:01 |
jdg | ok | 00:01 |
swill | cheers all. have a good evening... | 00:01 |
jdg | I modified glance-registry.conf | 00:02 |
vidd | try to retart all services | 00:02 |
*** swill has quit IRC | 00:02 | |
rmk | OK wow, if anyone is running windows VMs and trying to use vnc, you need this change.. | 00:02 |
rmk | https://bugs.launchpad.net/nova/+bug/885966 | 00:03 |
rmk | Massive improvement in usability | 00:03 |
jdg | Ok... one thing, when it gets to "Restarting glance-registry" is says: "restart: Unknown instance" | 00:03 |
jdg | index still give ECONNREFUSEd | 00:03 |
vidd | jdg, so answer "y" to the "did anything fail" | 00:04 |
vidd | jdg, can you run service glance-registry start | 00:05 |
jdg | vidd: yeah, I did and it gives the same message again. | 00:05 |
jdg | vidd: yes, it started | 00:05 |
jdg | glance-registry start/running process 49427 | 00:06 |
vidd | jdg, now run service glance-registry restart | 00:06 |
jdg | Ok.. good | 00:06 |
vidd | now try the index | 00:06 |
jdg | Same error | 00:06 |
jdg | It's the glance-api that's the problem. Restart won't work on it | 00:07 |
*** dolphm has quit IRC | 00:07 | |
vidd | what is glance-api complaining about? | 00:07 |
jdg | stacker@sfstack-38:~/openstack_installer$ sudo service glance-api restart | 00:07 |
jdg | restart: Unknown instance: | 00:07 |
vidd | jdg, run service glance-api start | 00:08 |
vidd | then restart unless you get an error | 00:08 |
jdg | WTF? run command not found | 00:08 |
vidd | jdg, loose the "run" =] | 00:09 |
jdg | Oh.. duhhh | 00:09 |
vidd | just service glance-api start | 00:09 |
jdg | I'm getting tired/lazy | 00:09 |
jdg | just copy verbatim | 00:09 |
jdg | OK... here we have some error info | 00:09 |
vidd | good | 00:09 |
jdg | stacker@sfstack-38:~/openstack_installer$ service glance-api start | 00:09 |
jdg | start: Rejected send message, 1 matched rules; type="method_call", sender=":1.221" (uid=1000 pid=51221 comm="start glance-api ") interface="com.ubuntu.Upstart0_6.Job" member="Start" error name="(unset)" requested_reply="0" destination="com.ubuntu.Upstart" (uid=0 pid=1 comm="/sbin/init") | 00:09 |
vidd | well...not really but its going to help us | 00:10 |
*** Pommi has quit IRC | 00:10 | |
*** dolphm has joined #openstack | 00:10 | |
vidd | service glance-api start | 00:11 |
jdg | vidd: same error message | 00:12 |
vidd | ok...some major issues with glance | 00:12 |
vidd | we are going to back up and punt...... | 00:12 |
jdg | :) | 00:12 |
jdg | How far back shall we go? | 00:13 |
vidd | copy the "Turn on Keystone for Glance" section of the convert_keystone script to the wtf script | 00:13 |
*** zorzar has quit IRC | 00:13 | |
vidd | let me know when that is done | 00:13 |
jdg | Ok.. done | 00:14 |
*** dolphm has quit IRC | 00:14 | |
vidd | apt-get autoremove --purge glance && ./glance_setup.sh && ./wtf.sh | 00:15 |
jdg | hmmm... | 00:16 |
*** CaptTofu has quit IRC | 00:16 | |
jdg | installing Glance errors | 00:16 |
vidd | what errors | 00:16 |
jdg | installing Glance | 00:17 |
jdg | E: Could not open lock file /var/lib/dpkg/lock - open (13: Permission denied) | 00:17 |
jdg | E: Unable to lock the administration directory (/var/lib/dpkg/), are you root? | 00:17 |
*** CaptTofu has joined #openstack | 00:17 | |
_rfz | yu need to be root jdg | 00:17 |
jdg | Got it... | 00:17 |
jdg | sorry | 00:17 |
vidd | jdg, have you been doing all this as standard user? my scripts require root | 00:17 |
*** Pr0toc0l has quit IRC | 00:18 | |
jdg | yes, but I've been doing "sudo" for everything, just missed the additional commands | 00:18 |
_rfz | vidd - have you upgraded to kiall' new packages? | 00:18 |
vidd | _rfz, my scripts use Kiall 's ppa | 00:18 |
jdg | vidd: ok, api, registry and scrubber have all started | 00:18 |
_rfz | vidd - yes but I mean have you upgrade your openstack installation packages? | 00:19 |
_rfz | I've upgraded mine with new libvirtd and something broke | 00:19 |
vidd | did the wtf script work? | 00:19 |
jdg | It appeasr to have | 00:19 |
jdg | Shall I try the index command again? | 00:19 |
vidd | yes | 00:20 |
vidd | you may need to restart the glance-api and glance-registry first | 00:20 |
jdg | DAMN IT JIM!! | 00:21 |
jdg | Same error | 00:21 |
*** lzyeval has joined #openstack | 00:21 | |
jdg | Shall I go all the way back to the beginning at this point and try again? | 00:22 |
vidd | sed -e "s,999888777666,$AUTH_TOKEN,g" -i /etc/glance/glance-registry.conf sed -e "s,999888777666,$AUTH_TOKEN,g" -i /etc/glance/glance-api.conf | 00:22 |
vidd | baaaa | 00:22 |
vidd | sudo sed -e "s,999888777666,$AUTH_TOKEN,g" -i /etc/glance/glance-registry.conf && sudo sed -e "s,999888777666,$AUTH_TOKEN,g" -i /etc/glance/glance-api.conf | 00:22 |
*** Pommi has joined #openstack | 00:22 | |
vidd | then restart api and registry | 00:23 |
jdg | WTF??? sed: can't read sed: No such file or directory | 00:23 |
jdg | got it, hold on | 00:24 |
vidd | wait..... | 00:24 |
jdg | Ok | 00:24 |
vidd | you need add the sed lines to the wtf script | 00:24 |
*** CaptTofu has quit IRC | 00:24 | |
jdg | Finished the two sed commands, haven't restarted anything | 00:25 |
jdg | Ok | 00:25 |
vidd | and run the wtf script again | 00:25 |
vidd | ok....check that the auth token was added to the two config files | 00:25 |
vidd | since the auth token was not added to the ENV you probably have them set to "" | 00:26 |
*** krow has quit IRC | 00:26 | |
jdg | Hmm... nope they weren't added | 00:27 |
Kiall | vidd, i take it your scripts are far from foolproof so? ;) | 00:28 |
*** rustam has joined #openstack | 00:28 | |
jdg | kiall: are you implying something there? | 00:28 |
vidd | Kiall, when user does "sudo" instead of actually being root .... =] | 00:28 |
Kiall | lol | 00:28 |
*** pweiss has left #openstack | 00:28 | |
Kiall | yea.. ah well ;) | 00:28 |
Kiall | Wait | 00:28 |
Kiall | That should work.. -_- | 00:29 |
jdg | Kiall: it didn't work with your script either. | 00:29 |
Kiall | really? | 00:29 |
vidd | no....no it shouldnt | 00:29 |
jdg | Nope, the script didn't even run before bombing out | 00:29 |
Kiall | pretty sure it should, but, whatever! | 00:29 |
jdg | kiall: Here's your output | 00:30 |
jdg | ERROR: (OperationalError) (1045, "Access denied for user 'root'@'localhost' (using password: YES)") None None | 00:30 |
jdg | root : ERROR (OperationalError) (1045, "Access denied for user 'root'@'localhost' (using password: YES)") None None | 00:30 |
*** lin has joined #openstack | 00:30 | |
Kiall | jdg, you didnt read the instructions so ;) | 00:30 |
Kiall | The part under the title "Grant root remote access to MySQL" | 00:31 |
vidd | Kiall, "sudo" appears to fool my "are you root" test but there are atherthings that sudo cant do (so it would appear) | 00:31 |
jdg | vidd: Rather than waste your time shall I just re-image and run as root instead of sudo? | 00:31 |
Kiall | vidd, once you sudo, everything ran by that script is also ran as root, it should work.. | 00:32 |
jdg | kiall: I did follow the README | 00:32 |
Kiall | weird, its worked for me on a clean install.. no extra steps... | 00:32 |
vidd | jdg, its not a waste of my time....but if you wish to...then yes | 00:32 |
_rfz | sudo/root is same | 00:33 |
_rfz | you have root axs | 00:33 |
jdg | It's your call, if I run as root and hit the same problem though that would SUCK! | 00:33 |
*** lin has quit IRC | 00:33 | |
jdg | _rfz: that's always been my assumption | 00:33 |
_rfz | jdg run as root :) | 00:33 |
Kiall | _rfz, I'm pretty sure once you sudo, your root. I've never hit something that didnt work like that | 00:33 |
_rfz | same | 00:34 |
vidd | _rfz, my assumption too...but there is this strange issue =\ | 00:34 |
*** rsampaio has joined #openstack | 00:34 | |
jdg | The only thing I wonder is if maybe there was a multi command line that I didn't catch and add the additional "sudo" to. | 00:34 |
_rfz | kiall - btw i've upgraded to your latest packages - with new libvirtd0 (from ubuntu) and openstack is dead :) | 00:34 |
jdg | Seems like I'd see something else befoer now though | 00:35 |
vidd | jdg, if you reimage, do "sudo su then run the scripts | 00:35 |
Kiall | _rfz, really? I upgraded everything that was available earlier... | 00:35 |
jdg | vidd: Ok, I'll kick that off tonight and see how it goes. | 00:35 |
jdg | Thanks for trying to help me out with this mess. | 00:35 |
vidd | jdg, there IS a possibility that my script has a bug in it somewhere | 00:36 |
*** arun has quit IRC | 00:36 | |
jdg | Actually, I'm going to leave the one node as is and run it as root on another system. | 00:37 |
Kiall | _rfz, any specific errors? | 00:37 |
vidd | but did you add the auth token to both glance config files? | 00:37 |
Kiall | I've 100% updated and everything is working for me.. | 00:37 |
*** miclorb_ has quit IRC | 00:37 | |
Kiall | I'm* | 00:37 |
jdg | No, the sed command didn't work. Shall we try that before punting for the evening? | 00:37 |
vidd | yes...add it manually | 00:38 |
vidd | both need the auth key | 00:38 |
*** debo-os has joined #openstack | 00:39 | |
vidd | both need to have keystone auth port set to 35357 | 00:39 |
vidd | then both need to be [re]started | 00:39 |
jdg | vidd: is it admin_toke or auth_token? | 00:39 |
jdg | admin_token | 00:39 |
vidd | $AUTH_TOKEN | 00:40 |
jdg | Ok, so just add the variable to the file anywhere? It doesn't exist currently | 00:40 |
vidd | from your server-path | 00:40 |
vidd | no...not the variable...the actual data | 00:41 |
Kiall | I think i just beat my own record.. how long ago did i come back online? | 00:41 |
Kiall | Start->Finish excluding ubuntu install.. since then ;) | 00:41 |
vidd | Kiall, you want to try my scripts? | 00:42 |
vidd | or should i say "PLEASE try my scripts" =] | 00:42 |
Kiall | So long as its press the button and go! Hate getting asked questions the entire way through ;) | 00:42 |
jdg | Ok, here's what I have in my registry.conf | 00:42 |
jdg | [filter:authtoken] | 00:42 |
jdg | paste.filter_factory = keystone.middleware.auth_token:filter_factory | 00:42 |
jdg | service_protocol = http | 00:42 |
jdg | service_host = | 00:42 |
jdg | service_port = 5000 | 00:42 |
jdg | auth_host = | 00:43 |
jdg | auth_port = 35357 | 00:43 |
jdg | auth_protocol = http | 00:43 |
jdg | auth_uri = http://:5000/ | 00:43 |
jdg | admin_token = | 00:43 |
Kiall | uhh auth_uri = http://:5000/ is wrong | 00:43 |
_rfz | kiall - http://pastebin.com/C2zv0k09 | 00:43 |
Kiall | _rfz, i had that error when i first tried OS.. But, even then i have no idea how I fixed it | 00:43 |
vidd | jdg all the urls are missing | 00:44 |
*** rnorwood has joined #openstack | 00:44 | |
jdg | I know, since we ran the WTF script everything is gone. | 00:44 |
jdg | That's what I was trying to say. | 00:44 |
_rfz | kiall it's weird - it just stopped working | 00:44 |
vidd | i tried to get to that.... | 00:45 |
jdg | So I added the ip to t service_host and auth_host | 00:45 |
vidd | we need to add ". server-path" to the wtf script.... | 00:45 |
vidd | and add the sed lines to the wtf script | 00:45 |
jdg | Ok | 00:46 |
Kiall | _rfz, have you given libvirt-bin a restart? | 00:46 |
jdg | got the sed lines, doing the server-path now | 00:46 |
Kiall | and then nova-compute after that | 00:46 |
jdg | Ok, now I understand what you were saying about the AUTH_TOKEN | 00:47 |
*** catbot has quit IRC | 00:47 | |
jdg | Ok, I've put the server-path entries at the top of the WTF.sh script | 00:47 |
jdg | shall I run: apt-get autoremove --purge glance && ./glance_setup.sh && ./wtf.sh | 00:47 |
vidd | jdg, make this your wtf script: http://paste.ubuntu.com/746545/ | 00:48 |
Kiall | the fact that you have a wtf script worries me :P | 00:48 |
vidd | =] safer this way =] | 00:48 |
jdg | Ok, done | 00:49 |
vidd | Kiall, its left over from my testing...i dont know how to pull it from the git repo =] | 00:49 |
jdg | YOu left it for me ;) | 00:49 |
vidd | jdg, now do the remove and reinstall =] | 00:49 |
*** deshantm_laptop has quit IRC | 00:49 | |
vidd | Kiall, by default the wtf script is empty =] | 00:50 |
Kiall | sure LO | 00:51 |
Kiall | :P | 00:51 |
jdg | Done | 00:51 |
vidd | and your jelous...i know =] | 00:51 |
*** marrusl has quit IRC | 00:51 | |
Kiall | lol | 00:51 |
vidd | jdg, ok...try the index command | 00:51 |
*** livemoon has joined #openstack | 00:51 | |
jdg | raise exception.NotAuthorized() | 00:52 |
*** ldlework has joined #openstack | 00:52 | |
_rfz | kialll - i've restarted it - still same error | 00:53 |
jdg | http://paste.openstack.org/show/3512/ | 00:53 |
jdg | glance-registry.conf still doesn't have the token | 00:55 |
Kiall | _rfz, i know I did what that error, but frankly, I have no clue how I fixed it. | 00:55 |
Kiall | I can say its not a version issue, "apt-get update && apt-get dist-upgrade" shows nothing for me, so I'm on the latest of everything aswell | 00:56 |
Kiall | i know I did have that error* | 00:56 |
*** adjohn has joined #openstack | 00:56 | |
livemoon | morning | 00:56 |
vidd | jdg, get the token from ./obtain-token.sh | 00:57 |
_rfz | kiall - no probs. Just wanted to make sure | 00:58 |
jdg | Ok | 00:58 |
*** rnorwood has quit IRC | 00:58 | |
vidd | now, copy it and put it in this line " glance -A $COPY_OF_TOKEN index | 00:59 |
vidd | replace $COPY_OF_TOKEN with the token of course | 00:59 |
jdg | NotAuthorized() | 01:00 |
jdg | http://paste.openstack.org/show/3514/ | 01:00 |
vidd | ok...something has to be wrong with keystone | 01:00 |
Kiall | _rfz, a reboot may help ;) | 01:01 |
Kiall | otherwise maybe #libvirt | 01:01 |
_rfz | tried it also :) stopped appamor and nada | 01:01 |
Kiall | thats basically a libvirt error wrapped in an OS exception | 01:01 |
Ryan_Lane | soren: so, our mysql client library is built with binary as the default charset. nova (I guess specifically sqlalchemy) doesn't like this. at all. | 01:02 |
Ryan_Lane | even adding ?charset=utf8 to the sql_connection flag doesn't seem to help | 01:03 |
jdg | vidd: So the token is in the database /keystone/token/ | 01:03 |
Ryan_Lane | it worked for nova-manage, but does not for the services | 01:03 |
vidd | yes the token is in there | 01:03 |
vidd | ./obtain-token gets it | 01:04 |
jdg | Yeah, good point | 01:04 |
jdg | Ok, I'll try running as root on a clean image and see what happens. | 01:05 |
vidd | there has to be an issue | 01:05 |
vidd | lets wipe the keystone database and run the ./keystone_template.sh | 01:06 |
*** dotdevops has quit IRC | 01:07 | |
jdg | Ok | 01:07 |
jdg | Best way to wipe, Just nuke it? | 01:07 |
vidd | go into phpmyadmin and delete all the tables | 01:08 |
jdg | Okee | 01:08 |
jdg | All tables are dropped | 01:09 |
jdg | running keystone_template.sh now | 01:09 |
vidd | as root i hope | 01:09 |
jdg | :) yes as root | 01:10 |
vidd | have you done the sudo su? | 01:10 |
jdg | yes | 01:10 |
debo-os | Hi once I have devstack running and I want to reinstall cleanly to include options etc, whats the std practice for doing that. | 01:11 |
vidd | debo-os, install the apps on a real machine | 01:11 |
*** dachary has quit IRC | 01:12 | |
jdg | Same friggin problem | 01:12 |
jdg | Wait, let me restart glance | 01:13 |
vidd | you want to restart keystone as well | 01:13 |
jdg | can I just run your restart.sh? | 01:13 |
vidd | restart does not restart keystone | 01:14 |
jdg | No luck | 01:16 |
jdg | restarted glance-reg, glance-api and keyston | 01:16 |
jdg | This sucks | 01:16 |
vidd | did the obtain token work? | 01:17 |
jdg | yes | 01:17 |
jdg | I need to head out of the office, will be back on later tonight. Maybe I'll try on a different system and then bug you tomorrow if I have the same problem. | 01:18 |
vidd | ill be here for a while yet | 01:19 |
jdg | Ok, thanks again for your help and your patience | 01:19 |
*** chomping has joined #openstack | 01:19 | |
*** mattstep has joined #openstack | 01:19 | |
mattstep | is there a way to setup a flavor that can oversubscribe with kvm? | 01:20 |
debo-os | vidd: i installed on a fresh ubuntu install | 01:21 |
*** hadrian_ has joined #openstack | 01:22 | |
*** woleium has quit IRC | 01:23 | |
Kiall | mattstep, all flavors can oversubscribe, as in, you can start 10 1G instances with 5GB ram.. | 01:23 |
vidd | debo-os, it is recommended to install from the ppa's (for 11.10 *buntu, i recommend Kiall's ppa) | 01:23 |
debo-os | devstack? | 01:24 |
debo-os | I want to change the code and reinstall for testing | 01:24 |
debo-os | so I cant use the ppa | 01:24 |
vidd | i have no advice for that | 01:24 |
Kiall | devstack is desgigned to give you a development environment, so yea. devstack is fine for that | 01:24 |
debo-os | ok how do you stop devstack instance | 01:24 |
*** sg[] has joined #openstack | 01:24 | |
Kiall | designed& | 01:24 |
Kiall | designed* | 01:24 |
*** miclorb_ has joined #openstack | 01:25 | |
debo-os | Kiall: then how do I reinstall with maybe some other components | 01:25 |
*** hadrian has quit IRC | 01:25 | |
*** hadrian_ is now known as hadrian | 01:25 | |
Kiall | debo-os, I believe you just re-run the script.. but bear in mind, the way devstack works you cant remove a component.. | 01:27 |
*** sg has quit IRC | 01:27 | |
Kiall | the OS dev's suggest using devstack inside a throwaway VM | 01:27 |
debo-os | just running gives me errors with multiple screen sessions | 01:27 |
debo-os | aah i c | 01:27 |
mattstep | Kiall : Can I instruct the scheduler to consider tiny's as not being allocated a full cpu versus smalls that are, as in, when I pack instances, I want to pack small/medium/large/etc. each with a dedicated cpu and then tinys without a dedicated cpu (so I don't oversubscribe a compute node without tinys, but I do oversubscribe a compute node if I have any tinys on it) | 01:27 |
Kiall | mattstep, not out of the box.. the the scheduler is interchangable... a little bit of custom code would probably let you do that | 01:28 |
mattstep | ok, cool | 01:28 |
*** dirakx1 has joined #openstack | 01:29 | |
*** pixelbeat has quit IRC | 01:29 | |
*** MarkAtwood has quit IRC | 01:29 | |
*** bryguy has quit IRC | 01:30 | |
*** bryguy has joined #openstack | 01:31 | |
sticky | http://www.hydis.com/eng/03_products/productKind.asp?L1=7&L2=2&pdG_Code=6&pd_Num=182&pdvName=Industrial&openBox=5&subBox=5&openSubBox=1 | 01:33 |
sticky | er | 01:33 |
sticky | wrong window, sorry | 01:33 |
*** jdurgin has quit IRC | 01:36 | |
*** dirakx1 has quit IRC | 01:39 | |
_rfz | Kiall, are you running a multinode openstack? | 01:41 |
Kiall | yep | 01:42 |
_rfz | does vnc work on all your nodes? | 01:42 |
*** lorin1 has joined #openstack | 01:43 | |
Kiall | yup | 01:43 |
_rfz | I get novnc working on my controller+compute node, but not on the others | 01:43 |
Kiall | what nova.conf VNC settings do you have? | 01:43 |
_rfz | on the compute host? | 01:44 |
livemoon | vidd, hi | 01:44 |
*** ohnoimdead has quit IRC | 01:44 | |
vidd | hello livemoon | 01:44 |
Kiall | _rfz, the nova.conf file should usually be identical on all nodes | 01:44 |
_rfz | ok | 01:44 |
_rfz | --vncserver_host=$CONTROLLER_IP | 01:44 |
_rfz | --vncproxy_port=6080 | 01:44 |
_rfz | --vncproxy_url=http://$CONTROLLER_IP:6080 | 01:44 |
Kiall | ---vncserver_host=$CONTROLLER_IP is your problem | 01:45 |
Kiall | ---vncserver_host=0.0.0.0 | 01:45 |
Kiall | use --vncserver_host=0.0.0.0* | 01:45 |
_rfz | cheers | 01:45 |
Kiall | its a bind address, rather than a connect to this address if you get me | 01:45 |
vidd | hehe i saw that =] | 01:45 |
Kiall | and.. the change will only apply to new instances | 01:45 |
Kiall | (that IP gets embedded in the libvirt XML) | 01:45 |
vidd | _rfz, i had the same issue | 01:46 |
rmk | So what is Rackspace using internally as their hypervisor? I've heard it's Xenserver. | 01:46 |
*** jimw_ has joined #openstack | 01:46 | |
_rfz | it is | 01:46 |
*** lzyeval has quit IRC | 01:46 | |
livemoon | vidd: what are u doing these days about openstack? | 01:46 |
rmk | Yeah so I'm curious why Xenserver as opposed to KVM. | 01:46 |
_rfz | Kiall, thanks | 01:46 |
_rfz | rackspace is moving to openstack | 01:47 |
vidd | livemoon, waiting for servers to do an actual install on | 01:47 |
