| openstackhudson | Project dashboard-tarmac build #2,288: STILL FAILING in 36 sec: http://hudson.openstack.org/job/dashboard-tarmac/2288/ | 00:01 |
|---|---|---|
| openstackhudson | Project nova-tarmac build #59,909: STILL FAILING in 36 sec: http://hudson.openstack.org/job/nova-tarmac/59909/ | 00:02 |
| *** ranger57 has quit IRC | 00:03 | |
| *** burris has quit IRC | 00:05 | |
| openstackhudson | Project dashboard-tarmac build #2,289: STILL FAILING in 35 sec: http://hudson.openstack.org/job/dashboard-tarmac/2289/ | 00:06 |
| openstackhudson | Project nova-tarmac build #59,910: STILL FAILING in 35 sec: http://hudson.openstack.org/job/nova-tarmac/59910/ | 00:07 |
| *** Ryan_Lane is now known as Ryan_Lane|away | 00:08 | |
| jk0 | jaypipes: all ready for the talk tonight? I've got VLC all ready to go :) | 00:09 |
| jk0 | don't be nervous | 00:09 |
| jk0 | hehe | 00:09 |
| *** arun_ has quit IRC | 00:10 | |
| *** burris has joined #openstack | 00:10 | |
| openstackhudson | Yippie, build fixed! | 00:10 |
| openstackhudson | Project dashboard-tarmac build #2,290: FIXED in 2 sec: http://hudson.openstack.org/job/dashboard-tarmac/2290/ | 00:10 |
| openstackhudson | Yippie, build fixed! | 00:11 |
| openstackhudson | Project nova-tarmac build #59,911: FIXED in 2.8 sec: http://hudson.openstack.org/job/nova-tarmac/59911/ | 00:11 |
| jaypipes | jk0: last week's talk was apparently 2.5 hours according to stephen. we talked today and agreed that was not going to happen tonight ;) | 00:12 |
| jk0 | haha | 00:12 |
| *** iammartian has quit IRC | 00:16 | |
| *** iammartian has joined #openstack | 00:18 | |
| *** fysa has quit IRC | 00:19 | |
| jarrod | man | 00:21 |
| jarrod | trying to publish a tarball keeps hanging | 00:22 |
| jarrod | funnnnnn | 00:22 |
| *** msassak has quit IRC | 00:23 | |
| *** j05h has quit IRC | 00:24 | |
| *** j05h has joined #openstack | 00:24 | |
| *** dragondm has quit IRC | 00:25 | |
| *** azneita has joined #openstack | 00:31 | |
| *** kashyap_ has quit IRC | 00:33 | |
| *** iammartian has quit IRC | 00:33 | |
| *** kashyapc has quit IRC | 00:34 | |
| *** kashyap_ has joined #openstack | 00:35 | |
| *** iammartian has joined #openstack | 00:35 | |
| *** grapex has left #openstack | 00:36 | |
| *** nelson has quit IRC | 00:36 | |
| *** nelson has joined #openstack | 00:37 | |
| nelson | creiht: gettin' there! | 00:37 |
| nelson | https://alsted.wikimedia.org:8080/wikipedia/commons/thumb/a/aa/Dzimbo_u_Beogradu_19.jpeg/448px-Dzimbo_u_Beogradu_19.jpeg | 00:37 |
| jarrod | crap, bad version of boto | 00:38 |
| nelson | creiht: I spent the last two days figuring out how to write a filter and put it into the pipeline for the proxy. | 00:38 |
| nelson | creiht: so now, it's rewriting the external name into an AUTH string, a container, and object name. | 00:38 |
| *** hub_cap has joined #openstack | 00:45 | |
| *** iammartian has quit IRC | 00:46 | |
| *** iammartian has joined #openstack | 00:48 | |
| jarrod | NoValidHost: No hosts found | 00:48 |
| *** kashyapc has joined #openstack | 00:50 | |
| *** rlucio has quit IRC | 00:54 | |
| *** iammartian has quit IRC | 00:58 | |
| jarrod | libvirtError: unknown OS type hvm | 00:58 |