rmk | Just curious what factors went into that decision. | 01:47 |
_rfz | rmk, i'd say timing | 01:47 |
rmk | As in, Xenserver was the best choice when they picked a hypervisor N years ago? | 01:47 |
_rfz | I'd be curious to see what they run when they migrate to openstack | 01:47 |
rmk | Not to say it isn't a good choice now. | 01:47 |
_rfz | rmk, yep | 01:48 |
rmk | Yeah that's fair. | 01:48 |
rmk | We picked Xenserver too a few years ago. | 01:48 |
jimw_ | xen is in the kernel now | 01:48 |
rmk | yes | 01:48 |
coli | xen != Xenserver | 01:48 |
rmk | Not the same at all. | 01:48 |
coli | (or I'm missing something) | 01:48 |
jimw_ | true | 01:49 |
*** mgius has quit IRC | 01:49 | |
*** woleium has joined #openstack | 01:49 | |
jimw_ | but having xen accepted in the kernel gives xenserver credibility | 01:49 |
rmk | Xen was always credible. | 01:50 |
coli | xen is not bad, been using it since version 1 | 01:50 |
rmk | No one is doubting that. | 01:50 |
coli | kvm is all new to me :-) | 01:50 |
rmk | I was just curious why Rackspace selected Xenserver. Not because it's a bad choice, just wanted to know what led to that decision one way or another. | 01:50 |
rmk | Mainly to compare their reasons against mine (for example). | 01:51 |
coli | I'm wondering what has rackspace added to openstack (steroids ?) to make it into production. | 01:51 |
rmk | coli: Heh | 01:51 |
rmk | coli: I lot of the pitfalls of Openstack can be negated operationally. | 01:51 |
*** rsampaio has quit IRC | 01:51 | |
rmk | Which is what we're planning to do. | 01:51 |
coli | rmk: they needed a strong enterprise player/name to go with them into these uncharted waters at that stage ? | 01:52 |
_rfz | coli, sign up to the beta | 01:52 |
jimw_ | seems like openstack/xenserver is not as good a fit as cloudstack/xenserver | 01:52 |
livemoon | vidd: Do you have installation doc? | 01:52 |
_rfz | Rackspace are offering openstack private cloud edition beta | 01:52 |
*** crescendo has quit IRC | 01:52 | |
vidd | docs? no ... i have an installer script | 01:52 |
coli | vidd: rotfl | 01:53 |
coli | vidd: I will record that quote as well ;-) | 01:53 |
_rfz | coli, http://www.rackspace.com/cloud/blog/2011/09/01/openstack-compute-enters-testing-stage-at-rackspace/ | 01:53 |
_rfz | sorry it's http://www.rackspace.com/cloud/blog/2011/09/01/openstack?-compute-enters-testing-stage-at-rackspace/ | 01:54 |
rmk | jimw_: Absolutely agree | 01:54 |
rmk | jimw_: Cloudstack works fairly well with Xenserver | 01:54 |
coli | if I only had 2 good openstack developers on my side.... | 01:54 |
coli | wishful thinking :-) | 01:54 |
rmk | coli: What issues are you having? | 01:55 |
Kiall | lol - i didnt even know you could use the â„¢ symbol in a URL.. nice ;) | 01:55 |
jimw_ | cloudstack being citrix now should be interesting | 01:55 |
coli | rmk: time or lack of it. no other issues. | 01:55 |
jimw_ | how about HP with ubuntu/openstack though? they are the ones I am watching | 01:55 |
rmk | Using Ubuntu here also. | 01:55 |
*** miclorb_ has quit IRC | 01:55 | |
rmk | Started Openstack with Xenserver and found it to be pretty rough. | 01:56 |
jimw_ | Openstack seems to be going the linux foundation route and that is very encouraging | 01:56 |
uvirtbot | New bug: #893840 in nova "test_libvirt:LibvirtUtilsTestCase.test_get_fs_info fails on Mac OS X" [Undecided,New] https://launchpad.net/bugs/893840 | 01:56 |
livemoon | vidd: your own script? | 01:56 |
vidd | yes | 01:56 |
*** stewart has quit IRC | 01:56 | |
jimw_ | I stood up cloudstack against rhel/rvm last week | 01:56 |
jimw_ | not too bad, but too many vendors in that picture | 01:56 |
rmk | jimw_: Yeah, Cloudstack + XCP is a decent combo though. | 01:56 |
_rfz | anyone used cloudstack? | 01:57 |
livemoon | vidd: Is it public? | 01:57 |
rmk | _rfz: In testing, yes. | 01:57 |
rmk | There are caveats with all of these solutions. | 01:57 |
vidd | yes..on github | 01:57 |
jimw_ | _rfz: me too | 01:57 |
*** stewart has joined #openstack | 01:57 | |
rmk | Cloudstack is much easier to get going. | 01:57 |
rmk | No comparison actually. | 01:57 |
rmk | You can have it running within hours. | 01:57 |
vidd | livemoon, https://github.com/vidd/openstack_installer | 01:57 |
jimw_ | works with vmware too if that is a consideration | 01:58 |
Kiall | rmk, openstack takes 15-30 minutes to install ;) | 01:58 |
*** troya has joined #openstack | 01:58 | |
rmk | Kiall: Sure, for us, since we've done it 50 times. :) | 01:58 |
_rfz | yep | 01:58 |
_rfz | openstack is a bitch to install | 01:58 |
jimw_ | slow and steady wins the race | 01:59 |
rmk | It's not really the installation which makes it difficult. Configuration and understanding all the components is time consuming. | 01:59 |
vidd | openstack is easy as pie to install | 01:59 |
_rfz | rmk, jimw_ what other pro's vs con's did you see with cloudstack vs openstack? | 01:59 |
rmk | vidd: Please. No it is not. | 01:59 |
jimw_ | I more excited for where it will be in two years.... | 01:59 |
vidd | its the configuation that is a biatch | 01:59 |
_rfz | okay vidd, your right :) | 01:59 |
rmk | I agree though at the conference I kept hearing that installation was hard. It's configuration which is hard. | 01:59 |
_rfz | but it's a bitch | 01:59 |
Kiall | most of the configuation was okay, there were only a few small things.. | 01:59 |
Glacee | rfz_ : I could go for hours doing comparaison between both :) | 01:59 |
Kiall | I dont understand why everyone thinks is so hard -_- | 01:59 |
rmk | Kiall: Introduce keystone and it's impossible unless you've been at this a while. | 02:00 |
_rfz | Glacee, give me a sumary | 02:00 |
coli | _rfz: i'm interested as well | 02:00 |
rmk | Kiall: Basic nova with deprecated auth is easy. | 02:00 |
livemoon | Kiall: If I want to learn nova code, how can I start it? | 02:00 |
Kiall | My first nova install took under an hour.. My second tried to use the ubuntu keystone packages (fail).. My 3rd was the day after, with all my own .deb's | 02:00 |
*** _diana_ has joined #openstack | 02:00 | |
Glacee | hmm cloudstack is more mature.. has way more features.. but the architecture is completely different | 02:00 |
Kiall | Okay.. maybe not the day after.. but not too far off ;) | 02:00 |
Glacee | so far.. openstack seems more adequate if you want to build extremly large webscale clouds | 02:01 |
debo-os | kiall: have you tried using quantum with devstack | 02:01 |
Kiall | no | 02:02 |
HugoKuo__ | 12345 | 02:02 |
Kiall | I'll take a look at it after essex is release, until then.. whatever, I dont need it | 02:02 |
HugoKuo__ | test test | 02:02 |
debo-os | kiall: makes sense | 02:02 |
*** reiddraper has left #openstack | 02:03 | |
livemoon | HugoKuo: morning | 02:03 |
*** dpippenger has quit IRC | 02:03 | |
coli | glacee: in your opinnion cloud with up to 200 nodes is better off with cloudstack or openstack ? | 02:03 |
livemoon | vidd: Does your script use Kiall's ppa for all ? | 02:03 |
Glacee | coli:depends on your use case :0 | 02:03 |
vidd | livemoon, yes | 02:04 |
livemoon | vidd: where you set the PPA | 02:04 |
vidd | in the "basics" script | 02:04 |
livemoon | oh, I see it | 02:05 |
livemoon | who can tell me how to learn nova source code? | 02:05 |
vidd | ? | 02:05 |
livemoon | I want to read nova code, how can I start it? | 02:06 |
*** wariola has joined #openstack | 02:06 | |
*** jimw_ has quit IRC | 02:06 | |
coli | Glacee: it seems cloudstack is Citrix now | 02:06 |
*** wariola has quit IRC | 02:07 | |
vidd | livemoon, the souce for Kiall 's ppas is here: https://github.com/managedit/ | 02:07 |
rmk | coli: It's hard to say honestly. | 02:07 |
rmk | coli: Depends on exactly what you want from your private cloud. | 02:07 |
_rfz | cloustack looks interesting | 02:08 |
*** stewart has quit IRC | 02:08 | |
*** rsampaio has joined #openstack | 02:09 | |
*** lorin1 has quit IRC | 02:10 | |
coli | rmk: make money providing similar service in public cloud to AWS and in private cloud compared to VM Ware experience | 02:10 |
rmk | yeah openstack is better suited to that | 02:10 |
coli | rmk: cloudstack has nicer looking dashboard :-) | 02:11 |
rmk | cloudstack has an ec2 api as a bolt on product | 02:11 |
rmk | coli: no question | 02:11 |
coli | rmk: and that sells.... | 02:11 |
rmk | yep | 02:11 |
_rfz | coli, we are building our own dashboard | 02:11 |
WormMan | I'm glad I don't have to deal with actual 'customers' | 02:11 |
_rfz | we also want to sell public cloud services | 02:11 |
rmk | openstack has always been more focused on the api being solid as opposed to the dash | 02:12 |
rmk | I mean consider there was no dash for a long time | 02:12 |
Kiall | rmk, yup.. | 02:12 |
coli | _rfz: you are lucky that you have means and ppl, I only have the first one. | 02:12 |
rmk | And realistically when you're dealing with hundreds or thousands of VMs, you're not very likely to be using a gui anyway. | 02:12 |
rmk | Especially if you're encorporating what makes IaaS compelling, which is dynamic scaling of resources (i.e. spinning VMs up and back down as needed) | 02:13 |
coli | rmk: you would be suprised how many large organisations on daily bases use the aws webfrontend instead of api | 02:13 |
_rfz | I agree coli | 02:13 |
rmk | coli: I don't doubt that at all | 02:14 |
rmk | I think both are equally important | 02:14 |
coli | rmk: aws has boomed since they have made the web interface accessible | 02:14 |
rmk | But if I had to choose one, I'd pick a strong API | 02:14 |
*** rustam has quit IRC | 02:14 | |
coli | rmk; I would as well, however we are able to make use of it, what's sad is that mose it ppl are unable to use it. | 02:14 |
_rfz | coli, they say globo.com is going to opensource their openstack dashboard. and from what people say it's the best by far | 02:15 |
rmk | With a strong API you'll see plenty of alternate dashboards soon enough. | 02:16 |
vidd | given a choice between strong api/weak gui; weak api/strong gui; and medioker api and gui....then i would choice mediocre | 02:16 |
coli | _rfz: i'm quickly googling globo :-) | 02:16 |
_rfz | coli, it was mentioned in the last openstack meetup | 02:16 |
*** mcclurmc has quit IRC | 02:17 | |
rmk | vidd: That's what you have now to be honest. | 02:17 |
*** mcclurmc_ has joined #openstack | 02:17 | |
troya | hi all | 02:18 |
vidd | rmk, yeah | 02:18 |
troya | hi vidd :) | 02:18 |
vidd | hello troya | 02:18 |
troya | vidd, i have remove all of database and generate it again | 02:19 |
Kiall | anyone else find its way too easy to hit "terminate" rather than "reboot" or "log" in the console? ;) | 02:19 |
WormMan | speaking of meetups/summits, anyone know where the 1st 2012 one is yet? | 02:19 |
*** nci has left #openstack | 02:19 | |
troya | vidd: what is different --fixed_range_v4 and --fixed_range ? | 02:19 |
Kiall | in the dashboard* | 02:19 |
coli | _rfz: I cannot find any info about globo.com dashboard, do you have any urls handy ? | 02:20 |
vidd | troya, dunno | 02:20 |
*** primeministerp has quit IRC | 02:21 | |
*** crescendo has joined #openstack | 02:22 | |
vidd | troya, i was not aware there was a --fixed_range_v4 | 02:23 |
vidd | i see "v6" but not v4 | 02:23 |
uvirtbot | New bug: #893843 in openstack-integration-tests "Create server with personality file that overwrites an existing file" [Undecided,New] https://launchpad.net/bugs/893843 | 02:26 |
troya | vidd: ok nothing, it's only curious about it | 02:26 |
uvirtbot | New bug: #893844 in openstack-integration-tests "Server personality file contents exceeds compute limit" [Undecided,New] https://launchpad.net/bugs/893844 | 02:26 |
troya | vidd | 02:27 |
troya | vidd: i have question about my ifconfig output, check it out please >> http://pastebin.com/Q4bKVFtR | 02:28 |
vidd | looks good | 02:29 |
troya | why on virbr0 was get ip 192.168.122.1, ? | 02:29 |
troya | where it get from ? | 02:29 |
vidd | because that is what libvirt sets it to by default | 02:29 |
troya | oo..it's automatically from nova ? | 02:30 |
Glacee | rmk: cloudstack really improved their api+ performance.. there is a cloudstack setup that was able to spin 10 000+vm one shot | 02:30 |
vidd | troya, no...it gets it from libvirt | 02:30 |
Glacee | they improved sec groups performance etc.. | 02:31 |
WormMan | I'd say something snide about how bad the openstack api code is in some places...(yay, get_all and then regexp match the IP address for cloud-init data) | 02:31 |
troya | but is it generate automatically ? or not ? because i have never configure float ip | 02:31 |
vidd | its done automagicly | 02:31 |
uvirtbot | New bug: #893845 in openstack-integration-tests "Server metadata key exceeds 255 bytes" [Undecided,New] https://launchpad.net/bugs/893845 | 02:31 |
uvirtbot | New bug: #893846 in openstack-integration-tests "Metadata value exceeds 255 bytes on server creation" [Undecided,New] https://launchpad.net/bugs/893846 | 02:31 |
coli | glacee: I heard just few hours ago that someone had to put nova-api on each compute node as it had problems when run only on managment node (if I unrestood correctly0 | 02:31 |
*** rnorwood has joined #openstack | 02:31 | |
troya | ok, now i'll create new insntaces, i hope there is no problem again :) | 02:32 |
WormMan | coli: I did that, it only had problems once I got beyond 100 launches at a time or so | 02:32 |
*** p0 has quit IRC | 02:33 | |
*** adjohn has quit IRC | 02:35 | |
coli | wormman: I'm wondering how did you solve talking to so many nova-api's at the same time to spread the load ? | 02:36 |
troya | vidd, i cannot ping instances again :( | 02:36 |
*** rnorwood has quit IRC | 02:36 | |
troya | it;s output of ifconfig and nova.conf >> http://pastebin.com/cpFutzdC | 02:36 |
troya | Kiall, are you still here ? | 02:36 |
vidd | troya, what have you done differently since last night? | 02:37 |
coli | wormman: I'm probably missing something, but as I undrestood you should be talkint to one nova-api which passes info to the rest of the system | 02:37 |
WormMan | coli: the issue was only the euca api calls made by the instances (169.254.169.254) so I just set ec2_dmz_host to be the local compute node IP | 02:38 |
WormMan | then when nova-network came up, it added the nat rule to the local host | 02:38 |
troya | i have not change anything, i have removed database only | 02:39 |
coli | wormman: it's gibbrish to me right now, however I just can get glimps of the idea, will remember to look into it when time comes. | 02:40 |
coli | wormman: thank you | 02:40 |
vidd | well if you have done nothing different, then why wouldyou expect to get different output? | 02:40 |
troya | anyone, can help me ?it's make me dizzy | 02:40 |
*** osier has joined #openstack | 02:41 | |
coli | just reading up on network vlan network mode, it seems that each tenant (project) get's it's own private vlan over which all instances communicate. taking into account L2 switching infrastructure it has a limitation of 4096 vlans (AFAIR), therefore do I undrestand correctly that using vlan network mode we cannot have more than approx. 4080 tenants ? | 02:47 |
WormMan | correct, this(I believe) is what things like quantum are trying to solve, among other issues | 02:49 |
coli | that's a massive limitation for large clouds, I wonder how rackspace is tackling it now. afaik quantum has some way to go before it's stable | 02:50 |
WormMan | I believe rackspace is handling it by not using VLAN mode | 02:52 |
coli | flatdhcp ? | 02:52 |
*** _diana_1 has joined #openstack | 02:52 | |
*** _diana_ has quit IRC | 02:53 | |
WormMan | not sure if flatdhcp or just plain flat, but it's only a rumor | 02:53 |
uvirtbot | New bug: #893847 in openstack-integration-tests "Number of metadata items exceeds compute limit on server creation" [Undecided,New] https://launchpad.net/bugs/893847 | 02:53 |
uvirtbot | New bug: #893848 in swift "can not make an exist container to public" [Undecided,New] https://launchpad.net/bugs/893848 | 02:53 |
coli | on the other hand it's intersting if vlan id is assigned dynamicly to tenant when first instance is started or is it fixed per tenant. if the second then it would mean that vlan ids are used up even if tenat is not generating money | 02:55 |
*** mattray has quit IRC | 02:55 | |
WormMan | the openstack default is hard assigned vlans | 02:55 |
WormMan | (I have 3, for my 3 tenants, all hard coded) | 02:55 |
WormMan | I also have a slightly not-customer type cloud | 02:56 |
coli | it's fine for large private clouds, but not for public use like aws with high number of small tenants. | 02:56 |
coli | do you know by any chance if AWS uses dhcp or injects the ip when imaging ? | 02:57 |
WormMan | my ec2 image looks to be dhcp | 02:57 |
WormMan | AWS/EC2 | 02:58 |
coli | I need to setup AWS account :-) | 02:58 |
coli | I belive aws is using xen ? | 02:58 |
WormMan | yes | 02:58 |
vidd | troya, have you looked into using Kiall 's scripts? | 02:58 |
WormMan | they have a free tier if you want to check out aws | 02:59 |
WormMan | http://aws.amazon.com/free/ | 02:59 |
troya | no i dont, where i get it ? | 02:59 |
coli | thank you. | 02:59 |
WormMan | it's always good to know the competition :) | 02:59 |
sticky | I'm not a fan of AWS to be honest | 02:59 |
vidd | troya, https://github.com/managedit/openstack-setup | 02:59 |
coli | I'm new to this cloud business, just learning | 03:00 |
*** mattstep has quit IRC | 03:00 | |
*** miclorb_ has joined #openstack | 03:00 | |
WormMan | strangely, our customers are also not a fan of AWS :) | 03:00 |
vidd | coli, you get a stack up and running? | 03:00 |
coli | vidd: same as yestarday, I had only two installs so far. I have asked for servers to be reprovisioned and probably in the monring (that is in few hours) someone will play with it. | 03:02 |
*** cmasseraf has joined #openstack | 03:02 | |
coli | vidd: I have asked for her to test your and kiall scrips | 03:02 |
*** primeministerp has joined #openstack | 03:03 | |
*** etarasov has quit IRC | 03:04 | |
*** rods has quit IRC | 03:04 | |
Kiall | Okay .. Clean 100% ubuntu install, Starting OS install @ 5 min past.. lets see how long this takes ;) | 03:05 |
*** Ryan_Lane has quit IRC | 03:06 | |
Kiall | (When i say clean install.. I mean, I installed ubuntu, logged in as root. Nothing more) | 03:07 |
*** bengrue has quit IRC | 03:07 | |
*** CaptTofu has joined #openstack | 03:07 | |
coli | a moment of truth :-) | 03:07 |
Kiall | coli, I'm about 50% of the way done already ;) | 03:08 |
coli | does it have nice graphical progress bar ? ;-) | 03:08 |
Kiall | keystone installed, glance installed, images uploading to glance... | 03:09 |
vidd | hehe | 03:09 |
*** debo-os has quit IRC | 03:09 | |
*** lorin1 has joined #openstack | 03:09 | |
troya | vidd:would you like to tell me more about flat dhcp injected? | 03:09 |
troya | is it same with flat dhcp only ? | 03:10 |
vidd | dunno ... dont use it | 03:10 |
troya | why vidd? | 03:11 |
coli | troya: is there such a thing as flat dhcp injected ? | 03:11 |
troya | coli, do you know about it ? can you tell me about it more? | 03:12 |
Kiall | done | 03:12 |
*** MarkAtwood has joined #openstack | 03:12 | |
Kiall | thats ... umm 7 minutes for keystone+nova+glance+dashboard+image upload ;) | 03:13 |
Kiall | including downloading everything ;) | 03:13 |
vidd | Kiall, wow | 03:13 |
coli | troya: I though that there was just "flat" of "flatdhcp", first injects ip during starting of instance, second grabs it from dhcp | 03:13 |
troya | then if flatdhcp on nova.conf i inserted --flat_injected=true ? what will happen ? | 03:14 |
vidd | Kiall, can you test my script to see if there is some issue? | 03:15 |
coli | troya; no idea | 03:15 |
Kiall | sure.. now that I've fixed everything up in mine ;) | 03:15 |
vidd | its not as fast as yours though | 03:15 |
*** mdomsch has quit IRC | 03:15 | |
Kiall | vidd, your one stops and asks questions rather than getting all that info upfront ;) | 03:16 |
Kiall | your's* | 03:16 |
vidd | right | 03:16 |
Kiall | gah and needing to chmod +x everything ;) | 03:17 |
Kiall | at* | 03:17 |
Kiall | vidd, it broke | 03:18 |
vidd | where? | 03:18 |
Kiall | it asked me if i would like to install the basics | 03:19 |