| *** iammartian has joined #openstack | 00:59 | |
| *** eikke has quit IRC | 01:01 | |
| *** vvuksan has joined #openstack | 01:03 | |
| *** vvuksan has quit IRC | 01:04 | |
| alekibango | jarrod: kvm? | 01:09 |
| alekibango | os type? :) | 01:09 |
| jarrod | vt not enabled :( | 01:09 |
| alekibango | use qemu ... | 01:09 |
| jarrod | i did on one server, but not this one | 01:09 |
| alekibango | or xen | 01:10 |
| alekibango | or uml | 01:10 |
| jarrod | i will use xen | 01:10 |
| jarrod | im going to reinstall everything doing to multiple server setup | 01:10 |
| jarrod | i didnt even have an option to specify hypervisor on bexar install | 01:10 |
| *** iammartian has quit IRC | 01:16 | |
| jarrod | libvirtError: internal error Could not get access to ACL tech driver 'ebiptables' | 01:17 |
| jarrod | yay | 01:17 |
| *** aimon has quit IRC | 01:17 | |
| *** rbergero1 has joined #openstack | 01:18 | |
| *** iammartian has joined #openstack | 01:18 | |
| jk0 | jaypipes: you have a URL handy for the stream? none on the website seem to be up (unless the stream just isn't running yet) | 01:23 |
| *** kpepple_ has joined #openstack | 01:23 | |
| jaypipes | jk0: still trying to get ahold of the hosts... :( conf line goes to voicemail, stephen's mobile goes to voicemail, and the host's skype contact doesn't reply... | 01:24 |
| jk0 | oh that's nice | 01:24 |
| jarrod | wtf is the ebiptables ACL tech driver | 01:25 |
| jaypipes | jarrod: I'm sensing a little frustration from you ;) | 01:26 |
| jarrod | its just around every turn heh | 01:27 |
| jarrod | i get something new | 01:27 |
| jaypipes | jarrod: keep pushing through :) | 01:27 |
| jk0 | stream is up now | 01:27 |
| jk0 | http://stream.tlltsarchive.org:8000/tllts | 01:27 |
| jaypipes | jarrod: I understand it can be frustrating, but we need input from ya. | 01:27 |
| *** aimon has joined #openstack | 01:29 | |
| *** kpepple_ has quit IRC | 01:30 | |
| *** kpepple_ has joined #openstack | 01:30 | |
| *** kpepple_ has quit IRC | 01:35 | |
| kpepple | jaypipes: can you hear them talking about you ? | 01:35 |
| jaypipes | kpepple: yep. | 01:36 |
| jaypipes | kpepple: I've been emailing Dave Yates this whole time... | 01:36 |
| jaypipes | really, really annoying. | 01:36 |
| kpepple | jaypipes: maybe you could call him on his "mexican" headset | 01:37 |
| jaypipes | what kind of podcast has people eating during the podcast? | 01:37 |
| jarrod | the database keeps locking | 01:37 |
| jarrod | for crying outloud | 01:37 |
| jaypipes | jarrod: how so? mysql? | 01:39 |
| jaypipes | jarrod: sqlite? | 01:39 |
| jarrod | sqlite -- tomorrow i get the rest up in a mysql | 01:39 |
| jaypipes | k | 01:39 |
| *** arun_ has joined #openstack | 01:39 | |
| *** arun_ has joined #openstack | 01:39 | |
| jarrod | yea i cant get past this | 01:40 |
| jarrod | gnite | 01:40 |
| *** hub_cap has quit IRC | 01:41 | |
| *** joearnold has quit IRC | 01:41 | |
| *** vvuksan has joined #openstack | 01:43 | |
| *** kpepple_ has joined #openstack | 01:47 | |
| *** iammartian has quit IRC | 01:49 | |
| *** MarkAtwood has joined #openstack | 01:51 | |
| *** kpepple_ has quit IRC | 01:51 | |