*** _rfz has quit IRC | 03:19 | |
Kiall | i typed "yes", and it moaned at me -_- | 03:19 |
vidd | y..not yes =] | 03:19 |
Kiall | and why is it installing a new kernel? lol | 03:19 |
vidd | what?!? | 03:19 |
*** etarasov has joined #openstack | 03:19 | |
Kiall | yea - | 03:19 |
Kiall | update-initramfs: Generating /boot/initrd.img-3.0.0-13-server | 03:19 |
Kiall | okay x11-common? | 03:20 |
vidd | its doing a dist-upgrade | 03:20 |
Kiall | why install that? | 03:20 |
Kiall | Pretty sure thats not a dep of anything... | 03:20 |
vidd | yes it is | 03:20 |
Kiall | really? havent noticed it get pulled it before | 03:21 |
vidd | probably python-pastewebkit | 03:21 |
*** CaptTofu has quit IRC | 03:23 | |
Kiall | pretty sure I'd be running VMs at this stage with mine :P | 03:24 |
Kiall | Error: invalid command 'set_permissions .* .* .* | 03:25 |
Kiall | your RMQ setup seems off | 03:25 |
vidd | =\ that never had an issue for me | 03:26 |
Kiall | well, its a real issue ;) | 03:28 |
coli | is there ubuntu 11.10 netinstall ? | 03:28 |
coli | i have decided to check kiall's claims of 7 minutes ;-) | 03:28 |
Kiall | coli, yea.. http://archive.ubuntu.com/ubuntu/dists/oneiric/main/installer-amd64/current/images/netboot/ | 03:29 |
*** krow has joined #openstack | 03:29 | |
Kiall | coli, go for it! | 03:29 |
coli | thanks, couldn';t find it on the webpage and documentation for pxe provisioning was suggesting going the nfs route | 03:30 |
vidd | rabbitmqctl set_permissions hermod '.*' '.*' '.*' works for me in cli | 03:30 |
Kiall | I'm sure that does, there is a username between the regexe's and the command | 03:30 |
Kiall | regex's | 03:30 |
Kiall | lol @ "Did any fail?" | 03:31 |
vidd | Kiall, how did you pass making the database file? | 03:31 |
Kiall | vidd, there are so god damn many questions -_- | 03:31 |
Kiall | if asked me if i wanted to use it or create a new one, i said new one since it was empty.. | 03:32 |
*** dirakx1 has joined #openstack | 03:32 | |
Kiall | it just .. moved on! | 03:32 |
*** livemoon has left #openstack | 03:32 | |
vidd | it creates a new one | 03:33 |
Kiall | *please* say this is nearly done -_- | 03:33 |
*** livemoon has joined #openstack | 03:33 | |
Kiall | woot - done... | 03:33 |
Kiall | now, does any of it work ;) | 03:33 |
Kiall | yea - nova is complaing about rabbitmq, glance is saying "NotAuthorized" .. | 03:35 |
*** dysinger has joined #openstack | 03:36 | |
Kiall | nearly there vidd ;) | 03:36 |
Kiall | coli, done yet? ;) | 03:36 |
coli | nope, just downloaded the netinstall, setingup the dhcp now | 03:37 |
*** primeministerp has quit IRC | 03:37 | |
*** primeministerp has joined #openstack | 03:37 | |
vidd | rabbit is not working because you somehow missed getting rabbit added to the database file | 03:38 |
*** primeministerp has quit IRC | 03:39 | |
*** primeministerp has joined #openstack | 03:40 | |
vidd | Kiall, do you see something wrong with the glance auth? | 03:40 |
Kiall | nope, nothing obvious.. | 03:41 |
Kiall | glance-test.sh just moans | 03:41 |
*** primeministerp1 has joined #openstack | 03:42 | |
*** primeministerp has quit IRC | 03:42 | |
Kiall | vidd, maybe i hit y+enter twice or something.. who knows! There were way too many questions -_- (just kidding) | 03:43 |
*** redconnection has quit IRC | 03:43 | |
Kiall | seriousally though, asking questions sucks. A config file means its repeatable... | 03:43 |
vidd | most of the scripts check for the file and accepts it | 03:45 |
*** primeministerp has joined #openstack | 03:45 | |
*** primeministerp1 has quit IRC | 03:45 | |
*** ejat has joined #openstack | 03:46 | |
*** ejat has joined #openstack | 03:46 | |
Kiall | coli, just updated the README with evey step necessary.. other than needing 2x NICs in the box ;) | 03:46 |
*** hallyn has quit IRC | 03:47 | |
coli | kiall, I got worse problem, I cannot find any test server which is got managment card with standard test password :-( | 03:48 |
Kiall | lol | 03:48 |
coli | it's not very often nowdays that I do these things.... lack of time. | 03:48 |
Kiall | stop trolling free-server-lists .. they arent as common as free proxy lists ;) | 03:48 |
*** primeministerp1 has joined #openstack | 03:49 | |
*** primeministerp has quit IRC | 03:49 | |
*** primeministerp has joined #openstack | 03:51 | |
*** primeministerp1 has quit IRC | 03:51 | |
chadh | Anyone here doing openstack with xen? | 03:51 |
Kiall | whoever set up primeministerp deserves a pat on the back, rather than just blocking IRC, they seem to have allowed the connection for 2-3 seconds, then kill it ;) | 03:52 |
Kiall | whoever set up primeministerp firewall deserves a pat on the back* | 03:52 |
livemoon | hi | 03:53 |
*** primeministerp has quit IRC | 03:53 | |
livemoon | does nova use libvirt and visrh to control instance? | 03:54 |
*** PeteDaGuru has quit IRC | 03:54 | |
*** primeministerp has joined #openstack | 03:54 | |
*** maplebed has quit IRC | 03:54 | |
primeministerp | Kiall: sorry I was tweaking my firewall | 03:55 |
primeministerp | Kiall: trying to get my altq queues right | 03:56 |
Kiall | primeministerp, lol.. I was just kidding ;) | 03:56 |
primeministerp | Kiall: I know | 03:56 |
primeministerp | Kiall: it was funny because I was the one tweaking the firewall | 03:56 |
*** dirakx1 has quit IRC | 03:56 | |
primeministerp | ;) | 03:56 |
Kiall | ;) | 03:56 |
coli | I'm so far back, I just found out that we have out local ubuntu repository :-) | 03:57 |
vidd | Kiall, have you looked any further into whats up with my script and glace/keystone? | 03:57 |
*** llang629 has joined #openstack | 03:59 | |
*** ejat has quit IRC | 04:05 | |
*** llang629 has left #openstack | 04:05 | |
*** ejat has joined #openstack | 04:07 | |
*** ejat has joined #openstack | 04:07 | |
*** JStoker has quit IRC | 04:11 | |
*** pradeep1 has joined #openstack | 04:12 | |
*** krow has quit IRC | 04:14 | |
*** jkyle has quit IRC | 04:15 | |
*** krow has joined #openstack | 04:16 | |
coli | got one ;-) | 04:17 |
coli | just starting pxe | 04:17 |
coli | how much space do you leave for dom0 ? | 04:17 |
*** JStoker has joined #openstack | 04:18 | |
*** krow has quit IRC | 04:21 | |
Kiall | coli, the VMs will live in the dom0 disk .. nova-volume stuff will be outside.. | 04:21 |
Kiall | so its hard to give a number | 04:21 |
Kiall | brb | 04:21 |
coli | VMs will live in dom0 disk ? somehow I dont undrestand. nova-volume will provide storage so how domU (VMs) will ive in dom0 ? | 04:22 |
*** debo-os has joined #openstack | 04:23 | |
coli | grrrr.... got everything working pxe, instaler, management card,.... guess what... ubuntu installer doesn't see my keyboard/mouse | 04:24 |
*** rnorwood has joined #openstack | 04:26 | |
Kiall | nova-volume provides the equilavant of amazon EBS | 04:26 |
*** dpippenger has joined #openstack | 04:26 | |
Kiall | the VMs themselves live in /var/lib/nova/instances, which can be backed by NFS/ISCSI or local disk etc | 04:27 |
*** dirakx1 has joined #openstack | 04:28 | |
*** fifieldt has quit IRC | 04:31 | |
*** hadrian has quit IRC | 04:36 | |
*** abecc has quit IRC | 04:38 | |
coli | kiall: thank you, I will investigate this further tomorrow, going to sleep now it's nearly 5am in here :-( | 04:39 |
*** JStoker has quit IRC | 04:39 | |
*** juddm has joined #openstack | 04:39 | |
*** ejat has quit IRC | 04:40 | |
*** JStoker has joined #openstack | 04:41 | |
etarasov | I have downloaded qcow2 ubuntu image from http://uec-images.ubuntu.com/releases/11.10/release-20111011/ How can I get root password for it? | 04:41 |
*** supriya has joined #openstack | 04:42 | |
*** troya has quit IRC | 04:42 | |
*** helfrez has joined #openstack | 04:44 | |
*** juddm has quit IRC | 04:44 | |
*** Nithya_HP has joined #openstack | 04:44 | |
*** CaptTofu has joined #openstack | 04:45 | |
*** CaptTofu has quit IRC | 04:50 | |
*** dotdevops has joined #openstack | 04:55 | |
*** stanchan has quit IRC | 04:55 | |
*** stanchan has joined #openstack | 04:56 | |
*** binbash_ has quit IRC | 04:56 | |
*** binbash_ has joined #openstack | 04:56 | |
*** debo-os has quit IRC | 04:58 | |
*** lorin1 has quit IRC | 04:58 | |
*** stanchan has quit IRC | 05:00 | |
*** lts has quit IRC | 05:02 | |
*** debo-os has joined #openstack | 05:04 | |
*** nerens has joined #openstack | 05:09 | |
*** lzyeval has joined #openstack | 05:10 | |
*** jdg__ has joined #openstack | 05:16 | |
*** cmasseraf has quit IRC | 05:17 | |
*** dnjaramba has joined #openstack | 05:19 | |
*** junaid has joined #openstack | 05:22 | |
*** helfrez has quit IRC | 05:25 | |
*** debo-os has quit IRC | 05:26 | |
*** rsampaio has quit IRC | 05:27 | |
*** krow has joined #openstack | 05:39 | |
*** supriya has quit IRC | 05:41 | |
*** troya has joined #openstack | 05:45 | |
*** jseutter has quit IRC | 05:48 | |
*** clauden has quit IRC | 05:48 | |
troya | hi all | 05:49 |
etarasov | hi | 05:50 |
troya | hi etarasov, i have some problm from yesterday that haven't solved yet | 05:51 |
troya | i use flat dhcp for my nova network, i can run instances, but i cannot access it via ssh and cannot to ping it.whereas i have configure security rule to allow port 22 | 05:52 |
troya | also icm | 05:52 |
*** jedi4ever_ has joined #openstack | 05:53 | |
*** clauden has joined #openstack | 05:53 | |
*** jedi4ever has quit IRC | 05:55 | |
*** jedi4ever_ is now known as jedi4ever | 05:55 | |
*** JesperA has quit IRC | 05:56 | |
etarasov | I have not reached that stage in configuration yet, I've just run instances without networking setup | 05:57 |
*** stewart has joined #openstack | 05:59 | |
*** localhost has quit IRC | 06:00 | |
*** CaptTofu has joined #openstack | 06:00 | |
*** mattstep has joined #openstack | 06:01 | |
*** localhost has joined #openstack | 06:01 | |
troya | etarasov: then you can ping your instances ? | 06:02 |
*** dirakx1 has quit IRC | 06:02 | |
*** stanchan has joined #openstack | 06:04 | |
*** debo-os has joined #openstack | 06:06 | |
*** debo-os_ has joined #openstack | 06:07 | |
*** yeming has joined #openstack | 06:10 | |
*** martine has quit IRC | 06:10 | |
*** debo-os has quit IRC | 06:11 | |
*** debo-os_ is now known as debo-os | 06:11 | |
*** vladimir3p has joined #openstack | 06:14 | |
*** dachary has joined #openstack | 06:23 | |
*** yeming has quit IRC | 06:23 | |
*** krow has quit IRC | 06:28 | |
*** debo-os has quit IRC | 06:30 | |
*** dachary has quit IRC | 06:31 | |
*** stewart has quit IRC | 06:33 | |
*** dotdevops has quit IRC | 06:35 | |
etarasov | troya: not yet, I don't know how to setup ip addresses for instances | 06:36 |
*** koolhead17 has joined #openstack | 06:37 | |
*** troya has quit IRC | 06:40 | |
*** krow has joined #openstack | 06:41 | |
*** evxd has joined #openstack | 06:42 | |
*** miclorb_ has quit IRC | 06:48 | |
*** stanchan has quit IRC | 06:50 | |
*** troya has joined #openstack | 06:52 | |
*** JStoker has quit IRC | 06:53 | |
*** krow has quit IRC | 06:55 | |
uvirtbot | New bug: #893881 in keystone "OS-KSCATALOG Create EndpointTemplate documentation missing Name attribute" [Undecided,New] https://launchpad.net/bugs/893881 | 06:56 |
*** JStoker has joined #openstack | 06:58 | |
*** arBmind has joined #openstack | 06:59 | |
*** ejat has joined #openstack | 07:01 | |
*** ejat has joined #openstack | 07:01 | |
*** arBmind has quit IRC | 07:01 | |
*** vladimir3p has quit IRC | 07:02 | |
troya | hi all | 07:06 |
koolhead17 | hi all | 07:06 |
troya | i have post my issue to https://answers.launchpad.net/nova/+question/179716 | 07:06 |
troya | hi koolhead | 07:06 |
troya | :) | 07:06 |
evxd | hai | 07:08 |
troya | hi exvd | 07:09 |
*** mgoldmann has joined #openstack | 07:09 | |
troya | evxd: i have some issue that haven't solved, may be you could to help me to solve this, check it out >> https://answers.launchpad.net/nova/+question/179716 | 07:10 |
evxd | you've got the wrong guy.. i'm new to this. :] | 07:12 |
troya | guys, there's wrong to try :) | 07:12 |
troya | there is no wrong :D | 07:12 |
*** kaigan_ has joined #openstack | 07:12 | |
*** arBmind has joined #openstack | 07:16 | |
*** mindpixel has joined #openstack | 07:17 | |
*** TheOsprey has joined #openstack | 07:18 | |
*** JStoker has quit IRC | 07:23 | |
*** arBmind has quit IRC | 07:24 | |
*** JStoker has joined #openstack | 07:29 | |
troya | vidd | 07:34 |
troya | are you there ? | 07:34 |
vidd | yes im here | 07:36 |
troya | actually flat dhcp can run on one ethernetcard machine, right ? | 07:38 |
*** zz_bonzay is now known as bonzay | 07:38 | |
troya | you work on 1 ethernet card with Flat DHCP too? | 07:38 |
vidd | yes | 07:39 |
*** guigui has joined #openstack | 07:39 | |
*** Razique has joined #openstack | 07:40 | |
troya | i have post my issue on https://answers.launchpad.net/nova/+question/179716 , i hope anyone will help me to solve this | 07:40 |
troya | hi Razique :) | 07:40 |
Razique | hey troya | 07:41 |
Razique | what's up . | 07:41 |
troya | yeah,,i'm very dizzy with my problem. somedays ago i have installed nova on ubuntu 11.10.everythings works well but when i try to acces instnaces.i can't | 07:42 |
troya | i have post my issue on https://answers.launchpad.net/nova/+question/179716 , i hope anyone will help me to solve this | 07:42 |
*** negronjl has quit IRC | 07:43 | |
troya | Razique: check it out, may be you can help me to solve my issue | 07:43 |
Razique | troya: let me see | 07:44 |
troya | https://answers.launchpad.net/nova/+question/179716 , check it out please | 07:44 |
Razique | ok can I see both commands output : select * from netwoks, and select * from fixed_ips | 07:45 |
Razique | oh and cat /var/lib/nova/networks/br100.conf | 07:45 |
*** adjohn has joined #openstack | 07:45 | |
*** yeming has joined #openstack | 07:48 | |
troya | Razique: i have paste to patstebin >> http://pastebin.com/58qbyFuH | 07:49 |
*** arBmind has joined #openstack | 07:49 | |
troya | there are two on br100.conf, once again >> 02:16:3e:7d:c4:b4,server-2.novalocal,10.0.0.35 | 07:49 |
Razique | troya: ok let me have a look :) | 07:51 |
troya | ok, please | 07:51 |
Razique | ok | 07:51 |
Razique | looks like you have too many networks here :p | 07:51 |
*** wariola has joined #openstack | 07:52 | |
Razique | I'd say that would explain the issue | 07:52 |
*** debo-os has joined #openstack | 07:52 | |
Razique | why 3 networks ? | 07:52 |
troya | which one ,Razique ? | 07:52 |
Razique | select * from networks; | 07:52 |
troya | o..yes, there are 3 | 07:53 |
Razique | ok | 07:53 |
Razique | first | 07:53 |
Razique | nova-manage network delete 10.0.0.32/28 | 07:54 |
Razique | nova-manage network delete 10.0.0.16/28 | 07:54 |
*** popux has joined #openstack | 07:54 | |
troya | yes, i have | 07:54 |
troya | then ? | 07:54 |
troya | is it wrong to make 3 networks once time, razique? | 07:55 |
Razique | delete from fixed_ips where project_id=NULL | 07:55 |
Razique | no waiy | 07:55 |
Razique | have u run it ? : | 07:55 |
Razique | wait* | 07:55 |
*** vidd is now known as vidd-away | 07:56 | |
Razique | no not really, but you need to make sure they are consistent | 07:56 |
uvirtbot | New bug: #893891 in horizon "Incorrect handling of expired tokens" [Undecided,New] https://launchpad.net/bugs/893891 | 07:56 |
*** shaon has joined #openstack | 07:56 | |
troya | Unknown column 'project_id' on fixed_ips | 07:56 |
Razique | ok great | 07:57 |
troya | Unknown column 'project_id' on fixed_ips, is it right ? | 07:57 |
Razique | now nova-manage network --modify --network 10.0.0.0/28 --project $project_name | 07:58 |
Razique | troya: yah it's cool | 07:58 |
troya | ok i have >> nova-manage network modify --network 10.0.0.0/28 --project successproject | 07:59 |
Razique | is the name's project ? | 07:59 |
*** stanchan has joined #openstack | 07:59 | |
troya | successproject is my project (it's for motivation with my dizzy =D) | 07:59 |
Razique | hehe | 07:59 |
Razique | ok great | 08:00 |
troya | ok then ? | 08:00 |
Razique | select * from networks; | 08:00 |
troya | this is output >> http://pastebin.com/hcC1ptmB | 08:01 |
troya | there is only one | 08:02 |
Razique | great | 08:02 |
Razique | now delete from fixed_ips where network_id=NULL | 08:02 |
Razique | then select * from fixed_ips | 08:02 |
troya | wait, on network table there is column dhcp_start which have value=2 | 08:03 |
troya | but on my nova.conf i have start dhcp with >> --flat_network_dhcp_start=10.0.0.34 . is it can make problem later ? | 08:04 |
troya | or i have change to 10.0.0.2 on my nova.conf first? | 08:04 |
livemoon | hi Razique | 08:05 |
Razique | hi | 08:05 |
*** reidrac has joined #openstack | 08:06 | |
troya | i have >> mysql> delete from fixed_ips where network_id is null; | 08:07 |
troya | there are 32 rows have delted | 08:07 |
troya | now only data which project_id=1 on fixed_ips | 08:08 |
*** lzyeval has quit IRC | 08:09 | |
etarasov | nova-compute at some point calls qemu-img. How to diagnose this error: http://pastebin.com/U7xyEfuk ? | 08:11 |
troya | razique: please wait for about 10 minutes, i have something which i must doing | 08:14 |
*** lzyeval has joined #openstack | 08:15 | |
*** quake has joined #openstack | 08:17 | |
*** dysinger has quit IRC | 08:21 | |
troya | Razique: ok, i have done | 08:22 |
*** arBmind has quit IRC | 08:22 | |
*** arun has joined #openstack | 08:24 | |
Razique | troya: ok cool :) | 08:24 |
Razique | now killall dnsmasq && service nova-network restart | 08:25 |
Razique | and try to run an instance again | 08:25 |
*** ejat has quit IRC | 08:25 | |
*** dachary has joined #openstack | 08:26 | |
troya | only one process dnsmasq which doesn't killed >> root 8755 2924 0 15:25 pts/0 00:00:00 grep --color=auto dnsmasq | 08:26 |
Razique | yah that is the grep itself :p | 08:26 |
*** quake has quit IRC | 08:27 | |
troya | must i terminate my previouse instances which running ? | 08:27 |
*** guigui has quit IRC | 08:27 | |
Razique | or reboot them ? | 08:27 |
*** ejat has joined #openstack | 08:27 | |
*** ejat has joined #openstack | 08:27 | |
troya | you are very cool Razique :D | 08:29 |
troya | now it's works :D | 08:29 |
Razique | hehe | 08:29 |
Razique | good job man | 08:29 |
troya | hahaha..it's realy succesproject ,right | 08:30 |
*** vdo has joined #openstack | 08:30 | |
Razique | a real successfull one :) | 08:30 |
troya | so, actually what's wrong with configuration, could you tell me ? | 08:30 |
Razique | ok | 08:30 |
*** ejat has quit IRC | 08:30 | |
Razique | the first thing was the fixed_ips rattached to a non existing network | 08:30 |
livemoon | Razique: Do you have blog? | 08:31 |
Razique | livemoon: no not really :p | 08:31 |
uvirtbot | New bug: #893899 in openstack-integration-tests "Append metadata on rebuild" [Undecided,New] https://launchpad.net/bugs/893899 | 08:31 |
Razique | that causes the ip picking by dnsmasq buggy | 08:31 |
Razique | the second thig is that u didn't have any network rattached to the project | 08:31 |
livemoon | oh | 08:31 |
Razique | third u had a /32 network | 08:31 |
livemoon | I want to learn more about openstack from you | 08:32 |
Razique | which means only one instance for that network, which is weird | 08:32 |
Razique | livemoon: hehe, just stay around | 08:32 |
Razique | I'm trying to learn also :p | 08:32 |
Razique | troya: so what we did : 1 remove weird / small networks | 08:32 |
Razique | 2- cleaned the orphans fixed_ip (used by dnsmas) | 08:32 |
Razique | 3- attached a network for the project | 08:33 |
Razique | and voila :) | 08:33 |
troya | yes,, we have removed 2 networks others | 08:33 |
*** sandywalsh_ has joined #openstack | 08:33 | |
*** sandywalsh has joined #openstack | 08:33 | |
troya | then what's you mean about >> third u had a /32 network | 08:33 |
Razique | troya: yah and the other one too | 08:33 |
troya | it's have 255.255.255.240 | 08:33 |
troya | Razique: is it not /28 ? | 08:34 |
Razique | troya: oh now sorry | 08:35 |
Razique | i was talking about the two /28 networks | 08:35 |
Razique | 10.0.0.16/28 | 08:35 |
Razique | and 10.0.0.32/28 | 08:35 |
*** Gollen has joined #openstack | 08:35 | |
Razique | in fact you could recreate them and affect them to another project if you want to :) | 08:36 |