| *** alekibango has quit IRC | 01:52 | |
| *** iammartian has joined #openstack | 01:52 | |
| *** kpepple_ has joined #openstack | 01:52 | |
| *** kpepple_ has joined #openstack | 01:53 | |
| *** kpepple_ has quit IRC | 01:54 | |
| *** kpepple_ has joined #openstack | 01:55 | |
| *** MarkAtwood has quit IRC | 01:55 | |
| *** kpepple has quit IRC | 01:55 | |
| *** kpepple_ is now known as kpepple | 01:55 | |
| *** troytoman-away has quit IRC | 01:56 | |
| *** troytoman-away has joined #openstack | 01:56 | |
| *** kpepple has quit IRC | 01:58 | |
| *** kpepple has joined #openstack | 01:59 | |
| *** kpepple has quit IRC | 02:03 | |
| *** rbergero1 has quit IRC | 02:05 | |
| justinsb | jarrod: The SQLite DB will keep locking, but as long as everything is working, then everything will eventually recover (i.e. it won't deadlock). Not that I'm a big fan of the locking... | 02:08 |
| *** kpepple has joined #openstack | 02:08 | |
| justinsb | jarrod: I also hit the ebiptables issue when I was trying a different networking mode. I can't remember exactly what I did, but I think I installed a package with a similar name, and then restarted /etc/init.d/libvirt-bin. If you don't restart libvirt-bin it doesn't work | 02:09 |
| justinsb | jarrod: _but_ I never got that other networking mode to work. In the end, I just use the contrib/nova.sh script (I think vish wrote this script) | 02:09 |
| justinsb | jarrod: I think it uses vlan networking | 02:09 |
| *** iammartian has quit IRC | 02:11 | |
| *** kpepple has quit IRC | 02:12 | |
| *** kpepple has joined #openstack | 02:12 | |
| *** iammartian has joined #openstack | 02:14 | |
| *** kpepple has quit IRC | 02:17 | |
| *** kpepple has joined #openstack | 02:17 | |
| *** drico has quit IRC | 02:19 | |
| *** troytoman-away is now known as troytoman | 02:21 | |
| *** kpepple has quit IRC | 02:21 | |
| *** troytoman is now known as troytoman-away | 02:22 | |
| *** kpepple has joined #openstack | 02:22 | |
| *** azneita has quit IRC | 02:23 | |
| *** azneita has joined #openstack | 02:23 | |
| *** azneita has quit IRC | 02:24 | |
| *** azneita has joined #openstack | 02:24 | |
| *** azneita has joined #openstack | 02:26 | |
| *** azneita has joined #openstack | 02:26 | |
| *** Dweezahr has quit IRC | 02:27 | |
| *** kpepple has quit IRC | 02:31 | |
| *** kpepple has joined #openstack | 02:32 | |
| *** kpepple has quit IRC | 02:37 | |
| *** kpepple has joined #openstack | 02:37 | |
| *** kpepple has quit IRC | 02:42 | |
| *** fcarsten has quit IRC | 02:42 | |
| *** kpepple has joined #openstack | 02:42 | |
| *** maplebed has quit IRC | 02:44 | |
| *** kpepple has quit IRC | 02:46 | |
| *** kpepple has joined #openstack | 02:48 | |
| *** kpepple has quit IRC | 02:48 | |
| *** kpepple has joined #openstack | 02:49 | |
| *** iammartian has quit IRC | 02:51 | |
| *** iammartian has joined #openstack | 02:52 | |
| *** kpepple has quit IRC | 02:53 | |
| *** kpepple has joined #openstack | 02:54 | |
| *** kpepple has quit IRC | 02:54 | |
| *** kpepple has joined #openstack | 02:55 | |
| *** kpepple has joined #openstack | 02:56 | |
| *** kpepple has joined #openstack | 02:57 | |
| *** kpepple has quit IRC | 03:02 | |
| *** zul has quit IRC | 03:02 | |
| *** reldan has quit IRC | 03:02 | |
| *** kpepple has joined #openstack | 03:02 | |
| *** zul has joined #openstack | 03:03 | |
| *** kpepple has quit IRC | 03:07 | |
| *** kpepple has joined #openstack | 03:08 | |
| *** kpepple has quit IRC | 03:08 | |
| *** kpepple has joined #openstack | 03:09 | |
| *** MarkAtwood has joined #openstack | 03:09 | |
| *** kpepple has quit IRC | 03:11 | |
| *** kpepple has joined #openstack | 03:11 | |
| *** adjohn has joined #openstack | 03:12 | |
| *** iammartian has quit IRC | 03:13 | |
| *** iammartian has joined #openstack | 03:14 | |
| *** kpepple has quit IRC | 03:17 | |
| *** kpepple has joined #openstack | 03:18 | |
| *** Ryan_Lane|away is now known as Ryan_Lane | 03:20 | |
| *** kpepple has quit IRC | 03:23 | |
| *** kpepple has joined #openstack | 03:24 | |
| *** kpepple has quit IRC | 03:26 | |
| *** KnightHacker has left #openstack | 03:27 | |
| *** kpepple has joined #openstack | 03:27 | |
| *** kpepple has quit IRC | 03:29 | |
| *** iammartian has quit IRC | 03:29 | |
| *** kpepple has joined #openstack | 03:29 | |
| *** kpepple has quit IRC | 03:32 | |
| *** iammartian has joined #openstack | 03:32 | |
| *** kpepple has joined #openstack | 03:32 | |
| *** kpepple has joined #openstack | 03:33 | |
| *** kpepple has quit IRC | 03:35 | |
| *** kpepple has joined #openstack | 03:36 | |
| *** kpepple has quit IRC | 03:40 | |
| *** kpepple has joined #openstack | 03:41 | |
| *** iammartian has quit IRC | 03:43 | |
| *** iammartian has joined #openstack | 03:44 | |
| *** kpepple has quit IRC | 03:45 | |
| *** kpepple_ has joined #openstack | 03:45 | |
| *** littleidea has joined #openstack | 03:46 | |
| *** kpepple_ has quit IRC | 03:47 | |
| *** kpepple has joined #openstack | 03:47 | |
| *** kpepple has quit IRC | 03:52 | |
| *** kpepple_ has joined #openstack | 03:52 | |
| *** kashyap_ has quit IRC | 03:52 | |
| *** kpepple_ has quit IRC | 03:54 | |
| *** kpepple has joined #openstack | 03:54 | |
| *** vvuksan has quit IRC | 03:55 | |
| *** littleidea has quit IRC | 03:56 | |
| *** iammartian has quit IRC | 03:57 | |
| *** iammartian has joined #openstack | 03:59 | |
| *** kpepple has quit IRC | 04:01 | |
| *** kpepple has joined #openstack | 04:02 | |
| *** kpepple has quit IRC | 04:02 | |
| *** kpepple has joined #openstack | 04:03 | |
| *** hadrian has quit IRC | 04:03 | |
| *** kpepple has quit IRC | 04:07 | |
| *** kpepple has joined #openstack | 04:08 | |
| *** kpepple has quit IRC | 04:09 | |
| *** kpepple has joined #openstack | 04:09 | |
| *** iammartian has quit IRC | 04:13 | |
| *** kpepple has quit IRC | 04:14 | |
| *** kpepple has joined #openstack | 04:14 | |
| *** iammartian has joined #openstack | 04:15 | |
| *** sateesh has joined #openstack | 04:17 | |
| *** kpepple has quit IRC | 04:19 | |
| *** kpepple has joined #openstack | 04:20 | |
| jarrod | libvir: Security Labeling error : internal error error calling aa_change_profile() <-- there is nothing but problem after problem | 04:21 |
| *** kpepple has quit IRC | 04:22 | |
| *** kpepple has joined #openstack | 04:22 | |