Razique | livemoon: what's up ? | 08:36 |
uvirtbot | New bug: #893902 in openstack-integration-tests "Flavor for server does not meet minRam requirements on create" [Undecided,New] https://launchpad.net/bugs/893902 | 08:36 |
livemoon | I need to spread openstack in china | 08:36 |
troya | Razique: so if i want to create network, i must create one by one, is it do you mean? | 08:36 |
Razique | troya: yah | 08:36 |
livemoon | Razique: so I need learn more skill about openstack from world | 08:37 |
Razique | and don't forget to attach them to a tenant afterwards if u want them to be used | 08:37 |
Razique | livemoon: sure | 08:37 |
troya | livemoon: it's great idea :D | 08:37 |
Razique | livemoon: any particular point ? | 08:37 |
livemoon | Europeans And Americans | 08:37 |
Razique | livemoon: what about France ? :( | 08:37 |
Razique | even better Lyon directly | 08:38 |
Razique | my city hehe :p | 08:38 |
livemoon | Fance is in Europeans | 08:38 |
Razique | yah | 08:38 |
Razique | I read too quickly this morning | 08:38 |
livemoon | I think Europeans And Americans are more people who killed in openstack | 08:38 |
Razique | I must be on speed or Redbull by inadvertance | 08:38 |
livemoon | who skilled in openstack | 08:38 |
Razique | livemoon: I'm not that sure actually | 08:39 |
livemoon | So I want to learn from you | 08:39 |
Razique | you can find good people everywhere | 08:39 |
livemoon | :) | 08:39 |
*** po has joined #openstack | 08:40 | |
Razique | Einstein said "You get knowledge by experimenting, the rest is nothing but Information" | 08:40 |
*** GheAway is now known as GheRivero | 08:41 | |
troya | Razique: it's wise quote :) | 08:41 |
troya | livemoon: how about indonesia ? hehe | 08:42 |
Razique | hehe I find it very accurate yah :) | 08:42 |
livemoon | troya: I think you need use two eth to test nova | 08:43 |
livemoon | one eth maybe has some issues; | 08:43 |
troya | eh Razique, i have set " ENABLED=0 " on /etc/default/dnsmasq, but whenever i reboot my pc and run instances, usually there is error message like " ndnsmasq: failed to create listening socket for 10.0.0.1: Address already in use " | 08:43 |
*** sandywalsh has quit IRC | 08:44 | |
*** sandywalsh_ has quit IRC | 08:44 | |
troya | Razique: is it one of effect because of many networks which i created few times ago ? | 08:44 |
troya | livemoon: wait for money to get it :D | 08:44 |
*** woleium has quit IRC | 08:48 | |
troya | livemoon: now i will configure my openstack on my server on internet.i hope get client soon :D | 08:49 |
Razique | troya: i'm not so sure | 08:50 |
Razique | can u send me both nova.conf and dpkg -l | grep dns | 08:50 |
*** junaid1 has joined #openstack | 08:51 | |
troya | Razique: this is >> http://pastebin.com/cCWN0HM4 | 08:52 |
*** itz_ has joined #openstack | 08:53 | |
*** junaid has quit IRC | 08:53 | |
troya | how about it Razique? | 08:54 |
*** doude_ has quit IRC | 08:56 | |
*** doude has joined #openstack | 08:56 | |
Razique | troya: remove dnsmasq | 08:59 |
*** guaqua2`` has joined #openstack | 09:00 | |
*** guaqua2` has quit IRC | 09:00 | |
*** adjohn has quit IRC | 09:00 | |
*** stanchan has quit IRC | 09:00 | |
Razique | aptitude remove --purge dnsmasq && aptitude purge ~c | 09:00 |
troya | is it don't use openstack to give dhcp ? | 09:00 |
Razique | nope, it uses dnsmasq-base | 09:01 |
troya | but first on documentation i must install dnsmasq, right ? | 09:01 |
Razique | troya: I'd say no | 09:01 |
*** dcs has joined #openstack | 09:01 | |
Razique | nova-network depends on dnsmasq-base | 09:02 |
troya | ok, i'll remove it | 09:02 |
Razique | try if that break simply reinstall it | 09:02 |
Razique | but I'm quite sure u don't need it | 09:02 |
Razique | it seems to be the daemon mainly | 09:02 |
Razique | and it's trying to listen on the port 53 | 09:02 |
Razique | but since dnsmasq process are launched by nova-network, the port is already in use | 09:03 |
Razique | so remove it, and restart nova-network | 09:03 |
Razique | then ps axl | grep dns | 09:03 |
Razique | if you see your processes that means it's fixed :) | 09:03 |
* Razique crosses the finger hehe | 09:04 | |
*** nacx has joined #openstack | 09:04 | |
*** uksysadmin has joined #openstack | 09:08 | |
*** guaqua2`` has quit IRC | 09:09 | |
troya | yes, there is process | 09:10 |
Razique | ok | 09:10 |
Razique | so try to reboot and see if you have the issue | 09:11 |
livemoon | who can teach me english? | 09:11 |
troya | Razique: reboot my machine do you mean ? | 09:11 |
Razique | yah reboot the pc | 09:11 |
troya | livemoon: i think your english is good | 09:11 |
doude | Hi all, do you know what is the objective of the flag 'zone_capabilities' ? | 09:11 |
troya | Razique: ok, please wait | 09:12 |
*** guigui1 has joined #openstack | 09:12 | |
*** troya has quit IRC | 09:12 | |
*** debo-os has quit IRC | 09:14 | |
*** koolhead17 has quit IRC | 09:17 | |
*** dobber has joined #openstack | 09:17 | |
*** stewart has joined #openstack | 09:20 | |
uvirtbot | New bug: #893918 in nova "os api needs to return more info than just build" [Undecided,New] https://launchpad.net/bugs/893918 | 09:21 |
*** troya has joined #openstack | 09:21 | |
troya | Razique: you are right, there is no problem now | 09:22 |
troya | :D | 09:22 |
Razique | o/ | 09:22 |
troya | usually after i restart my computer, nova-network service alwas XXX | 09:22 |
*** itz_ has quit IRC | 09:22 | |
troya | then on log show that dnsmasq already in used | 09:23 |
troya | so i must kill dnsmasq first | 09:23 |
troya | then rstart service again | 09:23 |
*** supriya has joined #openstack | 09:23 | |
Razique | ok | 09:24 |
Razique | now so it's not the case anymore ? | 09:24 |
*** dirkx_ has joined #openstack | 09:25 | |
troya | Razique, when creating network for example i execute : nova-manage network create public 10.0.0.0/8 3 16 | 09:26 |
troya | what the meaning of number 3 and number 16 ? | 09:27 |
*** junaid1 has quit IRC | 09:27 | |
*** troya has quit IRC | 09:27 | |
*** troya has joined #openstack | 09:28 | |
*** supriya has quit IRC | 09:28 | |
Razique | first and last ip for that network | 09:28 |
*** Nithya_HP has quit IRC | 09:28 | |
*** Nithya_HP has joined #openstack | 09:29 | |
troya | so on >> 10.0.0.0/8 3 16 number 3 is first ip for instances and number 16 is last ip for instances ? | 09:29 |
troya | so if my instances have out out range of number of ip range, i must create network again ? or can modify it ? | 09:30 |
Razique | 3 will go to nova-network , for the gateway | 09:31 |
Razique | and the last one will be used for broadcast | 09:32 |
troya | then if my instances have out of range of ip range, i must create network again ? or can modify it ? | 09:33 |
*** supriya has joined #openstack | 09:33 | |
Razique | well you won't be able to allocate more ip withing that netwrok | 09:33 |
Razique | thus you can create a new network | 09:33 |
Razique | and attach it to the project | 09:34 |
*** janpychou has quit IRC | 09:34 | |
Razique | but that would mean a bad anticipation | 09:34 |
Razique | usually go for a /24 or /28 | 09:34 |
*** redconnection has joined #openstack | 09:35 | |
*** pixelbeat has joined #openstack | 09:36 | |
troya | prevent be better than threatmen | 09:37 |
Razique | need to reboot | 09:38 |
Razique | be back :p | 09:38 |
*** Razique has quit IRC | 09:39 | |
*** koolhead11 has joined #openstack | 09:39 | |
koolhead11 | hi all | 09:39 |
jasona | anyone using kiall's scripts. when you create a test instance, what is the actual login/passwd you can use for an instance ? | 09:40 |
*** mattstep has quit IRC | 09:41 | |
*** Gollen has quit IRC | 09:41 | |
koolhead11 | kodapa, how far you reached :) | 09:42 |
uksysadmin | jasona, without knowing any background, but instances are set up using keys, not user/pass | 09:45 |
koolhead11 | hola uksysadmin | 09:45 |
uksysadmin | hello koolhead11 | 09:47 |
koolhead11 | uksysadmin, how have you been? Seems like too busy :D | 09:48 |
uksysadmin | as always | 09:48 |
uksysadmin | in terms of openstack I've a test cluster on my desk (2 nodes plus laptop) and a dev environment in the datacentre | 09:49 |
uksysadmin | just at the stage of trying to use Orchestra to provision my OpenStack env | 09:49 |
koolhead11 | waoo. | 09:49 |
koolhead11 | i have eye on your blog page as usual :P | 09:49 |
uksysadmin | :) | 09:50 |
*** stewart has quit IRC | 09:50 | |
uksysadmin | I've got a nice set up of running cobbler as a VM under VirtualBox which pxe boots my physical test cluster on my desk | 09:51 |
uksysadmin | cobb ler/Orchestra | 09:51 |
*** perestrelka has quit IRC | 09:52 | |
*** debo-os has joined #openstack | 09:52 | |
*** redconnection has quit IRC | 09:53 | |
troya | hi koolhead | 09:53 |
koolhead11 | troya, hello sir | 09:53 |
*** perestrelka has joined #openstack | 09:53 | |
troya | koolhead:where i can get log of freenode channle openstack | 09:53 |
troya | i forget somethings | 09:54 |
koolhead11 | uksysadmin, i have only worked on cobbler, not tried orchestra yet | 09:54 |
jasona | uksysadmin: hmm | 09:54 |
*** javiF has joined #openstack | 09:54 | |
koolhead11 | troya, http://eavesdrop.openstack.org/irclogs/ | 09:54 |
koolhead11 | uksysadmin, are you using preseed file too | 09:54 |
koolhead11 | i am yet to get my custom partitioning part working via preseed :( | 09:55 |
jasona | uksysadmin: uksysadmin: wonder where the key generation happened then and how i can get an actual login to an instance | 09:55 |
*** foexle has joined #openstack | 09:55 | |
foexle | hiho | 09:55 |
livemoon | hi, I want to buy a vps out of china, is there any cheap vps? | 09:56 |
*** zatix has quit IRC | 09:56 | |
koolhead11 | foexle, hellos | 09:57 |
koolhead11 | hola livemoon | 09:57 |
livemoon | I will see it | 09:58 |
uksysadmin | jasona, you should've at some point ran the command something like: euca-add-keypair openstack > cloud/creds/openstack.pem | 09:58 |
foexle | hey koolhead11 :) | 09:59 |
uksysadmin | this adds the keypair named "openstack" to the environment and it put the output private key as that end path (change to suit) | 09:59 |
livemoon | koolhead11: what is hola? | 09:59 |
uksysadmin | make sure its readable only by you (chmod 0400 openstack.pem) | 09:59 |
koolhead11 | livemoon, http://en.wikipedia.org/wiki/Hola :D | 09:59 |
uksysadmin | then when you launch your instance you should specify it on your command line with the -k option | 09:59 |
foexle | livemoon: hola = hello :) | 09:59 |
*** rustam has joined #openstack | 09:59 | |
koolhead11 | :P | 10:00 |
foexle | ^^ | 10:00 |
uvirtbot | foexle: Error: "^" is not a valid command. | 10:00 |
uksysadmin | euca-run-instances -k openstack $emi -t m1.small | 10:00 |
*** mcclurmc_ has quit IRC | 10:02 | |
*** chadh has quit IRC | 10:02 | |
uksysadmin | and yeah - using preseed | 10:02 |
*** chadh has joined #openstack | 10:03 | |
uksysadmin | but if I see failed to run preseed command one more time I'm gonna scream | 10:03 |
*** stewart has joined #openstack | 10:03 | |
*** mcclurmc has joined #openstack | 10:04 | |
uksysadmin | I'm confident on this one - after the 17th reboot and pxeboot - I might have an extra OpenStack Compute Node | 10:05 |
*** yeming has quit IRC | 10:05 | |
uksysadmin | 17th time's a charm | 10:06 |
*** redconnection has joined #openstack | 10:06 | |
koolhead11 | uksysadmin, i spent a day inside my datacentre to get in some love with preseed. :) | 10:07 |
uksysadmin | koolhead11, you should get out mre | 10:09 |
uksysadmin | more | 10:09 |
uksysadmin | ;-) | 10:09 |
koolhead11 | uksysadmin, i can give u presseed which does everything excluding the manual partition part | 10:09 |
koolhead11 | like putting keys and all too | 10:10 |
koolhead11 | :P | 10:10 |
uksysadmin | awesome - that be v handy | 10:10 |
uksysadmin | guess in return I'll play with partitioning for you ;-) | 10:10 |
koolhead11 | deal | 10:10 |
livemoon | bye all | 10:10 |
uksysadmin | bye livemoon | 10:10 |
livemoon | I am go home | 10:11 |
*** livemoon has left #openstack | 10:11 | |
uvirtbot | New bug: #893932 in nova "Error in Glance plugin while uploading the VHD file after snapshot / nova image-create/ or nova backup" [Undecided,New] https://launchpad.net/bugs/893932 | 10:11 |
troya | by livemoon | 10:12 |
etarasov | how to add fixed ip to instance? | 10:13 |
*** kaigan_ has quit IRC | 10:14 | |
*** kaigan_ has joined #openstack | 10:15 | |
*** rods has joined #openstack | 10:18 | |
*** katkee has joined #openstack | 10:22 | |
*** superbobry has joined #openstack | 10:23 | |
*** debo-os has quit IRC | 10:23 | |
*** redconnection has quit IRC | 10:26 | |
*** shang has quit IRC | 10:27 | |
*** PotHix has joined #openstack | 10:32 | |
*** dirkx_ has quit IRC | 10:38 | |
*** dirkx_ has joined #openstack | 10:39 | |
*** shang has joined #openstack | 10:39 | |
*** dev_sa has joined #openstack | 10:41 | |
*** miclorb_ has joined #openstack | 10:47 | |
*** troya has quit IRC | 10:51 | |
*** dirkx_ has quit IRC | 10:51 | |
*** dirkx_ has joined #openstack | 10:51 | |
*** darraghb has joined #openstack | 10:55 | |
*** lionel has quit IRC | 10:55 | |
*** lionel has joined #openstack | 10:56 | |
*** popux has quit IRC | 10:58 | |
*** dnjaramba has quit IRC | 10:58 | |
*** Razique has joined #openstack | 10:59 | |
Razique | anyone here have been able to create a boot from volume ? | 11:00 |
Razique | a bootable volume | 11:00 |
Daviey | kodapa / koolhead11 / vidd-away / jdg / Kiall: Would you mind summarsing what happend last night regarding upgrade woes? I've not been able to follow. Thanks | 11:01 |
koolhead11 | Razique, sirr | 11:01 |
Razique | hey koolhead11 :) | 11:01 |
*** Razique has left #openstack | 11:02 | |
koolhead11 | Razique, where have you been last day man? | 11:02 |
*** Razique has joined #openstack | 11:02 | |
foexle | hey Razique :) | 11:06 |
Razique | hey foexle | 11:06 |
*** supriya has quit IRC | 11:06 | |
Razique | 'sup ? | 11:06 |
*** cereal_bars has joined #openstack | 11:07 | |
Razique | any boot from volume pioneer around ? :F | 11:08 |
Razique | :D | 11:08 |
foexle | no sry | 11:09 |
foexle | but if you have a little bit time today => https://answers.launchpad.net/swift/+question/179733 ;) | 11:09 |
*** miclorb_ has joined #openstack | 11:17 | |
*** chmouel has quit IRC | 11:20 | |
*** lzyeval has quit IRC | 11:21 | |
*** chmouel has joined #openstack | 11:21 | |
foexle | chmouel: heyho :) .... where i can find the keystone version ? ... --version puts a hex | 11:22 |
chmouel | how did you check it out ? | 11:22 |
chmouel | did you install from package/git? | 11:22 |
foexle | it was from Razique a stable version: openstack-keystone-85a49ae | 11:23 |
foexle | was a zip file cant find some git informations | 11:24 |
*** dirkx_ has quit IRC | 11:24 | |
*** ahasenack has joined #openstack | 11:24 | |
foexle | the last file change was 8. nov | 11:25 |
*** shaon has quit IRC | 11:27 | |
chmouel | humm so it's probably the old swift/keystone middleware | 11:28 |
foexle | so swift are verry new. i have answered the post | 11:32 |
foexle | swift version: 1.4.5~20111117.1632-0ubuntu0ppa1~lucid1 | 11:32 |
*** popux has joined #openstack | 11:33 | |
*** oubiwann has quit IRC | 11:33 | |
doude | Hi all, do you know what is the objective of the flag 'zone_capabilities' ? | 11:36 |
*** oubiwann has joined #openstack | 11:41 | |
*** primeministerp has quit IRC | 11:42 | |
*** catintheroof has joined #openstack | 11:47 | |
*** lorin1 has joined #openstack | 11:47 | |
*** catintheroof has quit IRC | 11:48 | |
*** livemoon has joined #openstack | 11:50 | |
*** dirkx_ has joined #openstack | 11:50 | |
*** troya has joined #openstack | 12:00 | |
*** troya has quit IRC | 12:02 | |
*** katkee_ has joined #openstack | 12:03 | |
*** katkee has quit IRC | 12:03 | |
*** wi3dzma has joined #openstack | 12:05 | |
*** primeministerp has joined #openstack | 12:07 | |
*** katkee_ has quit IRC | 12:10 | |
*** markvoelker has joined #openstack | 12:11 | |
Razique | Back all http://blogs.orange-business.com/live/assets_c/2011/11/Cloud%20Journey%20infographic%20from%20Orange-thumb-596x3656-7722.jpg | 12:21 |
*** etarasov has quit IRC | 12:27 | |
*** dnjaramba has joined #openstack | 12:27 | |
*** AntoniHP has quit IRC | 12:30 | |
*** lorin1 has quit IRC | 12:30 | |
koolhead11 | Razique, good one | 12:32 |
Razique | foexle: u around ? | 12:33 |
foexle | yeah | 12:33 |
*** troya has joined #openstack | 12:33 | |
Razique | koolhead11: yah interesting infos | 12:33 |
Razique | sorry I was eating | 12:33 |
Razique | let me have a look at ur issue | 12:33 |
foexle | :) | 12:33 |
*** mdupont has quit IRC | 12:34 | |
foexle | °° http_connect_raw() got an unexpected keyword argument 'timeout' | 12:36 |
*** abecc has joined #openstack | 12:38 | |
troya | Hi Razique, are you still here? | 12:39 |
*** supriya has joined #openstack | 12:40 | |
*** dcs has quit IRC | 12:40 | |
Razique | troya: yh | 12:40 |
*** AntoniHP has joined #openstack | 12:40 | |
Razique | working on openstack failures causes identification and action plan | 12:40 |
Razique | :p | 12:40 |
troya | Razique: you are right | 12:41 |
troya | :D | 12:41 |
troya | Razique: i have installed nova on real server justnow.Still have same issue with on my computer.Then i have following instruction like some times ago.Now it still failed to ping instnaces | 12:42 |
Razique | foexle: run SELECT tokens.id AS tokens_id, tokens.user_id AS tokens_user_id, tokens.tenant_id AS tokens_tenant_id, tokens.expires AS tokens_expires | 12:43 |
Razique | FROM tokens | 12:43 |
Razique | WHERE tokens.id = "30abaeac-b8dc-46f9-9aa1-db146c513099" | 12:43 |
Razique | foexle: It looks like the roles for swift are incorrect | 12:43 |
*** miclorb_ has quit IRC | 12:44 | |
*** praefect has joined #openstack | 12:44 | |
*** cloudgeek has joined #openstack | 12:45 | |
chmouel | yeah i'm fixing that | 12:45 |
chmouel | just testing my change first :) | 12:45 |
livemoon | troya: so bad | 12:45 |
Razique | chmouel: u work on swift project ? | 12:45 |
*** lorin1 has joined #openstack | 12:46 | |
*** lorin1 has quit IRC | 12:46 | |
praefect | hi Razique, ça va? | 12:46 |
Razique | hey praefect :) | 12:46 |
praefect | are you in Europe? | 12:47 |
Razique | nice nice :) | 12:47 |
Razique | yup France | 12:47 |
*** primeministerp has quit IRC | 12:47 | |
troya | livemoon: i dont know why. i have following instruction from Razique like on my computer which have running sometimes ago | 12:47 |
praefect | Ã Nice !! (just kidding) | 12:47 |
Razique | hehe what about you ? | 12:47 |
praefect | I'm from Québec city, I've read somewhere you were from France | 12:47 |
Razique | troya: don't put this on me :x | 12:47 |
troya | livemoon: but it's run on Qemu, and my computer run on KVM, is it effect ? | 12:47 |
Razique | hehe :p | 12:47 |
troya | Razique : hehehe | 12:48 |
Razique | troya: could be | 12:48 |
*** bsza has joined #openstack | 12:48 | |
Razique | what are the images you trying to spawn | 12:48 |
Razique | look for "unable to spawn, is virtualization enabled" into nova-compute.log | 12:48 |
chmouel | i work on openstack stuff yeah and i'm based in France | 12:48 |
troya | Razique:this is ubuntu 9.10. but running on qemu, my computer which running sometimes ago running on kvm.is it effect ? | 12:48 |
praefect | hi chmouel, I work on openstack too and I work for a french company... in Quebec city, canada | 12:49 |
livemoon | troya: how many eth do you have? | 12:49 |
praefect | I tried cloudstack yesterday... I feel like I cheated on openstack.. kind of... | 12:49 |
mattt | :) | 12:49 |
troya | livemoon: still one | 12:49 |
Razique | oh great chmouel | 12:49 |
Razique | your blog feeds are now into my inbox :) | 12:50 |
Razique | troya: try a tty image | 12:50 |
livemoon | whose blog? | 12:50 |
praefect | yeah what blog? where? | 12:50 |
troya | Razique: what is tty image? | 12:50 |
Razique | http://blog.chmouel.com/ | 12:51 |
*** hadrian has joined #openstack | 12:51 | |
troya | on my computer which runnning with your instruction is ubuntu 11.04 | 12:51 |
livemoon | chmouel: you are work in openstack? | 12:51 |
Razique | http://www.minimalinux.org/ttylinux/ | 12:51 |