| *** kpepple has joined #openstack | 04:24 | |
| *** kpepple has quit IRC | 04:26 | |
| *** kpepple has joined #openstack | 04:27 | |
| *** kpepple has joined #openstack | 04:28 | |
| *** kpepple has quit IRC | 04:33 | |
| *** kpepple has joined #openstack | 04:33 | |
| *** kpepple has joined #openstack | 04:34 | |
| *** kpepple has quit IRC | 04:36 | |
| *** kpepple has joined #openstack | 04:36 | |
| *** kpepple has joined #openstack | 04:38 | |
| *** aimon has quit IRC | 04:38 | |
| *** aimon has joined #openstack | 04:40 | |
| *** nelson has quit IRC | 04:41 | |
| *** littleidea has joined #openstack | 04:41 | |
| justinsb | jarrod: Did you restart the libvirt daemon? It | 04:42 |
| *** kpepple has quit IRC | 04:42 | |
| justinsb | jarrod: It has a habit of misbehaving in odd ways if it hasn't been restarted since a config change | 04:42 |
| *** kpepple has joined #openstack | 04:43 | |
| jarrod | yes | 04:44 |
| jarrod | and i tried stopping apparmor | 04:44 |
| *** iammartian has quit IRC | 04:47 | |
| *** KnightHacker has joined #openstack | 04:48 | |
| *** iammartian has joined #openstack | 04:49 | |
| *** guynaor has joined #openstack | 04:49 | |
| *** guynaor has left #openstack | 04:49 | |
| *** kpepple has quit IRC | 04:49 | |
| *** kpepple has joined #openstack | 04:50 | |
| justinsb | jarrod: I had lots of trouble until I used vish's script (contrib/nova.sh I think). Otherwise it felt like there were too many conflicting sets of instructions, none of which really worked. I followed this page: http://wiki.openstack.org/NovaInstall/DevInstallScript | 04:51 |
| jarrod | that installs bexar? | 04:52 |
| jarrod | ill try it | 04:52 |
| jarrod | thanks | 04:52 |
| justinsb | jarrod: I think it pulls from trunk actually. So Bexar++ :-) | 04:53 |
| jarrod | wow | 04:53 |
| jarrod | dangit | 04:53 |
| jarrod | i did it the hard way heh | 04:53 |
| justinsb | Yes - I went through the same process :-) | 04:54 |
| *** kpepple_ has joined #openstack | 04:55 | |
| *** kpepple has quit IRC | 04:55 | |
| jarrod | i can do a multiserver install with his script? | 04:55 |
| jarrod | ill figure it out | 04:56 |
| jarrod | thanks for pointing me there | 04:56 |
| jarrod | i think im going to switch to Xen | 04:56 |
| *** pquerna has joined #openstack | 04:56 | |
| *** littleidea has quit IRC | 04:57 | |
| justinsb | jarrod: I don't think you can do a multi-server install per-se, it's really the dev install | 04:57 |
| justinsb | jarrod: It brings up a basic stand-alone config | 04:57 |
| justinsb | jarrod: You have to reconfigure it to do anything different | 04:57 |
| justinsb | jarrod: But I found it a great starting point | 04:58 |
| justinsb | jarrod: I'd stick with KVM initially, just because that's what ships in the default script, and then only change once you've got it going | 04:58 |
| justinsb | jarrod: But it's up to you! | 04:58 |
| *** freeflying has quit IRC | 05:00 | |
| *** freeflying has joined #openstack | 05:00 | |
| *** kpepple_ has quit IRC | 05:02 | |
| *** kpepple has joined #openstack | 05:03 | |
| *** kpepple has joined #openstack | 05:04 | |
| *** pvo has joined #openstack | 05:05 | |
| *** kpepple has quit IRC | 05:06 | |
| *** kpepple has joined #openstack | 05:06 | |
| *** pvo has quit IRC | 05:07 | |
| *** jfluhmann has joined #openstack | 05:08 | |
| *** kpepple has quit IRC | 05:13 | |
| *** kpepple has joined #openstack | 05:13 | |
| *** iammartian has quit IRC | 05:15 | |
| jarrod | maybe something is conflicting, but even the nova.sh never launches the vm | 05:15 |
| jarrod | ughh | 05:15 |
| jarrod | ill just reinstall fresh tomorrow | 05:15 |
| *** iammartian has joined #openstack | 05:17 | |
| *** kpepple has quit IRC | 05:18 | |
| *** MarkAtwood has quit IRC | 05:18 | |
| *** kpepple has joined #openstack | 05:18 | |
| *** MarkAtwood has joined #openstack | 05:18 | |
| *** kpepple has quit IRC | 05:22 | |
| *** kpepple has joined #openstack | 05:23 | |
| *** kpepple has joined #openstack | 05:25 | |
| *** kpepple has quit IRC | 05:26 | |
| *** kpepple has joined #openstack | 05:27 | |
| *** kpepple has quit IRC | 05:28 | |
| *** kpepple has joined #openstack | 05:29 | |
| *** kpepple has quit IRC | 05:30 | |
| *** kpepple has joined #openstack | 05:31 | |
| *** Ryan_Lane is now known as Ryan_Lane|away | 05:31 | |
| *** iammartian has quit IRC | 05:34 | |
| *** kpepple has quit IRC | 05:35 | |
| *** littleidea has joined #openstack | 05:36 | |
| *** iammartian has joined #openstack | 05:36 | |
| *** kpepple has joined #openstack | 05:36 | |
| *** littleidea_ has joined #openstack | 05:37 | |
| *** littleidea has quit IRC | 05:40 | |
| *** littleidea_ is now known as littleidea | 05:40 | |
| *** kashyapc has quit IRC | 05:43 | |
| *** f4m8_ is now known as f4m8 | 05:44 | |
| *** kpepple has quit IRC | 05:45 | |
| *** kpepple has joined #openstack | 05:45 | |
| *** kpepple has quit IRC | 05:50 | |
| *** kpepple has joined #openstack | 05:50 | |
| *** kpepple has quit IRC | 05:52 | |
| *** kpepple has joined #openstack | 05:52 | |
| *** kpepple has quit IRC | 05:54 | |
| *** kpepple has joined #openstack | 05:54 | |
| *** kpepple_ has joined #openstack | 05:59 | |
| *** kpepple has quit IRC | 05:59 | |
| *** kpepple_ has quit IRC | 06:04 | |
| *** kpepple has joined #openstack | 06:04 | |
| *** jfluhmann_away has quit IRC | 06:07 | |
| *** kpepple has quit IRC | 06:08 | |
| *** kpepple has joined #openstack | 06:09 | |
| *** kpepple has quit IRC | 06:11 | |
| *** kpepple has joined #openstack | 06:11 | |
| *** Hayati has joined #openstack | 06:14 | |
| *** kpepple has quit IRC | 06:15 | |
| *** kpepple has joined #openstack | 06:16 | |
| *** iammartian has quit IRC | 06:18 | |
| *** iammartian has joined #openstack | 06:19 | |
| *** kpepple has quit IRC | 06:20 | |
| *** kpepple has joined #openstack | 06:21 | |
| *** Cybodog has joined #openstack | 06:26 | |
| *** kpepple has quit IRC | 06:29 | |
| *** kpepple_ has joined #openstack | 06:29 | |
| *** omidhdl has joined #openstack | 06:30 | |
| *** iammartian has quit IRC | 06:32 | |
| *** kpepple_ has quit IRC | 06:34 | |
| *** kpepple has joined #openstack | 06:35 | |
| *** iammartian has joined #openstack | 06:36 | |
| *** kpepple has joined #openstack | 06:36 | |
| *** kpepple_ has joined #openstack | 06:41 | |
| *** kpepple has quit IRC | 06:41 | |