Razique | troya: let me find you a clouded image | 12:51 |
praefect | troya: http://smoser.brickies.net/ubuntu/ttylinux-uec/ | 12:51 |
chmouel | yeah not much interesting on my blog but probably should make a lengthy post to explain how swift+keystone works | 12:51 |
coli | praefect:: what's your expiereince with cloudstack then, compared to openstack ? | 12:51 |
Razique | troya: there http://smoser.brickies.net/ubuntu/ttylinux-uec/ | 12:51 |
koolhead11 | chmouel, here you are | 12:52 |
troya | it's tarball, right ? | 12:52 |
koolhead11 | i was reading your blog :d | 12:52 |
praefect | well cloudstack is really really easy to setup, you guys would laugh, it is also more mature in terms of being all packaged together (auth, billback, dashboard), everything is working out of the box | 12:52 |
koolhead11 | wanted to know about how/what python-novaclient is | 12:52 |
*** osier has quit IRC | 12:52 | |
coli | chmouel: I would love that sort of post, as well as explantion of this note "Keystone currently allows any valid token to do anything with any account." in keystone documentation regarding swift | 12:53 |
chmouel | koolhead11: it's an os project for the client binding of nova (and keystone) | 12:53 |
praefect | cloudstack does not have something like FlatDHCP, you need to have VLANs if you want to assign floating ips... | 12:53 |
*** marrusl has joined #openstack | 12:53 | |
chmouel | coli: that's not true in trunk 8-) | 12:53 |
Razique | troya: simply download a targ | 12:53 |
Razique | z | 12:53 |
Razique | and use uec-publish-tarball | 12:53 |
chmouel | coli: and don't think with the old one as well | 12:53 |
coli | praefect: therefore cloudstack has a limitation to 4080 tenants (or there abouts) ? | 12:54 |
chmouel | coli: any user can do anything in an tenant/account that's true in old middleware but not the new one | 12:54 |
praefect | coli: you are right, it limits the number of tenants... it is explained in the doc | 12:54 |
troya | ok, i will try now. first of all i have try using tarball from >> http://uec-images.ubuntu.com/server/releases/natty/release/ubuntu-11.04-server-uec-amd64.tar.gz.but i cannot to login into | 12:54 |
koolhead11 | chmouel, i have keystone up and running. where and how i should install python-novaclient and get things going. i am yet not able to get dash working via keystone | 12:54 |
praefect | the limit is actually inferior to 4096 | 12:54 |
Razique | troya: yah you cannot run that image into a qemu system | 12:54 |
Razique | have you looked at nova-compute erorrs ? | 12:54 |
coli | chmouel: I have waited for this sort of explenation before enduring on a chalenge to get it tested, thank you. | 12:54 |
coli | praefect:: i know, I just allow few vlans for management and special use ;-) | 12:55 |
troya | Razique: nova-compute appear normal | 12:55 |
chmouel | koolhead11: well you probably want to do the same way we do on devstack | 12:55 |
troya | nova-network also like normall | 12:55 |
troya | please wait, i'm eating :D | 12:55 |
praefect | while we are on the subject, I never managed to get keystone working, I know there was a new documentation waiting for review, is this out? Where is the real document that will help me run keystone, I've tried twice to get it running I want to succeed next time | 12:55 |
chmouel | koolhead11: this should be pretty easy to read as this is a shell script | 12:55 |
coli | praefect: in your opinnion what's the main cons in cloudstack ? | 12:56 |
chmouel | coli: yep sorry about that it's a bit of a mess right know but hopefully this should get better soon when we finish to squash all those bugs | 12:56 |
Razique | praefect: where did you stop ? | 12:56 |
koolhead11 | chmouel, we are working on documentation and am more intersted to play with it and install use it from package :) | 12:56 |
chmouel | koolhead11: i know it's on pypi if that's help not sure about deb/rpm (although for deb you can easily use python-stdeb to generate package from pypi) | 12:57 |
coli | chmouel: do I undrestand correctly that in order to have things working as it should then I should use trunk version ? | 12:57 |
praefect | coli: IMHO, it is written in Java so less "hackable" for me, also it is more "magic" which means that all is well as long as it works, it something does not work you will have a hard time figuring it out, that's when you call citrix (who bought it) to the rescue I guess | 12:57 |
koolhead11 | chmouel, ubuntu/oneiric has a python-novaclient | 12:57 |
coli | praefect: I heard about cloudstack yestarday and it was news taht Citrix has acquired it just few months ago, Citrix is my client ;-) | 12:57 |
koolhead11 | coli, o.0 | 12:58 |
*** po has quit IRC | 12:58 | |
coli | praefect:: openstack main advantage for me ist the I'm able to read the code and change it when needed. | 12:58 |
cloudgeek | Razique: Livemoon:Hey my exam going on | 12:58 |
praefect | coli: I have a question for you, about VLANs, I'm in a work environment where all VLANS are taken, could I isolate myself wit a cisco switch that supports vlan or will this impact the whole network (playing with VLANs) | 12:58 |
chmouel | koolhead11: yeah i am not sure about distro specific stuff, hopefully when packagers will wakes-up they could tell you about it | 12:58 |
chmouel | hint there is the openstack-packaging irc chan | 12:58 |
chmouel | less busy than here | 12:58 |
praefect | coli: yeah, you are right, I modify it also, python is great for that | 12:58 |
koolhead11 | chmouel, awesome | 12:58 |
Razique | cloudgeek: been a while | 12:59 |
Razique | how it's going ? | 12:59 |
coli | praefect: you can have your own switch with own vlan id as long as the uplink is "mode access" and not "mode trunk" | 12:59 |
cloudgeek | Razique: I score well but final exam reaming only i got internal grades | 12:59 |
praefect | thanks coli that's what I wanted to know. | 12:59 |
*** dnjaramba has quit IRC | 12:59 | |
koolhead11 | Razique, i am in a cool bug currently vidd-away helped me but no luck yet | 12:59 |
Razique | koolhead11: enlighten me | 13:00 |
Razique | cloudgeek: congrats :) | 13:00 |
cloudgeek | Razique:cloud in my syllabus <Professor impress from openstack , might have good score | 13:00 |
livemoon | cloudgeek: I havenot seen you for a long time | 13:00 |
Razique | I hate exams | 13:00 |
Razique | make me feel like shit haha | 13:00 |
koolhead11 | Razique, me2 :P | 13:00 |
*** dnjaramba has joined #openstack | 13:00 | |
koolhead11 | cloudgeek, congrats :) | 13:00 |
livemoon | cloudgeek: are you a student? | 13:00 |
coli | praefect: how long did it take you to install cloudstack ? | 13:01 |
Razique | cloudgeek: yah cool | 13:01 |
praefect | so.. there isn't any good documentation on keystone to help me, I'm afraid to try and tackle it a third time | 13:01 |
cloudgeek | livemoon:Razique: m,y exams going on , i also hate exams i believe in practical things not in writing on paper | 13:01 |
cloudgeek | Thanks to all | 13:01 |
livemoon | cloudgeek: In china, there are a lot of exams | 13:01 |
Razique | praefect: what's the rock u're hitting right now about KS ? | 13:01 |
praefect | coli: 4 hours, the cloud is running, even running qcow2 images taken from the openstack cloud, it is that simple... I could redo it in 2hours | 13:02 |
livemoon | If you are in china, you will be crazy | 13:02 |
cloudgeek | livemoon: Here in India too , | 13:02 |
coli | praefect: how many nodes ? | 13:02 |
coli | I will try it out today then | 13:02 |
cloudgeek | livemoon: i write a blog on cloud computing security | 13:02 |
koolhead11 | btw cloudgeek where in india u study :) | 13:02 |
*** redconnection has joined #openstack | 13:02 | |
livemoon | cloudgeek: give us your blog | 13:02 |
praefect | Razique: I don't even know where to start, I guess last time I failed when I tried and move nova away from old auth... then everything stopped working I guess.. I kind of understand the concepts better now, but I need a doc to take me by the hand on this one | 13:03 |
praefect | sorry to say but it's trye | 13:03 |
praefect | true | 13:03 |
cloudgeek | Koolhead11:in MDU university in Rohtak (Govt uni) | 13:03 |
Razique | praefect: mm we have updated the doc a bit about Keystone | 13:03 |
Razique | and I've a pending schema | 13:03 |
praefect | coli: I've got 2 nodes only | 13:03 |
koolhead11 | cloudgeek, wa00. | 13:03 |
praefect | Razique: so my best bet is the identity starter guide? | 13:03 |
praefect | Razique: official doc? | 13:04 |
cloudgeek | livemoon: blog link is http://learnhackstuff.blogspot.com/ | 13:04 |
livemoon | cloudgeek: are you still a student? | 13:04 |
*** popux has quit IRC | 13:04 | |
Razique | praefect: have you read that ? http://docs.openstack.org/diablo/openstack-identity/admin/content/what-is.html | 13:04 |
cloudgeek | livemoon: I am in 2nd year | 13:04 |
foexle | chmouel: its anything wrong with parsing of the result set in swift-keystone middleware? | 13:04 |
cloudgeek | livemoon: Undergrad | 13:05 |
praefect | Razique: it's funny, all the links are dark blue because I've read EVERYTHING!!! =) I'll re-read it, I should work on keystone today and get it over with, I need another coffee | 13:05 |
* Razique remembr old school times | 13:05 | |
livemoon | cloudgeek: :) | 13:06 |
livemoon | I miss you very much | 13:06 |
Razique | praefect: hehe initiatives come to thems that wait | 13:06 |
Razique | praefect: which version of Keysone and nova/glance are u running ? | 13:06 |
cloudgeek | livemoon: i also miss you ,vidd also other all Friends too Razique also | 13:06 |
cloudgeek | I missed all openstack | 13:07 |
livemoon | :) | 13:07 |
praefect | Razique: that's probably the main problem, I'm on centos, scientific linux, RHEL, call it what you want... so I'm stuck with the packages from yum.griddynamics.net (they do a good job don't get me wrong) | 13:07 |
Razique | :) | 13:07 |
livemoon | now I will go to read blog about openstack | 13:07 |
*** dnjaramba has quit IRC | 13:07 | |
livemoon | see you tomorrow my friends | 13:07 |
foexle | bye livemoon | 13:07 |
Razique | praefect: oh ok | 13:07 |
Razique | bye livemoon | 13:07 |
praefect | ciao livemoon | 13:07 |
*** dnjaramba has joined #openstack | 13:07 | |
Razique | first things first | 13:08 |
Razique | is today KS installed ? :D | 13:08 |
praefect | Razique: yes, | 13:08 |
Razique | how is it configured ? | 13:08 |
praefect | keystone-manage 0.9 (keystone --version says something about a function at 0x7287342383 | 13:09 |
Razique | 0.9 | 13:09 |
Razique | mmm | 13:09 |
Razique | which version of diablo ? | 13:09 |
Razique | stable or trunk ? | 13:09 |
praefect | Razique: I don't know, it's probably stable, they built those packages after the release of Diablo, it says 2011.3-LOCALBRANCH:LOCALREVISION I don't know it that tells anything | 13:10 |
Razique | ok 2011.3 : stable :) | 13:10 |
Razique | praefect: I've a version of keystone that works with stable | 13:11 |
koolhead11 | praefect, i doubt keystone will have any issue if ur using from source | 13:11 |
Razique | maybe you would try with minue ? | 13:11 |
koolhead11 | i have keystone running and i can make request via curl to get data | 13:11 |
chmouel | foexle: i am not sure i understand | 13:11 |
chmouel | foexle: the schemas has changed a few times so the middleware had to track them between stable/diablo and trunk | 13:12 |
chmouel | foexle: if that's what you were asking | 13:12 |
koolhead11 | Razique, now when am passing username and password via dashboard, the django server crashes | 13:12 |
praefect | Razique: thansk for your help, let me check where I was regarding keystone I'm looking at the db, it looks like I have all my endpoints set | 13:12 |
Razique | praefect: what the error when u request Keystone ? | 13:12 |
praefect | with curl? what kind of command should I try (let me check) | 13:13 |
Razique | praefect: ok | 13:13 |
koolhead11 | praefect, yes try curl command | 13:13 |
foexle | chmouel: no and yes ;) .... if you see my question again at launchpad, i get now again http 412 .... have new version of swift and trunk of keystone | 13:13 |
koolhead11 | Razique, vidd-away suggested me to edit the dashboard with some random data | 13:13 |
koolhead11 | in one of the table | 13:13 |
*** aliguori has joined #openstack | 13:14 | |
*** dnjaramba has quit IRC | 13:14 | |
chmouel | foexle: I am not sure yet why you gett this 412, you really have nothing int he logs? | 13:14 |
*** dnjaramba has joined #openstack | 13:15 | |
*** javiF has quit IRC | 13:15 | |
*** aliguori has quit IRC | 13:15 | |
*** Teknix has quit IRC | 13:15 | |
foexle | chmouel: noo keystone returns 200 and swift gets 412 ... no backtraces | 13:16 |
foexle | chmouel: from syslog http://pastebin.com/T4cEiLJf | 13:17 |
praefect | Razique: ok when I try curl auth and username/password I get "The service could not be found for glance" which means I must have a problem with this endpoint only (maybe) | 13:18 |
Razique | yup :) looks like so | 13:19 |
Razique | you use mysql ? | 13:19 |
chmouel | foexle: not sure yet, let me fix a few stuff and give you to test that new middleware to see if you can reproduce with it | 13:19 |
praefect | Razique: no sqllite, I see 3 enpoints in enpoint_templates (nova, identity and glance), I'll double-verify the glance endpoint, give me 2 minutes | 13:20 |
*** GheRivero_ has joined #openstack | 13:21 | |
troya | All: i have published tty images, but still cannot ping it | 13:21 |
foexle | chmouel: ok thx | 13:21 |
troya | is use qemu is different with kvm ? | 13:21 |
*** lts has joined #openstack | 13:22 | |
*** livemoon has left #openstack | 13:24 | |
cloudgeek | bye for now need to prep for exam | 13:27 |
cloudgeek | bye all | 13:27 |
praefect | Razique: ok, I checked the glance endpoint, everything seems right but I still get 404 -service not found when I curl "auth" ----- curl -d '{"auth":{"passwordCredentials":{"username": "joeuser", "password": "secrete"}}}' -H "Content-type: application/json" http://localhost:35357/v2.0/tokens | 13:28 |
koolhead11 | praefect, tht looks correct :P | 13:28 |
DW-10297 | So when you're using nova to handle VMs for you does it store the VMs directly in a filesystem? | 13:29 |
praefect | it returns "itemNotFound" "message" the service could not be found for glance--404 | 13:29 |
Razique | praefect: i'll be back, need to finish a doc =D | 13:29 |
praefect | Razique: no problem | 13:29 |
troya | koolhead: are there different on running instances between using qemu and kvm on nova ? | 13:31 |
*** develtech has joined #openstack | 13:31 | |
smoser | Razique, praefect troya i suggest/request using http://launchpad.net/cirros rather than the ttylinux-uec. its kind of my 2.0 version of ttylinux-uec. | 13:33 |
smoser | (sorry if i sent that more than once) | 13:33 |
smoser | network issues here. | 13:33 |
Razique | smoser: oh yes I recall taht project:) | 13:33 |
Razique | thanks :) | 13:34 |
praefect | smoser: nice, you're kind of a legend man! smoser brickies I know the URL by heart and now I'm talking to the man behind the image!! =) haha | 13:34 |
smoser | yeah. i would have to suspect that 90% of traffic to brickies.net is for ttylinux-uec | 13:35 |
smoser | :) | 13:35 |
smoser | as there isn't relaly any other reason that you'd go there. | 13:35 |
praefect | I use ubuntu cloud images the latest one is big and slow to boot but really nice still boots faster than windows!) | 13:35 |
praefect | smoser: I browsed the other stuff there... | 13:35 |
praefect | come for the image stay for the other stuff(tm) | 13:36 |
smoser | outside of you customizing with user data, the images should boot to usability in probably 15 seconds. | 13:36 |
Razique | smoser: hehe | 13:36 |
smoser | well... except for bug https://bugs.launchpad.net/nova/+bug/851159 | 13:36 |
koolhead11 | smoser, hello sir :) | 13:36 |
troya | smoser: i run on machine with qemu, is it no problem ? | 13:36 |
smoser | well, nothing is going to be lightning fast with software virtualization. | 13:38 |
Razique | is it me ; or is Swift complicated to run ? | 13:38 |
Razique | harder than the other Openstack projects | 13:39 |
koolhead11 | Razique, i bett for others too :P | 13:39 |
Razique | ok ok | 13:41 |
troya | Razique: i have running tty image, but still unreachable.i check on nova-network.log nova-compute,l.log it's appear running normal | 13:43 |
Razique | same as this morning :p | 13:45 |
troya | yes, but it's on different machine. | 13:45 |
Razique | 1- is the network defined 2- is it attached to the projec 3- are the SSH and icmp rules added 4- check /Var/lib/nova/network/br100.conf | 13:45 |
Razique | 4- can I see the output of nova-compute when you spawn the instance ? :) | 13:46 |
troya | ok, i will check step by step again :ou are my teacher :) | 13:46 |
*** develtech has quit IRC | 13:46 | |
Razique | hehe ;) | 13:47 |
*** dirakx1 has joined #openstack | 13:51 | |
*** PiotrSikora has quit IRC | 13:54 | |
*** PiotrSikora has joined #openstack | 13:55 | |
troya | Razique: 1. i have defined network,2 and i have attache network to project 3, i have enable ssh and icmp >> i have pastebin sir : http://pastebin.com/waLQZWU2 | 13:58 |
*** dnjaramba has quit IRC | 13:58 | |
*** Nithya_HP has quit IRC | 13:58 | |
*** dnjaramba has joined #openstack | 13:59 | |
troya | Razique: please check this, may be there is some missing ? | 14:04 |
coli | PiotrSikora: Piotr z polboxu ? | 14:06 |
chmouel | foexle: I have just added some fixed to keystone middleware https://review.openstack.org/1867 would be cool if you can do your testing against it so I can work from there on the fixes | 14:06 |
chmouel | ^^^ same goes for all the swift+keystone testers | 14:07 |
uvirtbot | chmouel: Error: "^^" is not a valid command. | 14:07 |
Razique | another interesting blog : http://coderslike.us/ | 14:07 |
Razique | troya: can you show me the leases ? | 14:07 |
Razique | var/lib/nova/networks/br100.conf | 14:08 |
*** Razique_ has joined #openstack | 14:09 | |
foexle | chmouel: i'll try :D | 14:09 |
chmouel | now this is comming it's time for lunch for me i'm starving (3PM here) | 14:10 |
chmouel | if people can test and make sure this works | 14:10 |
*** mestery has left #openstack | 14:10 | |
chmouel | i'll spend some time to write some nice blog post explaining all of this | 14:10 |
*** deshantm_laptop has joined #openstack | 14:12 | |
troya | Razique: these like this >> http://pastebin.com/3GN3xQWm | 14:13 |
*** Razique has quit IRC | 14:13 | |
*** Razique_ is now known as Razique | 14:13 | |
koolhead11 | chmouel, please do that :D | 14:13 |
*** dirkx_ has quit IRC | 14:15 | |
*** wi3dzma has quit IRC | 14:15 | |
praefect | Coudl I get a paste of a glance-api.conf file that include valid keystone config? | 14:17 |
*** dnjaramba has quit IRC | 14:17 | |
foexle | chmouel: in swiftauth file you use keystone_url without admin port is this correct ? | 14:18 |
foexle | chmouel: only in your doc at the top | 14:19 |
foexle | keystone_url = http://keystone_url:5000/v2.0 | 14:19 |
*** pradeep1 has quit IRC | 14:19 | |
*** dnjaramba has joined #openstack | 14:19 | |
foexle | chmouel: and i get the same issue | 14:20 |
*** hugokuo has joined #openstack | 14:20 | |
coli | does dashboard support nova-volume with iscsi storage ? | 14:22 |
*** dnjaramba has quit IRC | 14:23 | |
*** dnjaramba has joined #openstack | 14:24 | |
*** dirakx1 has quit IRC | 14:40 | |
*** al has quit IRC | 14:41 | |
*** vikedios has joined #openstack | 14:42 | |
*** katkee has joined #openstack | 14:43 | |
chmouel | foexle: this shoudl actually be http://localhost:35357/v2.0 | 14:44 |
*** GheRivero_ has quit IRC | 14:44 | |
*** popux has joined #openstack | 14:45 | |
foexle | chmouel: yes ok .... i have the admin port configured ... but i get every time the same http code ... and i cant find any errors | 14:47 |
hugokuo | chmouel , would keystone2 support keystone Essx-m1 SSL feature ? thx | 14:48 |
hugokuo | chomouel , and aslo swift v2.0 ? | 14:49 |
*** dnjaramba has quit IRC | 14:54 | |
*** dnjaramba has joined #openstack | 14:55 | |
*** dgags has joined #openstack | 14:57 | |
foexle | chmouel: it's any way to find out where the problem is ? ... logging level of swift or someting like this | 14:58 |
*** chomping has left #openstack | 14:58 | |
*** jseutter has joined #openstack | 15:03 | |
*** cp16net has joined #openstack | 15:06 | |
*** cloudgeek has left #openstack | 15:06 | |
*** deshantm_laptop_ has joined #openstack | 15:06 | |
*** supriya has quit IRC | 15:09 | |
*** deshantm_laptop has quit IRC | 15:10 | |
uvirtbot | New bug: #894027 in glance "The wrong glance module is used when building docs" [Undecided,In progress] https://launchpad.net/bugs/894027 | 15:11 |
*** popux has quit IRC | 15:11 | |
*** pasik has quit IRC | 15:16 | |
*** MarkAtwood has quit IRC | 15:17 | |
*** guitarzan has quit IRC | 15:20 | |
*** pasik has joined #openstack | 15:20 | |
*** dnjaramba has quit IRC | 15:21 | |
*** praefect_ has joined #openstack | 15:22 | |
*** MarkAtwood has joined #openstack | 15:23 | |
*** vikedios has quit IRC | 15:23 | |
*** andrewbogott has joined #openstack | 15:24 | |
*** praefect has quit IRC | 15:24 | |
*** troya has quit IRC | 15:24 | |
*** 64MAA8CBP has joined #openstack | 15:24 | |
*** debo-os has joined #openstack | 15:27 | |
Glacee | is there a current redundancy method for keystone? Is it possible to have active/active db and/or multiple keystone services? | 15:27 |
*** mattray has joined #openstack | 15:31 | |
*** swill has joined #openstack | 15:31 | |
vidd-away | Glacee, i dont see much of a point to have keystone redundancy | 15:32 |
swill | chmouel: i see you did this today: https://review.openstack.org/#change,1867 | 15:32 |
swill | chmouel: the previous patch is still required right? | 15:32 |
swill | chmouel: or does this replace it? | 15:32 |
Glacee | vidd-away: really? keystone have to be up and running 100% if you want your authentication to work | 15:33 |
Glacee | seems pretty critical to me | 15:33 |
vidd-away | i can see having the database server being the same machine with the dashboard and keystone on it, and have that machine mirrored | 15:33 |
*** dendrobates is now known as dendro-afk | 15:34 | |
vidd-away | run it behind a load balancer | 15:34 |
Glacee | yeah but what about database? | 15:34 |
*** vidd-away is now known as vidd | 15:34 | |
vidd | ? | 15:35 |
Glacee | needs to have some kind of replication method if you have 2 dbs | 15:35 |
vidd | i just said that....database...keystone....horizon | 15:35 |
Glacee | database in cluster or something | 15:36 |
vidd | yes | 15:36 |
*** dolphm has joined #openstack | 15:36 | |
vidd | not something for the average user | 15:37 |
*** rsampaio has joined #openstack | 15:38 | |
foexle | Glacee: Master-Master replication but the second db are hotstandby and you use only one db server. if this server has a failure u can with arp request grap the ip of the fail host. | 15:38 |
foexle | Glacee: which method do you use for balancing ? http requests ? | 15:38 |
foexle | Glacee: mod_proxy ha-proxy or keepalived? | 15:39 |
Glacee | foexle: still on the planning phase | 15:39 |
vidd | i think OS is more of a "we need more/less infrastruction"-orientated...not "have 6 copies in case one fails"iented | 15:39 |
Glacee | yes.. I was pretty mich thinking about master-master replication as for the db | 15:39 |
foexle | Glacee: so i have keystone as a redundancy service | 15:39 |
chmouel | swill: yeah it goes on top of the previous patch (which is merged on github mirror) | 15:40 |
*** rnirmal has joined #openstack | 15:40 | |
foexle | Glacee: but i have some problems with kestone and swift :( | 15:40 |
swill | foexle: what are your problems? i have a working version now. | 15:40 |
chmouel | hugokuo: there is code in it (to connect via ssl) but i haven't tested it yet | 15:40 |
foexle | swill: i get every time ttp code 412 | 15:41 |
Glacee | foexle: we will propably use netscaler | 15:41 |
Glacee | but for now ha-proxy | 15:41 |
foexle | swill: keystone returns 200 but swift 412 | 15:41 |
notmyname | foexle: 412 from swift means that the request was malformed | 15:41 |
*** stewart has quit IRC | 15:41 | |
foexle | notmyname: yes i know but why ? .... i cant see no backtraces in logs and keystond says 200 | 15:42 |
swill | foexle: pastebin me your proxy conf, and your endpoint templates and the command you are using. | 15:42 |
foexle | swill: https://answers.launchpad.net/swift/+question/179733 | 15:42 |
notmyname | foexle: with curl or the swift cli? | 15:42 |
foexle | notmyname: curl | 15:42 |
notmyname | that link show a 401 from swift | 15:43 |
chmouel | swill: that configuration need probably to be adjusted | 15:43 |
chmouel | for the swif keystone middleware | 15:43 |
swill | foexle: put 'catch_errors' first in your proxy pipeline to get the errors in the logs. | 15:43 |
swill | chmouel: yep | 15:43 |
foexle | notmyname: look at the end | 15:43 |
foexle | swill: i have | 15:43 |
foexle | swill: no errors look my post at the end | 15:44 |
*** debo-os has quit IRC | 15:44 | |
*** cereal_bars has quit IRC | 15:44 | |
swill | yep. looking at it now... | 15:45 |
*** katkee_ has joined #openstack | 15:45 | |
*** katkee has quit IRC | 15:45 | |
chmouel | can you add http://pastie.org/private/od6kohq4bdcsc5akxzz6g in the [GENERAL] section | 15:46 |
foexle | chmouel: i have this :) | 15:46 |
*** javiF has joined #openstack | 15:47 | |
chmouel | can you paste your full proxy-server.conf ? | 15:48 |
swill | foexle: two of your identity endpoint templates have the admin port. that should only be the admin one. the other two should have the service port. | 15:48 |
foexle | yep | 15:48 |
*** mrjazzcat has joined #openstack | 15:49 | |
swill | foexle: my identity endpoint Template looks like this | 15:49 |
swill | keystone-manage endpointTemplates add RegionOne identity http://127.0.0.1:5000/v2.0 http://127.0.0.1:5001/v2.0 http://127.0.0.1:5000/v2.0 1 1 | 15:49 |
*** mrjazzcat has left #openstack | 15:50 | |
*** deshantm_laptop_ has quit IRC | 15:50 | |
*** deshantm_laptop has joined #openstack | 15:50 | |
foexle | chmouel: done | 15:50 |
*** troya has joined #openstack | 15:50 | |
uvirtbot | New bug: #894034 in openstack-ci "Milestone-proposed branches should be +2able by PTLs" [High,Confirmed] https://launchpad.net/bugs/894034 | 15:51 |
*** imsplitbit has joined #openstack | 15:51 | |
foexle | swill: i have exactly this endpoint | 15:51 |
chmouel | foexle: did you bind your keystone server to 5001? | 15:51 |
chmouel | foexle: which version of keystone you were saying you are using ? because 5001 is old | 15:51 |
*** debo-os has joined #openstack | 15:51 | |
foexle | chmouel: its the old default .... but i use in each conf this port | 15:51 |
chmouel | foexle: k | 15:52 |
chmouel | and you are sure that token is the admin token in keystone? | 15:52 |
foexle | yes keystone response are 200 with all informations | 15:52 |
swill | foexle: right now your internalUrl is using the admin port 5001, not the service port 5000 | 15:52 |
*** kbringard has joined #openstack | 15:53 | |
swill | according to the printout in lauchpad | 15:53 |
*** kbringard has quit IRC | 15:53 | |
foexle | swill: oh your right ! | 15:53 |
*** Ryan_Lane has joined #openstack | 15:53 | |
*** stewart has joined #openstack | 15:54 | |
chmouel | swill: nice catch | 15:54 |
hugokuo | chmouel , thx … I'll have a look at the code . and test it tar | 15:54 |
swill | chmouel: foexle trust me. i have combed my config for problems. hahaha | 15:54 |
hugokuo | s/tar/tmr/g | 15:54 |
*** vidd is now known as vidd-away | 15:54 | |
*** dev_sa has quit IRC | 15:54 | |
chmouel | swill: hey hey | 15:54 |
swill | chmouel: so this already has the first patch merged right? https://github.com/openstack/keystone/blob/master/keystone/middleware/swift_auth.py | 15:55 |
swill | chmouel: i recognize the changes. it is merged | 15:55 |
chmouel | swill: yeah that's correct | 15:55 |
*** kaigan_ has quit IRC | 15:56 | |
foexle | swill: no success :( | 15:56 |
chmouel | swill: the other review is on top of i | 15:56 |
*** nerdstein has joined #openstack | 15:56 | |
foexle | swill: same error | 15:56 |
swill | chmouel: cool, cause my deploy script pulls directly from there. i will do a clean install and review the new patch. | 15:56 |
swill | foexle: ok, do this for me | 15:56 |
chmouel | swill: cheers | 15:56 |
swill | foexle: you said you have catch_errors in your pipeline right? | 15:56 |
*** dirkx_ has joined #openstack | 15:57 | |
foexle | yeah i posted my complete conf to my post | 15:57 |
foexle | swill: pissibly wrong with my roles for this user ? | 15:58 |
*** rnorwood has quit IRC | 15:58 | |
foexle | swill: or something in the db | 15:59 |
swill | foexle: ya, i am looking | 15:59 |
swill | one sec | 15:59 |
*** debo-os_ has joined #openstack | 16:00 | |
swill | foexle: so your tenant and your password are both 'testing'? | 16:00 |
*** stanchan has joined #openstack | 16:01 | |
annegentle | koolhead11: around? Got a question for you. | 16:02 |
*** jsh_ has joined #openstack | 16:02 | |
koolhead11 | annegentle, shoot | 16:02 |
*** jaypipes has quit IRC | 16:02 | |
* koolhead11 is not in happy state though | 16:02 | |
annegentle | koolhead11: aw man sorry to hear that. | 16:02 |
*** 64MAA8CBP has quit IRC | 16:02 | |
* annegentle is in a grateful state | 16:02 | |
*** debo-os has quit IRC | 16:02 | |
*** vladimir3p has joined #openstack | 16:02 | |
hugokuo | ….. | 16:02 |
koolhead11 | hugokuo, hola | 16:03 |
*** debo-os_ is now known as debo-os | 16:03 | |
hugokuo | tired XD | 16:03 |
chmouel | thanksgiving is on tomo so everything will be okay | 16:03 |
annegentle | so someone pointed out to me that if you've got Ubuntu 11.04 and your repositories are up-to-date, you actually end up installing Diablo OpenStack... is that true? | 16:03 |
*** nyeates has joined #openstack | 16:03 | |
annegentle | I mean, it makes sense to me. | 16:03 |
annegentle | But I wanted to think about what that means for the guides... we really have an 11.04 guide and an 11.10 guide. Not really Cactus and Diablo guides (for the Starter guides) | 16:04 |
*** stanchan has quit IRC | 16:04 | |
annegentle | I think it's a good thing! All we have to do is point the 11.04 guide to the version of Dash that works without Keystone and poof Starter Guide with Dash? | 16:05 |
*** adjohn has joined #openstack | 16:05 | |
*** hallyn has joined #openstack | 16:05 | |
*** reidrac has quit IRC | 16:05 | |
*** jsh_ is now known as jsh | 16:05 | |
hugokuo | annegentle , for the starter …..agree with you | 16:05 |
*** TheOsprey has quit IRC | 16:06 | |
*** dolphm has quit IRC | 16:07 | |
*** debo-os has quit IRC | 16:07 | |
annegentle | though if there's a bug that prevents a smooth experience... then we wouldn't want to make that path an unhappy path. Need a happy path. | 16:07 |
*** troya_ has joined #openstack | 16:08 | |
hugokuo | for sure …. as my recent experience with keystone ….. not a happy path currently XD | 16:09 |
*** troya has quit IRC | 16:09 | |
*** andreas__ has joined #openstack | 16:09 | |
koolhead11 | hugokuo, am still stuck with dashboard | 16:09 |
koolhead11 | bloody dash dies once i put password and username | 16:10 |
hugokuo | koolheadll , ok .. I just finish one major task . Maybe I can have a integration test for you | 16:10 |
hugokuo | plz mail your entire system's conditions to me | 16:11 |
koolhead11 | k | 16:11 |
*** freeflyi1g has joined #openstack | 16:11 | |
*** jdg has quit IRC | 16:11 | |
*** ahasenack has quit IRC | 16:11 | |
hugokuo | also your purpose …. for example all from packages or source code …. versions of openstack related projects | 16:11 |
*** jdg has joined #openstack | 16:11 | |
*** popux has joined #openstack | 16:12 | |
*** freeflying has quit IRC | 16:14 | |
*** hugokuo has quit IRC | 16:14 | |
*** krow has joined #openstack | 16:17 | |
*** Razique has quit IRC | 16:18 | |
*** dolphm has joined #openstack | 16:20 | |
*** map_nw has joined #openstack | 16:24 | |
*** obino has quit IRC | 16:25 | |
*** mindpixel has quit IRC | 16:25 | |
*** Razique has joined #openstack | 16:26 | |
*** guitarzan has joined #openstack | 16:26 | |
*** Razique has joined #openstack | 16:26 | |
*** jdg has quit IRC | 16:30 | |
*** jdg has joined #openstack | 16:30 | |
Razique | back | 16:31 |
Razique | :D | 16:31 |
Razique | meeeting | 16:31 |
troya_ | oh..ok | 16:32 |
troya_ | what time in you razique ? in here is midnight :D | 16:33 |
*** MarkAtwood has quit IRC | 16:33 | |
Razique | 5:33 | 16:33 |
Razique | troya_: which country ? | 16:34 |
troya_ | i;m indonesian :) | 16:34 |
troya_ | south east asia :D | 16:34 |
Razique | ok i'll add you to my clocks :D | 16:34 |
Razique | any city for me ? :p | 16:35 |
troya_ | hahaha...you have said about it sometimes ago, now you saying it again :D | 16:35 |
troya_ | i'm on Jakarta | 16:35 |
*** dobber has quit IRC | 16:35 | |
Razique | but why aren't you on my list :( | 16:35 |
troya_ | oh..on mail i mean .. | 16:36 |
troya_ | anwar@troyans.net :) | 16:36 |
troya_ | it's my mail | 16:36 |
*** bonzay is now known as zz_bonzay | 16:36 | |
*** juddm has joined #openstack | 16:36 | |
Razique | I have it :p | 16:37 |
*** bryguy has quit IRC | 16:37 | |
Razique | ok so know Anwar ? | 16:37 |
Razique | you know | 16:38 |
troya_ | yes, i'm Anwar :D | 16:38 |
Razique | haha | 16:38 |
troya_ | hahaha | 16:38 |
troya_ | my i always use nick name troya eveywere on cyberworld | 16:38 |
troya_ | i like this name :D | 16:38 |
*** sandywalsh has joined #openstack | 16:38 | |
*** sandywalsh_ has joined #openstack | 16:39 | |
*** anotherjesse has joined #openstack | 16:39 | |
*** uksysadmin has quit IRC | 16:39 | |
troya_ | ok, can i continue with our discussion again sir ? hehe ( you are my teacher, so i must call you as sir) | 16:39 |
*** gamars has joined #openstack | 16:40 | |
gamars | sup all | 16:40 |
*** bryguy has joined #openstack | 16:40 | |
Razique | haha | 16:41 |
Razique | let me take a coffee | 16:41 |
Razique | but speak meanwhile:p | 16:41 |
gamars | joining the openstack (haven't sleep in a while hence). I'm running into a wall on something that must be stupid. Not an expert by any means on networking, I've setup 2 nodes (server1 is pretty much everything, and server2 a simple nova-compute). When trying to launch an image, it fails at the bridging. | 16:42 |
gamars | what I cannot understand is: I have no br100 in /etc/networking/interfaces. Do I have to set it up there by hand? | 16:43 |
gamars | (Ubuntu 11.10 Diablo release (I guess)) | 16:43 |
*** bryguy has quit IRC | 16:44 | |
*** clauden has quit IRC | 16:44 | |
troya_ | ok, from your step i have checked network list, then attach network to project, and about security rule to allowing port 22 and icmp.I have pastbin on >> http://pastebin.com/waLQZWU2 | 16:44 |
*** clauden_ has joined #openstack | 16:44 | |
troya_ | then /var/lib/nova/network/br100.conf >> http://pastebin.com/3GN3xQWm | 16:44 |
*** bryguy has joined #openstack | 16:45 | |
gamars | @troya_ interesting, in my case I don't have a /var/lib/nova/network/br100.conf file | 16:46 |
troya_ | gamars: from experince and some guidance, ip address of br100 is generated automatically.you don't need configure it on /etc/network/interfaces | 16:46 |
larissa | gamars: Error: "troya_" is not a valid command. | 16:46 |
gamars | troya_: ok, but then how do I get rid of this >> http://paste.openstack.org/show/3516/ | 16:47 |
troya_ | do you have inserted : --bridge_interfaces=br100 and --flat_network_bridge=br100 on nova.conf ? | 16:48 |
*** cereal_bars has joined #openstack | 16:48 | |
gamars | I guess the important part is "trying to add VLAN #100 to IF -:br100:- error: No such device" | 16:48 |
troya_ | what type of network method do use ? flat dhcp ? or vlan ? | 16:49 |
gamars | troya_: first time I hear about flat_network_bride :D >> http://paste.openstack.org/show/3517/ | 16:49 |
gamars | I'm doing as much as possible out of the book (default). So it must be a vlan configuration (needed because of MCAST between hosted image) | 16:50 |
Ryan_Lane | I'm getting a really weird stacktrace: http://pastebin.com/YTr4XJhi | 16:50 |
Ryan_Lane | I should mention that I'm using the facebook version of mysql and mysql client | 16:50 |
Ryan_Lane | which has binary as the default charset for the server and the client | 16:50 |
troya_ | gamars: honestly, i have not use vlan on my network, i'm sory | 16:51 |
gamars | troya_: ok, how would I go about switching it to the apparently more easy to use FlatDHCP? | 16:51 |
Ryan_Lane | but, we've worked around those issues on the server side by setting collation_server=utf8_unicode_ci, character_set_server=utf8, skip-character-set-client-handshake | 16:51 |
gamars | (I'll get some sleep and tackle that vlan later on) | 16:52 |
Ryan_Lane | could it be possible that bad data was injected into rabbit before that was set? | 16:52 |
*** CaptTofu1 has joined #openstack | 16:52 | |
*** nati2 has joined #openstack | 16:52 | |
troya_ | gamars: may be you can look at this >> http://pastebin.com/cpFutzdC | 16:54 |
gamars | troya_: thx. and in that case, it's just an edit of the nova.conf and nova-network restart will update the interfaces? | 16:55 |
coli | anybody working on Dashboard is here today ? | 16:55 |
*** afm has quit IRC | 16:57 | |
troya_ | it's usually need modifying database | 16:58 |
*** katkee_ has quit IRC | 16:58 | |
troya_ | if you haven't nessecary data on your database, be better that you drop your database and creat e again with "nova-manage db sync" after modify nova.conf | 16:59 |
gamars | troya_: ok that's cool. Another question, in that nova.conf you have no public eth. Do you have 2 nice? | 17:00 |
*** jaypipes has joined #openstack | 17:00 | |
gamars | erf nics | 17:00 |
*** MarkAtwood has joined #openstack | 17:00 | |
Razique | troya_: ok can I see the euca-get-console-output for the instance ? | 17:00 |
troya_ | gamars: no, in this case i only use 1 eth0 | 17:01 |
gamars | troya_: mmm ok. that's my main problem :( | 17:01 |
*** jeblair has quit IRC | 17:03 | |
troya_ | Razique: this is my output : http://pastebin.com/046WW60i | 17:04 |
troya_ | i look there are some unreachable | 17:04 |
*** woleium has joined #openstack | 17:05 | |
troya_ | Razique: this output is from tty images, do you want other instance ? | 17:05 |
*** dpippenger has quit IRC | 17:06 | |
*** TheOsprey has joined #openstack | 17:07 | |
*** jkyle has joined #openstack | 17:08 | |
*** jeblair has joined #openstack | 17:08 | |
*** KumarKR has joined #openstack | 17:09 | |
*** jseutter has quit IRC | 17:11 | |
*** maplebed has joined #openstack | 17:12 | |
*** clauden__ has joined #openstack | 17:12 | |
*** clauden_ has quit IRC | 17:12 | |
*** ccorrigan has joined #openstack | 17:13 | |
*** adjohn has quit IRC | 17:14 | |
foexle | chmouel: Unknown column 'endpoint_templates_1.version_id' in 'field list'") | 17:15 |
*** adiantum has joined #openstack | 17:15 | |
*** gregp76 has joined #openstack | 17:16 | |
chmouel | foexle: where do you see that | 17:16 |
koolhead11 | Razique, better be online for another few hour :P | 17:16 |
foexle | chmouel: keystone log | 17:16 |
Razique | koolhead11: hehe | 17:16 |
Razique | troya_: ok | 17:16 |
*** jdg has quit IRC | 17:16 | |
Razique | looks like the instance can't retrieve network | 17:16 |
foexle | chmouel: its the new version .... so which one you posted | 17:16 |
Razique | troya_: can I see nova-network.log ? | 17:17 |
chmouel | foexle: probably a wrong keystone initial data but i'm no keystone expert | 17:17 |
*** nati2_ has joined #openstack | 17:17 | |
smoser | troya_, it sure seems like your guests are not getting a dhcp address. | 17:18 |
*** nati2 has quit IRC | 17:18 | |
troya_ | Razique: it's piece of that nova-network.log | 17:21 |
troya_ | http://pastebin.com/YhN8CVXE | 17:21 |
foexle | chmouel: hmmm .... so i get this error since the daily version | 17:21 |
troya_ | it's too long, i can't pastebin all | 17:21 |
andyandy_ | is it possible in flatDHCP to assign for each project a different subnetwork? I mean in fixed range '10.x.x.x' assign '10.0.1.x' to a project etc. etc? | 17:21 |
chmouel | foexle: keystone daily version? | 17:22 |
foexle | chmouel: this one of your changes | 17:22 |
chmouel | yeah so that's not the swift keystone middleware but that's from keystone itself | 17:23 |
chmouel | you probably pulled keystone straight from trunk | 17:23 |
chmouel | and still have a old database | 17:23 |
chmouel | can you try from a new database? | 17:24 |
chmouel | I am not entirely sure what changed exactly with the old database and new | 17:24 |
chmouel | as I am not an expert in keystone itself | 17:24 |
troya_ | Razique, can we continuing it yesterday, it's about 12.24 am in here, i'm very sleepy @_@ | 17:25 |
troya_ | Razique: i'mean tomorrow | 17:25 |
troya_ | i have lose my control now, sleepy have mastering on my self :D | 17:26 |
*** supriya has joined #openstack | 17:26 | |
*** vdo has quit IRC | 17:27 | |
troya_ | bye all | 17:28 |
troya_ | see you later | 17:28 |
*** troya_ has quit IRC | 17:29 | |
*** rnirmal has quit IRC | 17:32 | |
*** stewart has quit IRC | 17:33 | |
*** dotdevops has joined #openstack | 17:35 | |
*** cp16net has quit IRC | 17:36 | |
*** mattstep has joined #openstack | 17:36 | |
andyandy_ | I am testing the flatDhcp, I am trying to figure out using the (iptables)/(nova security group) how to "isolate" different projects in a similar way a Vlan works, have you any idea? thank you | 17:37 |
*** cp16net has joined #openstack | 17:38 | |
*** py__ has quit IRC | 17:39 | |
*** py__ has joined #openstack | 17:40 | |
*** redconnection has quit IRC | 17:40 | |
*** popux has quit IRC | 17:40 | |
*** koolhead11 has quit IRC | 17:41 | |
*** gamars has quit IRC | 17:42 | |
*** rnirmal has joined #openstack | 17:44 | |
*** ohnoimdead has joined #openstack | 17:45 | |
*** stewart has joined #openstack | 17:45 | |
*** al has joined #openstack | 17:47 | |
*** ohnoimdead has quit IRC | 17:47 | |
*** ohnoimdead has joined #openstack | 17:48 | |
*** mattstep has quit IRC | 17:49 | |
*** rnorwood has joined #openstack | 17:50 | |
*** mattstep has joined #openstack | 17:52 | |
*** ohnoimdead_ has joined #openstack | 17:54 | |
*** ohnoimdead has quit IRC | 17:54 | |
*** ohnoimdead_ is now known as ohnoimdead | 17:54 | |
*** mattstep_ has joined #openstack | 17:54 | |
*** rnirmal has quit IRC | 17:54 | |
*** jdurgin has joined #openstack | 17:55 | |
*** PeteDaGuru has joined #openstack | 17:56 | |
*** mattstep has quit IRC | 17:56 | |
*** andrewbogott has quit IRC | 17:57 | |
*** rnirmal has joined #openstack | 17:57 | |
*** mattstep_ has quit IRC | 17:58 | |
*** mattstep has joined #openstack | 18:00 | |
*** rnorwood has quit IRC | 18:01 | |
*** rustam has quit IRC | 18:01 | |
*** dolphm has quit IRC | 18:04 | |
*** ohnoimdead has quit IRC | 18:04 | |
*** ohnoimdead_ has joined #openstack | 18:04 | |
*** mattstep_ has joined #openstack | 18:04 | |
*** rnorwood has joined #openstack | 18:05 | |
*** dolphm has joined #openstack | 18:05 | |
*** ohnoimdead_ has quit IRC | 18:05 | |
*** ohnoimdead has joined #openstack | 18:05 | |
*** rnorwood has quit IRC | 18:05 | |
*** nacx has quit IRC | 18:06 | |
*** mattstep has quit IRC | 18:06 | |
*** dragondm has quit IRC | 18:06 | |
*** Razique has quit IRC | 18:07 | |
*** mattstep_ has quit IRC | 18:08 | |
*** ohnoimdead has quit IRC | 18:08 | |
*** ohnoimdead has joined #openstack | 18:09 | |
*** mattstep has joined #openstack | 18:10 | |
*** dolphm has quit IRC | 18:10 | |
*** jdg has joined #openstack | 18:11 | |
*** afm has joined #openstack | 18:11 | |
*** ohnoimdead has quit IRC | 18:11 | |
*** ohnoimdead has joined #openstack | 18:11 | |
*** ohnoimdead_ has joined #openstack | 18:12 | |
*** ohnoimdead has quit IRC | 18:12 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:12 | |
*** stanchan has joined #openstack | 18:17 | |
*** afm has quit IRC | 18:17 | |
*** ohnoimdead has quit IRC | 18:18 | |
andyandy_ | I try to reprashe: does exists a way in flatDHCP network to avoid a user inside a VM to sniff the network traffic from a VM of a different project? | 18:20 |
Glacee | notmyname: I think you mentioned that swift keeps all the drives running at all time? Is it due to the auditors? | 18:20 |
*** afm has joined #openstack | 18:24 | |
*** cereal_bars has quit IRC | 18:24 | |
*** ohnoimdead has joined #openstack | 18:25 | |
*** ohnoimdead_ has joined #openstack | 18:27 | |
*** ohnoimdead has quit IRC | 18:27 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:27 | |
*** javiF has quit IRC | 18:27 | |
*** nati2 has joined #openstack | 18:27 | |
*** ohnoimdead has quit IRC | 18:27 | |
*** ohnoimdead has joined #openstack | 18:27 | |
*** floehmann has joined #openstack | 18:29 | |
*** darraghb has quit IRC | 18:29 | |
*** nati2_ has quit IRC | 18:30 | |
*** ohnoimdead_ has joined #openstack | 18:31 | |
*** ohnoimdead has quit IRC | 18:31 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:31 | |
*** obino has joined #openstack | 18:32 | |
*** cp16net has quit IRC | 18:33 | |
*** ohnoimdead_ has joined #openstack | 18:34 | |
*** ohnoimdead has quit IRC | 18:34 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:34 | |
*** ohnoimdead_ has joined #openstack | 18:35 | |
*** ohnoimdead has quit IRC | 18:35 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:35 | |
*** ohnoimdead_ has joined #openstack | 18:36 | |
*** ohnoimdead has quit IRC | 18:36 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:36 | |
*** rnorwood has joined #openstack | 18:36 | |
*** dirkx_ has quit IRC | 18:36 | |
*** imsplitbit has quit IRC | 18:37 | |
*** dpippenger has joined #openstack | 18:37 | |
*** ohnoimdead_ has joined #openstack | 18:38 | |
*** ohnoimdead has quit IRC | 18:38 | |
*** ohnoimdead_ has quit IRC | 18:38 | |
*** ohnoimdead has joined #openstack | 18:38 | |
*** adjohn has joined #openstack | 18:39 | |
*** ohnoimdead_ has joined #openstack | 18:40 | |
*** ohnoimdead has quit IRC | 18:40 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:40 | |
*** llang629 has joined #openstack | 18:43 | |
*** nati2 has quit IRC | 18:45 | |
*** ohnoimdead has quit IRC | 18:46 | |
*** ohnoimdead_ has joined #openstack | 18:46 | |
*** gregp76 has left #openstack | 18:47 | |
*** ohnoimdead has joined #openstack | 18:48 | |
*** ohnoimdead_ has quit IRC | 18:48 | |
*** markvoelker has quit IRC | 18:50 | |
*** ohnoimdead_ has joined #openstack | 18:50 | |
*** ohnoimdead has quit IRC | 18:50 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:50 | |
*** ohnoimdead_ has joined #openstack | 18:52 | |
*** ohnoimdead has quit IRC | 18:52 | |
*** ohnoimdead_ is now known as ohnoimdead | 18:52 | |
*** ohnoimdead has quit IRC | 18:52 | |
*** koolhead17 has joined #openstack | 18:52 | |
*** redconnection has joined #openstack | 18:54 | |
*** Ryan_Lane has quit IRC | 18:57 | |
*** jkyle has quit IRC | 18:57 | |
*** rustam has joined #openstack | 18:58 | |
*** praefect_ has quit IRC | 18:59 | |
*** praefect has joined #openstack | 18:59 | |
*** nati2 has joined #openstack | 19:00 | |
*** Vipul_ has joined #openstack | 19:00 | |
*** Ryan_Lane has joined #openstack | 19:01 | |
*** sandywalsh_ has quit IRC | 19:02 | |
*** dachary has quit IRC | 19:10 | |
*** KumarKR has quit IRC | 19:17 | |
*** debo-os has joined #openstack | 19:17 | |
*** redconnection has quit IRC | 19:19 | |
*** peteroplus has joined #openstack | 19:23 | |
*** supriya has quit IRC | 19:25 | |
peteroplus | Hi all, anybody knows how to get READ permissions on other container than the user owns on swift , i see that keystone is forming the URL with AUTH_%tenant_id% but when i try to retrieve something from another container from another user | 19:25 |
bsza | Glacee, yes drive are busy due to replicators and auditors | 19:26 |
peteroplus | i set the ACL , but it seems that the way that keystone forms the url doest allow me to use cross user / cross container permissions | 19:26 |
swill | peteroplus: which auth method are you using? | 19:26 |
peteroplus | anybody know if this is possible | 19:26 |
peteroplus | swill keystone2 | 19:26 |
peteroplus | swiftauth | 19:27 |
peteroplus | not tokenauth | 19:27 |
*** negronjl_mobile has joined #openstack | 19:27 | |
swill | peteroplus: ok, i am using swiftauth as well. i am assuming you have already fixed the bug with the 'timeout' in swift_auth.py file? | 19:27 |
*** dolphm has joined #openstack | 19:28 | |
*** RobertLaptop has joined #openstack | 19:28 | |
*** catintheroof has joined #openstack | 19:28 | |
peteroplus | swill i dont have timeout problems, it seems like the ACLs are ignored | 19:28 |
*** rnorwood has quit IRC | 19:28 | |
peteroplus | can the swift_auth.py fix affec this ? | 19:28 |
*** redconnection has joined #openstack | 19:28 | |
swill | peteroplus: so you can give acl access to a container to other users in the same tenant | 19:28 |
*** negronjl_mobile is now known as negronjl | 19:28 | |
swill | peteroplus: it will give you a 500 error if it fails with that bug. | 19:29 |
peteroplus | swill its possible to define acl for users that are in different tenants ? thats actually my question | 19:29 |
swill | peteroplus: you are using a url similar to this: | 19:29 |
*** ccorrigan has quit IRC | 19:30 | |
swill | swift -A http://127.0.0.1:5000/v1.0 -U username -K password post -r User -w User container | 19:30 |
swill | where User is the role of the user you want to give access to. | 19:30 |
peteroplus | thats right swill | 19:31 |
swill | in different tenants. No as far as i can tell | 19:31 |
*** rnorwood has joined #openstack | 19:31 | |
peteroplus | ok swill so thats my issue, its not working cause they are on different tenants | 19:31 |
*** andrewbogott has joined #openstack | 19:31 | |
swill | peteroplus: when i did testing across tenants i was not able to get it to work. | 19:31 |
*** dailylinux has joined #openstack | 19:31 | |
peteroplus | so if i want to integrate this with nova, so projects are mapped to tenants i cant get it work cross tenant | 19:32 |
swill | peteroplus: i believe that is 'by design' | 19:32 |
peteroplus | thats a big fail | 19:32 |
dailylinux | Hi | 19:32 |
dailylinux | what does Cloudstack have that Openstack doesn't? | 19:32 |
peteroplus | a project o nova that is mapped to a tenant and mapped so to a account in swift will not be able to read a file from another user/project/tenant | 19:32 |
swill | peteroplus: hmmm... i have not worked with nova at all. | 19:32 |
*** negronjl has quit IRC | 19:33 | |
peteroplus | ok swill, but that anwser my question so far | 19:33 |
*** negronjl has joined #openstack | 19:33 | |
peteroplus | i being fighting hours with this :) | 19:33 |
swill | peteroplus: my answer is specific to swiftauth. it may be possible with a different auth method. but i really dont know that. | 19:33 |
*** dolphm has quit IRC | 19:33 | |
*** andrewbogott has quit IRC | 19:34 | |
*** dolphm has joined #openstack | 19:34 | |
*** Razique has joined #openstack | 19:34 | |
Spirilis | hmm, so I just got devstack up and running, but I'm noticing noVNC giving me problems. Says WebSock error: [object Event] ... this is with firefox 8.0.1... | 19:35 |
peteroplus | ok i seems that is possible with tokenauth cause any token can do anything to any container | 19:35 |
swill | peteroplus: this patch is in the pipe. let me see if that helps us https://review.openstack.org/#change,1867 | 19:35 |
peteroplus | but thats actually a big security issue swill | 19:35 |
Spirilis | let me see if safari works | 19:35 |
swill | peteroplus: exactly, which is why i switched to swiftauth | 19:35 |
Spirilis | ahh k safari worked | 19:36 |
*** slop has quit IRC | 19:36 | |
peteroplus | swill ok that patch its in review by now, maybe when its approved can work | 19:37 |
peteroplus | i already have the admin role options | 19:37 |
peteroplus | and its not working | 19:37 |
swill | peteroplus: i dont think it will change this functionality, looking at it. | 19:37 |
swill | peteroplus: i have it working per tenant. i can give read and write permissions based on role | 19:38 |
peteroplus | swill perfect thats the catch | 19:38 |
swill | peteroplus: looking at the swiftauth file this shows that it is NOT possible | 19:39 |
swill | if account != '%s_%s' % (self.reseller_prefix, tenant[0]): | 19:39 |
swill | that is a check to make sure that you can not do cross tenant calls | 19:39 |
vidd-away | Kiall, are you here? | 19:39 |
*** vidd-away is now known as vidd | 19:40 | |
swill | peteroplus: so right now, i think the answer is that it is not possible to have users in different tenants access the same container. | 19:40 |
swill | using swiftauth | 19:41 |
peteroplus | swill excelent crystal clear pal | 19:41 |
vidd | Kiall, you here? | 19:44 |
*** juddm has left #openstack | 19:44 | |
*** dailylinux has quit IRC | 19:45 | |
*** jkyle has joined #openstack | 19:45 | |
peteroplus | swill thanks for all your help | 19:45 |
swill | peteroplus: no problem. sorry i did not have better news. ;) | 19:45 |
peteroplus | swill haha, no big issue i prefer to know if that was possible to not spend more time on it | 19:46 |
*** dachary has joined #openstack | 19:47 | |
*** JesperA has joined #openstack | 19:51 | |
*** jkerby has joined #openstack | 19:52 | |
*** hingo has joined #openstack | 19:52 | |
*** dnjaramba has joined #openstack | 19:54 | |
*** rnorwood has quit IRC | 20:00 | |
jkerby | Is anyone familiar with HybridFox connectivity to a single node stackops install of openstack. I am having authentication issues. From the docs it seems straight forward on setting up the credentials, however, I am not connecting and getting Authentication Failure: Signature does not match in the nova-api.log. | 20:00 |
Ryan_Lane | there's a change in diablo's defaults from cactus's regarding hostnames | 20:02 |
*** rnorwood has joined #openstack | 20:02 | |
Ryan_Lane | my instances are now getting their hostname in dhcp based on the instance's name, rather than the instanceid... | 20:02 |
Ryan_Lane | how do I change that back> | 20:02 |
*** cereal_bars has joined #openstack | 20:03 | |
uvirtbot | New bug: #894128 in openstack-ci "sync script not removing people from groups" [Undecided,New] https://launchpad.net/bugs/894128 | 20:03 |
vidd | Kiall, you there? | 20:06 |
Ryan_Lane | is there an option to change that back? | 20:07 |
coli | vidd: hi :-) | 20:10 |
vidd | hello coli | 20:10 |
*** afm has quit IRC | 20:10 | |
vidd | coli, have you seen Kiall lately? | 20:11 |
vidd | it appears his keystone package is broken | 20:11 |
coli | vidd: not since this morning (its 9pm in here now) | 20:11 |
swill | vidd: which one is his? | 20:11 |
vidd | managedit | 20:11 |
coli | vidd: we have installed his keystone, glance, dashboard and so far it works | 20:12 |
coli | vidd: installing compute node right now (first reading scripts then executing) | 20:12 |
*** mjfork has joined #openstack | 20:12 | |
*** nerdstein has left #openstack | 20:13 | |
coli | vidd: are you familiar with his scrips ? (kiall) | 20:15 |
vidd | yes | 20:15 |
vidd | alot of mine are based on his | 20:15 |
coli | he is got two interfaces eth0 - for floating ip (public) and eth1 for fixed ip (private ip range) | 20:17 |
vidd | right | 20:17 |
*** dolphm has quit IRC | 20:17 | |
Ryan_Lane | vishy, soren, any other nova dev that may know: my display name is now being set as the hostname in the dhcp file, rather than the instance-id. this is a change from cactus to diablo, and it's causing all my new instance creations to fail (since puppet assumes the names will be unique), any idea how to change this back? | 20:17 |
*** camm has quit IRC | 20:17 | |
coli | do i undrestand correctly that I need a single vlan which is going to connect all eth0 interfaces on all compute nodes and another one which is going to connect eth1 ? | 20:18 |
vishy | Ryan_Lane: "Display Name"? | 20:18 |
vishy | in the db? | 20:18 |
Ryan_Lane | yes | 20:18 |
vishy | are you booting using ec2_api? | 20:19 |
Ryan_Lane | yep | 20:19 |
vishy | on stable/diablo? | 20:19 |
Ryan_Lane | right. from the ppa | 20:19 |
praefect | Do I need kernel support for NBD to be able to properly inject key data? | 20:19 |
vidd | eth0 is managed by your internal router | 20:19 |
Ryan_Lane | I'm setting DisplayName via ec2 api | 20:20 |
Ryan_Lane | that's now ending up in the dhcp file, rather than the instanceid | 20:20 |
vishy | Ryan_Lane: stable/diablo isn't in a ppa i think | 20:20 |
praefect | I manage to inject ssh keys properly but the injection fails later on (child died uenxpectedly) and it deadlocks with stale lockfiles because libvirt crashes after the injection | 20:20 |
vidd | a vlan is set up upon the eth1 to connect all your compute nodes | 20:20 |
* Ryan_Lane twitches | 20:20 | |
praefect | I'm on centos | 20:20 |
Ryan_Lane | where am I supposed to be installing nova from, now? | 20:20 |
*** rnorwood has quit IRC | 20:20 | |
*** rnorwood has joined #openstack | 20:21 | |
vishy | Ryan_Lane: packaged version your best bet is probably ubuntu | 20:21 |
praefect | under centos: all injections logs an error (was not like that in cactus) | 20:21 |
vishy | Ryan_Lane: but you need to change generate_default_hostname | 20:21 |
Ryan_Lane | I'm running lucid | 20:21 |
Ryan_Lane | where's that at? | 20:21 |
vishy | in compute/api.py | 20:21 |
Ryan_Lane | ah | 20:21 |
*** PotHix has quit IRC | 20:22 | |
swill | i have built a bash scrpit that installs and runs a chef-solo recipe to install swift and keystone (swiftauth) for testing. should i clean it up and make it public? would people be interested in it? | 20:22 |
Ryan_Lane | heh. well, that's an unwelcome change ;) | 20:22 |
Ryan_Lane | it should at least be configurable. I'll need to open a bug for that | 20:23 |
vidd | swill, anything that makes installing and configuring stuff easier would be useful to ppl =] | 20:23 |
*** catintheroof has quit IRC | 20:24 | |
*** peteroplus has quit IRC | 20:24 | |
swill | vidd: this is really for testing at this point and is just for setting up a local environment. i dont want people to think it is a magic bullet. | 20:24 |
vishy | Ryan_Lane: in essex it is settable | 20:24 |
Ryan_Lane | ah | 20:24 |
Ryan_Lane | ok | 20:24 |
jdg | vidd: Ran again on a fresh OS image and just used the defaults, had the same problem. :( | 20:24 |
vishy | Ryan_Lane: but it defaults to display_name | 20:25 |
vidd | swill there are no magic bullets =] | 20:25 |
Ryan_Lane | I keep running into all kinds of fun issues :) | 20:25 |
swill | vidd: haha, ok... :) | 20:25 |
vishy | Ryan_Lane: lucid sounds painful | 20:25 |
Ryan_Lane | lucid isn't necessarily the issue | 20:25 |
Ryan_Lane | https://launchpad.net/~openstack-release/+archive/2011.3 <- | 20:25 |
Ryan_Lane | there's a ppa for diablo that has lucid packages | 20:26 |
vishy | Ryan_Lane: yeah but have you seen how many bugfixes have been backported into stable/diablo? | 20:26 |
Ryan_Lane | -_- | 20:26 |
*** nati2 has quit IRC | 20:26 | |
vishy | (post diablo release) | 20:26 |
*** nati2 has joined #openstack | 20:26 | |
* Ryan_Lane groans | 20:26 | |
*** adjohn has quit IRC | 20:26 | |
Ryan_Lane | it's really annoying to not have fixed backported to the LTS | 20:27 |
Ryan_Lane | *fixes | 20:27 |
vishy | Ryan_Lane: yeah, we decided to try to stop maintaining packages and let downstream do it | 20:27 |
*** coli has quit IRC | 20:27 | |
*** adjohn has joined #openstack | 20:27 | |
vishy | but i don't think they are supporting openstack in lucid | 20:28 |
Ryan_Lane | it's an infrastructural headache to support distro release of the day | 20:28 |
Ryan_Lane | so, basically, I'm stuck, now, with whatever I get from the distro> | 20:28 |
Ryan_Lane | so, once the new LTS is released, I'll be stuck on essex for four years? | 20:28 |
Ryan_Lane | (or are server LTS 2 years? :) ) | 20:29 |
vidd | 2 years | 20:29 |
Ryan_Lane | that's kind of shitty | 20:29 |
*** camm has joined #openstack | 20:30 | |
*** anotherjesse has quit IRC | 20:30 | |
Ryan_Lane | so, now I either have to support oneric, which is an unstable version, or I need to have a broken nova build? | 20:31 |
*** stevegjacobs_ has quit IRC | 20:32 | |
vishy | Ryan_Lane: I don't know what the update plan is for Precise | 20:32 |
vidd | oneric is not unstable =\ | 20:32 |
Ryan_Lane | vidd if it isn't a LTS, it's unstable | 20:32 |
Ryan_Lane | vishy: is there any easy property I can use to format the hostname back to what it previously was? | 20:33 |
*** coli has joined #openstack | 20:34 | |
vishy | it was ec2_id before? | 20:34 |
Ryan_Lane | yes | 20:34 |
Ryan_Lane | id_to_ec2_id in ec2utils? | 20:34 |
vishy | ec2utils. … yeah | 20:35 |
vishy | or just copy it directly | 20:35 |
*** rnirmal has quit IRC | 20:35 | |
Ryan_Lane | ok. cool. thanks | 20:35 |
vishy | 'i-%08x' % int(instance_id) | 20:35 |
* Ryan_Lane nods | 20:35 | |
*** debo-os has quit IRC | 20:37 | |
*** cmasseraf has joined #openstack | 20:38 | |
*** cp16net has joined #openstack | 20:40 | |
*** adiantum has quit IRC | 20:40 | |
*** cereal_bars has quit IRC | 20:41 | |
*** debo-os has joined #openstack | 20:42 | |
*** mgoldmann has quit IRC | 20:45 | |
*** andrewbogott has joined #openstack | 20:46 | |
*** RobertLaptop has left #openstack | 20:47 | |
*** sandywalsh has quit IRC | 20:48 | |
dweimer | I'm just starting to look into nova and am looking for an install method that can get a single node or small cluster up and running quickly. Something to play with over the long weekend so that I can start learning how it works. Is devstack stable or should I use something like UEC? | 20:51 |
vidd | dweimer, check out Kiall 's ppa and install script https://github.com/managedit/openstack-setup | 20:52 |
j^2 | hey all | 20:54 |
dweimer | vidd: Perfect, thanks. | 20:56 |
*** negronjl has quit IRC | 21:01 | |
*** negronjl has joined #openstack | 21:01 | |
*** debo-os has quit IRC | 21:03 | |
*** negronjl has quit IRC | 21:04 | |
*** negronjl has joined #openstack | 21:05 | |
*** PeteDaGuru has quit IRC | 21:08 | |
*** jdg has quit IRC | 21:14 | |
coli | vidd: I think I found few things which should be fxied in kiall's scripts ;-) | 21:16 |
coli | hopefuly he will be here today | 21:16 |
*** statik has quit IRC | 21:16 | |
andrewbogott | I'm about to push to a (new) nova branch on launchpad, and it's telling me to use bzr. This puzzles me since I got the nova source from git. Can anyone explain the relationship to me? Or is there a page of explanation someplace? | 21:19 |
coli | vlan100 is created only by nova-network ? | 21:23 |
vidd | coli, what have you found? | 21:23 |
vidd | yes...the networking is created by nova-network | 21:24 |
coli | vidd: so on management node where I don't have nova-network I need to create vlan100@eth1 (vlan_id: 100) by hand and attach it to virbr0 ? | 21:25 |
coli | vidd: his scripts don't check if eth1 is up or not, assuimg that the interface is up (our's wasn't) | 21:25 |
vidd | what services do you have on the management node? | 21:25 |
*** jdg has joined #openstack | 21:26 | |
coli | /usr/bin/nova-objectstore /usr/bin/nova-api /usr/bin/nova-scheduler /usr/bin/nova-vncproxy /usr/bin/nova-ajax-console-proxy | 21:27 |
vidd | coli, you will need nova-network on that node | 21:27 |
*** juddm has joined #openstack | 21:27 | |
vidd | unless you are going with an alternative networking alternative | 21:27 |
vidd | the vncproxy alone will require access to the nova-network | 21:28 |
Ryan_Lane | andrewbogott: did you read this? http://wiki.openstack.org/GerritWorkflow | 21:28 |
andrewbogott | Yes, but I will read it again :) | 21:29 |
coli | vidd: do you think kiall missed that in his scripts ? or did he assume that everyone will run nova-compute on the management node ? ;-) | 21:29 |
*** Ryan_Lane is now known as Ryan_Lane|away | 21:29 | |
vidd | you should be able to get away without the compute on the management node...the networking you need | 21:30 |
*** sandywalsh has joined #openstack | 21:30 | |
vidd | also, im sure his script are not geared to production-grade installs | 21:31 |
coli | justt checked, his scripts assume that nova-controler doesn't need nova-network, only nova-compute.sh will install nova-network | 21:32 |
*** miclorb_ has joined #openstack | 21:32 | |
*** bengrue has joined #openstack | 21:32 | |
vidd | strange | 21:32 |
coli | i'm still very glad that yourself and kiall have prepared these scripts for us all to learn from. | 21:32 |
vidd | my scripts work on the assumption that everything is set up on the controlloer | 21:33 |
vidd | and that you manually install nova compute on other machines and manually copy the config file | 21:34 |
coli | in his coping is manual as well, just he has a script called nova-controller.sh and nova-compute.sh to run on managment node or on compute nodes. | 21:35 |
coli | i'm just wondering why would you need nova-network to work wit nova-vncproxy | 21:36 |
*** ahale- is now known as ahale | 21:37 | |
vidd | to path the vnc image from the compute node to the vnc proxy | 21:37 |
*** Razique has quit IRC | 21:37 | |
coli | what do you mean "path" ? | 21:38 |
*** GheRivero_ has joined #openstack | 21:38 | |
vidd | the vm is on computer "B".... | 21:38 |
vidd | you tell computer "a" that you want to see that vm | 21:39 |
coli | vncproxy works on management node, which has access to private vlan over which compute node is available, vnc is available on comput enode on port 59xx | 21:39 |
coli | vidd: it's all in the mysql | 21:39 |
vidd | right...and nova-network controlls the private vlan | 21:40 |
*** deshantm_laptop has quit IRC | 21:40 | |
coli | and that database was populated with data by "nova-manage network create....." | 21:40 |
vidd | its possible that my understanding of how the pieces work together is off =] | 21:41 |
coli | true that nova-network controls private vlan but except for communication over that vlan these is nothing to control (or is there) on the managment node ? (no nat, no iptables, etc.) | 21:41 |
*** odyi has quit IRC | 21:41 | |
coli | as you know I'm completly new to this (openstack) and trying to undrestand how it works. | 21:41 |
vidd | hehe...im trying to understand it all as well =] | 21:42 |
coli | to me it seems that if I create by hand vlan100 with vlan_id 100 and attach virbr0 to it then dnsmasq will be able to communicate with the vlan | 21:42 |
coli | virbr0 has already an ip assigned from private range | 21:42 |
*** odyi has joined #openstack | 21:43 | |
*** odyi has joined #openstack | 21:43 | |
*** sdake has quit IRC | 21:43 | |
*** HowardRoark has joined #openstack | 21:43 | |
coli | I will try that before I install nova-network and we will see :-) | 21:43 |
coli | my undrestanding is that nova-network manages all the pieces of networking for instances | 21:43 |
coli | and on management node there are no instances as nova-compute doesn live here :-) | 21:44 |
*** HowardRoark has left #openstack | 21:44 | |
*** nati2 has quit IRC | 21:44 | |
*** HowardRoark has joined #openstack | 21:45 | |
vidd | coli, now that i think about it....its possible that im attributing rabbitmq jobs to the nova-network =] | 21:45 |
*** dailylinux has joined #openstack | 21:48 | |
*** ayates83 has quit IRC | 21:51 | |
*** Ryan_Lane|away is now known as Ryan_Lane | 21:55 | |
*** sandywalsh has quit IRC | 21:56 | |
*** pradeep has joined #openstack | 21:57 | |
*** swill has quit IRC | 21:57 | |
*** stanchan has quit IRC | 21:58 | |
*** stanchan has joined #openstack | 21:58 | |
coli | vidd: any idea what starts and stops dnsmasq ? | 21:58 |
vidd | no | 21:59 |
*** sdake has joined #openstack | 21:59 | |
*** mattray has quit IRC | 22:02 | |
*** nati2 has joined #openstack | 22:03 | |
*** negronjl has quit IRC | 22:04 | |
*** vipul__ has joined #openstack | 22:04 | |
*** jdg has quit IRC | 22:06 | |
*** rnorwood has quit IRC | 22:08 | |
*** GheRivero_ has quit IRC | 22:09 | |
*** dspano has joined #openstack | 22:10 | |
*** Vipul_ has quit IRC | 22:12 | |
*** jaypipes has quit IRC | 22:17 | |
Kiall | vidd, you've mentioned my name like 20 times? lol .. whats up? | 22:18 |
Kiall | coli, nova-network manages dnsmasq instances | 22:18 |
*** Ryan_Lane has quit IRC | 22:21 | |
*** bsza has quit IRC | 22:21 | |
*** nati2 has quit IRC | 22:22 | |
*** andrewbogott has quit IRC | 22:22 | |
*** HowardRoark has quit IRC | 22:23 | |
*** cereal_bars has joined #openstack | 22:23 | |
*** lts has quit IRC | 22:26 | |
uvirtbot | New bug: #894172 in openstack-qa "nova.compute.manager has unreachable path" [Undecided,New] https://launchpad.net/bugs/894172 | 22:26 |
jasona | hey kiall | 22:28 |
Kiall | heya, how did you get on yesterday? get everything going? | 22:29 |
jasona | it is all working. i have been poking various things now to understand | 22:29 |
jasona | people were happy to see it up. then freaked a bit that its up. so now i have to move it to somewhere more secure :) | 22:29 |
jasona | hence a q around changing ips after, or is that ./clean-all and reinstall again ? :) | 22:29 |
vidd | Kiall, i suspected a ug got introduced into your last keystone update | 22:30 |
vidd | but i had a missing : | 22:30 |
*** jdg has joined #openstack | 22:30 | |
Kiall | lol | 22:30 |
Kiall | I test the basic functionality of all the updates before pushing to launchpad ;) | 22:31 |
vidd | im trying to recreated the reported issues with my script | 22:31 |
jdg | Kiall: Nice job on your script! I have a node up and running, can access dashboard and launch instances. | 22:31 |
jdg | couple little things I've apparently gotten wrong, but at least it's working | 22:31 |
Kiall | heh .. as always ;) | 22:32 |
jdg | :) Yeah, for some reason I can't reach the VM's via the network or vnc. | 22:32 |
jdg | But at least it's progress | 22:32 |
*** sandywalsh has joined #openstack | 22:33 | |
vidd | i have to go take the family to dinner | 22:34 |
vidd | bbs | 22:34 |
*** CaptTofu1 has quit IRC | 22:34 | |
*** vidd is now known as vidd-away | 22:34 | |
Kiall | jdg, have you assigned them floating IPs? ;) | 22:35 |
jdg | Yes, assigned them through dashboard | 22:35 |
Kiall | and re VNC.. are you using chrome? nova-vncproxy doesnt seem to work with chrome. | 22:35 |
jdg | Ahh... yes, I'm using chrome, let me re-launch and try firefox | 22:36 |
*** nerens has quit IRC | 22:36 | |
*** Ryan_Lane has joined #openstack | 22:36 | |
*** dailylinux has quit IRC | 22:37 | |
*** rsampaio has quit IRC | 22:37 | |
Kiall | FF 7.0.1 is the only other browser I've tested in, I've heard IE works too | 22:37 |
*** jkyle has quit IRC | 22:37 | |
*** Ryan_Lane has joined #openstack | 22:38 | |
jdg | Hmm... ok, well at least in firefox it tried, but it turns out it's going to a 127 address | 22:38 |
*** mrjazzcat has joined #openstack | 22:39 | |
Kiall | Ah yea.. 1 sec i'll find the setting | 22:39 |
*** mrjazzcat has left #openstack | 22:39 | |
*** adjohn has quit IRC | 22:40 | |
Kiall | line 18 and 19 https://github.com/managedit/openstack-setup/blob/master/nova.conf.tmpl#L18 | 22:40 |
Kiall | (Leave L18 and 0.0.0.0!) | 22:40 |
*** llang629 has left #openstack | 22:40 | |
Kiall | as* | 22:41 |
jdg | ? | 22:41 |
*** arrsim has quit IRC | 22:42 | |
jdg | So it is 0.0.0.0 | 22:42 |
Kiall | the setting from L18 should be taken as is.. | 22:42 |
Kiall | the one from L19 should point at the vncproxy host | 22:42 |
*** nati2 has joined #openstack | 22:44 | |
jdg | Ok, so the template has it correct, but my /etc/nova/nova.conf has 127.0.0.1 | 22:44 |
jdg | So I can just modify nova.conf to match your template and restart nova perhaps? | 22:44 |
*** MarkAtwood has quit IRC | 22:44 | |
Kiall | Then you must have left HOST_IP as 127.0.0.1 in the settings file? | 22:44 |
*** MarkAtwood has joined #openstack | 22:44 | |
*** nyeates has quit IRC | 22:45 | |
Kiall | Despite the " Do actually change this! This should be an IP address accessible by your end users. Not 127.0.0.1." ;) | 22:45 |
*** redconnection has quit IRC | 22:48 | |
*** redconnection has joined #openstack | 22:48 | |
jdg | kiall: So actually I should probably edit nova.conf and change all of the 127 entries to my controller node IP anyway. Not sure why this didn't work from the settings file. | 22:48 |
*** jkerby has left #openstack | 22:49 | |
*** RobertLaptop has joined #openstack | 22:49 | |
Kiall | You must not have changed it in the settings file ;) | 22:49 |
Kiall | But Yea.. Basically nothing should be 127.0.0.1 | 22:50 |
jdg | Guess not, just thought I did :) | 22:50 |
*** debo-os has joined #openstack | 22:50 | |
*** zul has quit IRC | 22:50 | |
*** dspano has quit IRC | 22:50 | |
*** zul has joined #openstack | 22:50 | |
Kiall | time to force people to change that ;) https://github.com/managedit/openstack-setup/commit/4fe63fa0e49536595f9884678a0b2b48fd4e19d0 | 22:51 |
*** zul has quit IRC | 22:51 | |
*** zul has joined #openstack | 22:51 | |
Kiall | 192.0.2.0/24 == Block reserved for documentation ;) | 22:51 |
*** afm has joined #openstack | 22:51 | |
coli | Kiall: when you have time I have few questions about your scripts :-) | 22:52 |
Kiall | shoot.. | 22:52 |
coli | Kiall: do you script allow for situation where a management node doesn't have a nova-compute (so we don't run nova-compute.sh on it) ? | 22:52 |
Kiall | Yea, Dont use the "All in one server" instructions :) | 22:53 |
*** redconnection has quit IRC | 22:53 | |
*** zul has quit IRC | 22:54 | |
*** zul has joined #openstack | 22:54 | |
coli | Kiall: on compute node we have vlan100 and br100 interfaces, vlan100 is member of br100 bridge | 22:55 |
coli | Kiall; how is vlan100 communicating with management node ? | 22:55 |
coli | Kiall: on management node after running your scripts (excluding nova-compute) we only have virbr0 which doesn't have any interfaces attached to it | 22:56 |
Kiall | the vlan traffic is for instance->instance stuff, so the mgmt node doesnt need to see it | 22:56 |
Kiall | only nova-network and nova-compute nodes would ever get those bridges/vlans | 22:57 |
coli | that clarifies the vlan100 issue | 22:57 |
coli | in other words compute nodes and management node don't communicate over eth1 interfaces ? | 22:57 |
Kiall | And, the vlan is instance->nova-network aswell (where it gets NAT'd) | 22:57 |
*** debo-os has quit IRC | 22:57 | |
Kiall | Nope - eth0 typically is where all the nova controller traffic goes.. | 22:58 |
coli | therefore managemnt node is not using eth1 at all ? | 22:59 |
Kiall | Yes - the mgmt node doesnt use it.. | 22:59 |
Kiall | In larger deployments you might go with 3 or 4 interfaces per server, but my scripts assume the minimun .. | 23:00 |
coli | ufff.... got it now :-) | 23:00 |
coli | your scripts are very nice, it's just lack of my knowledge of how things work :-) | 23:00 |
coli | so vncproxy communicates on eth0 interfaces in order to connec tto port 59xx on the compute nodes ? | 23:00 |
Kiall | 1) might be the HW node IPs+nova comunication...2) might be a 10Gb port for nova-volume traffic.. 3) might be for floating IP traffic from the world, and 4) might be for project VLANs etc | 23:01 |
Kiall | coli, yes.. | 23:01 |
Kiall | We take #1+2+3 and merge them onto eth0, then leave eth1 for #4 in my scripts for simplicity | 23:01 |
*** dachary has quit IRC | 23:03 | |
*** afm has quit IRC | 23:07 | |
*** jkyle has joined #openstack | 23:08 | |
*** rnorwood has joined #openstack | 23:09 | |
*** rsampaio has joined #openstack | 23:10 | |
jdg | Kiall: interesting, whatever I did to change nova.conf is now giving me the same issue that I was having with Vidd's script last night. | 23:11 |
uvirtbot | New bug: #894186 in horizon "Bad credentials casuses "unable to find tenant" message" [Low,Fix committed] https://launchpad.net/bugs/894186 | 23:11 |
jdg | ECONNREFUSED when trying to go to instances or images | 23:11 |
Kiall | "to go to instances"? | 23:12 |
jdg | sorry... from inside dashboard | 23:12 |
jdg | When I click on "Instances" or "Images" | 23:13 |
Kiall | Yea 0 you used 127, so stuff is probably off in the keystone DB | 23:13 |
jdg | I changed my nova.conf: http://paste.openstack.org/show/3525/ | 23:13 |
jdg | Ahhh, that would make sense | 23:13 |
Kiall | ~ at the end? | 23:14 |
Kiall | Anyway - nova.conf looks fine other than that + mysql pass, which I assume you just removed | 23:15 |
Kiall | have a look at the keystone DB in the endpoint_templates table | 23:15 |
Kiall | change any 127's to real IPs | 23:15 |
Kiall | should be 9 of them I think | 23:15 |
*** n88 has quit IRC | 23:16 | |
*** JesperA has quit IRC | 23:16 | |
jdg | Hmm... no I didn't remove password from the mysql entry. | 23:16 |
Kiall | lol | 23:16 |
Kiall | I guess you dont have a password set so ;) | 23:17 |
jdg | :) I must have nuked it accidentally when adding the IP's | 23:17 |
jdg | It is set when I try to access mysql | 23:17 |
*** TheOsprey has quit IRC | 23:17 | |
jdg | Ok, so you were saying change in endpoint_templates... | 23:18 |
Kiall | yea - check for 127.0.0.1's | 23:19 |
coli | kiall: it seems like vncproxy is using hostnames in order to connec to compute-nodes to connect to 59xx ports. shouldn't it use an ip address and not a hostname which in my case resolves to totaly different ip ? | 23:21 |
Kiall | coli, what vnc related settings have you got in your nova.conf? | 23:21 |
coli | kiall: it's all installed from your script :-) no changes (so far) | 23:22 |
*** CaptTofu has quit IRC | 23:22 | |
Kiall | Okay, that makes no sense then ;) What makes you think its using a hostname? | 23:22 |
jdg | kiall: Yep, all the keystone IP entries in the DB are still 127.0.0.1 | 23:23 |
coli | kiall: I'm sure it's using hostname to connect as it resolves it :-) how did I get there ? | 23:23 |
Kiall | coli, I've no idea how its using a hostname! when you say it resolves it.. what do you mean, how/where/when do you see that happening? | 23:24 |
coli | kiall: as noVNC was stopping at "starting vnc handshake" I have tcpdumped all traffic on eth0 on managemnt node for 5900 port | 23:24 |
coli | kiall: it has shown that manadement node is trying to connect to 5900 port on an ip which resolves to wildcard for our domain name | 23:25 |
Kiall | whats is "--vncproxy_url" set to in nova.conf? | 23:25 |
coli | kiall: /etc/resolve has our domain name as search option | 23:25 |
Kiall | Humm - actually... | 23:26 |
coli | kiall: then I have tcpdumped eth0 for port 53 and restarted novnc. bang there it goes a request for A record for hostname.our-domain (hostname for compute node) | 23:26 |
coli | kiall: inserted hostname in /etc/hosts and works fine :-) | 23:26 |
Kiall | Yea - just figured it out.. | 23:26 |
coli | therefore it must use hostname instead of ip of a compute node/ | 23:26 |
Kiall | Go have a look at the nova DB's services table, it uses the hostname (not FQDN, just the host part) | 23:27 |
Kiall | Thats what nova-vncproxy tries to connect to.. | 23:27 |
*** MarkAtwood has quit IRC | 23:27 | |
*** nid0 has quit IRC | 23:27 | |
coli | does it mean that all hostames for compute nodes have to resolve to proper ip addresses ? | 23:27 |
Kiall | Well - Thats standard practice isnt it? ;) | 23:28 |
coli | hehehe..... | 23:28 |
coli | rotfl | 23:28 |
Kiall | I think thats what's happening anyway.. but it makes sense | 23:28 |
coli | if it uses just the host part then it's fgoignt to have lots of problems.... | 23:29 |
Kiall | I've never setup openstack on servers without valid forward DNS+reverse DNS+hostname .. Except for single server tests.. So I actually dont know what its going to do with invalid info | 23:29 |
coli | it's not going to connect (vncproxy) :-) | 23:30 |
Kiall | Anyway - I'm off! cya later... | 23:31 |
*** Ryan_Lane has quit IRC | 23:32 | |
coli | thank you | 23:33 |
*** nid0 has joined #openstack | 23:33 | |
*** vipul__ has quit IRC | 23:33 | |
*** po has joined #openstack | 23:35 | |
*** dirkx_ has joined #openstack | 23:38 | |
*** dotdevops has quit IRC | 23:40 | |
*** dirkx_ has quit IRC | 23:40 | |
*** MarkAtwood has joined #openstack | 23:43 | |
*** bryguy has quit IRC | 23:45 | |
*** bryguy has joined #openstack | 23:45 | |
*** rsampaio has quit IRC | 23:47 | |
*** DW-10297 has quit IRC | 23:49 | |
*** po has quit IRC | 23:52 | |
*** redconnection has joined #openstack | 23:53 | |
*** martine has joined #openstack | 23:53 | |
*** _rfz has joined #openstack | 23:54 | |
*** lzyeval has joined #openstack | 23:56 | |
_rfz | Howdy | 23:57 |
_rfz | Kiall, the libvirt issues fixed itself | 23:57 |
_rfz | :) | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!