| *** kpepple_ has quit IRC | 06:45 | |
| *** iammartian has quit IRC | 06:50 | |
| *** Hayati has quit IRC | 06:51 | |
| *** iammartian has joined #openstack | 06:51 | |
| *** jfluhmann has quit IRC | 06:55 | |
| *** miclorb_ has quit IRC | 06:57 | |
| *** fysa has joined #openstack | 07:03 | |
| *** matiu has quit IRC | 07:03 | |
| *** iammartian has quit IRC | 07:04 | |
| *** iammartian has joined #openstack | 07:06 | |
| *** omidhdl has quit IRC | 07:06 | |
| *** guigui has joined #openstack | 07:09 | |
| *** ramkrsna has joined #openstack | 07:12 | |
| *** ramkrsna has joined #openstack | 07:12 | |
| *** omidhdl has joined #openstack | 07:15 | |
| *** iammartian has quit IRC | 07:17 | |
| *** iammartian has joined #openstack | 07:18 | |
| *** littleidea has joined #openstack | 07:19 | |
| *** berendt has joined #openstack | 07:27 | |
| *** iammartian has quit IRC | 07:30 | |
| *** kpepple has joined #openstack | 07:30 | |
| *** iammartian has joined #openstack | 07:31 | |
| *** CloudChris has joined #openstack | 07:37 | |
| *** kpepple_ has joined #openstack | 07:38 | |
| *** kpepple has quit IRC | 07:39 | |
| *** kpepple_ is now known as kpepple | 07:39 | |
| *** kpepple has quit IRC | 07:40 | |
| *** kpepple has joined #openstack | 07:40 | |
| *** iammartian has quit IRC | 07:43 | |
| *** iammartian has joined #openstack | 07:45 | |
| *** miclorb_ has joined #openstack | 07:53 | |
| *** kpepple has quit IRC | 07:54 | |
| *** kpepple has joined #openstack | 07:54 | |
| *** iammartian has quit IRC | 07:58 | |
| *** kpepple has quit IRC | 07:59 | |
| *** kpepple has joined #openstack | 07:59 | |
| *** iammartian has joined #openstack | 08:00 | |
| *** allsystemsarego has joined #openstack | 08:02 | |
| *** allsystemsarego has joined #openstack | 08:02 | |
| *** kpepple has quit IRC | 08:02 | |
| *** kpepple has joined #openstack | 08:03 | |
| *** kpepple has quit IRC | 08:10 | |
| *** kpepple has joined #openstack | 08:11 | |
| *** kpepple has quit IRC | 08:15 | |
| *** kpepple has joined #openstack | 08:16 | |
| *** littleidea has quit IRC | 08:16 | |
| *** littleidea has joined #openstack | 08:17 | |
| *** westmaas has quit IRC | 08:18 | |
| *** rcc has joined #openstack | 08:18 | |
| *** iammartian has quit IRC | 08:19 | |
| *** kpepple has quit IRC | 08:20 | |
| *** iammartian has joined #openstack | 08:21 | |
| *** kpepple has joined #openstack | 08:21 | |
| *** thimble has joined #openstack | 08:24 | |
| *** MarcMorata has joined #openstack | 08:25 | |
| *** jfluhmann has joined #openstack | 08:26 | |
| *** kpepple has quit IRC | 08:27 | |
| *** kpepple has joined #openstack | 08:27 | |
| *** MarcMorata has quit IRC | 08:31 | |
| *** kpepple has quit IRC | 08:32 | |
| *** perra has joined #openstack | 08:32 | |
| *** kpepple has joined #openstack | 08:32 | |
| *** kpepple has joined #openstack | 08:33 | |
| *** kpepple has quit IRC | 08:34 | |
| *** littleidea has quit IRC | 08:34 | |
| *** kpepple has joined #openstack | 08:34 | |
| *** kpepple has quit IRC | 08:35 | |
| *** kpepple has joined #openstack | 08:35 | |
| *** drico has joined #openstack | 08:37 | |
| *** jfluhmann has quit IRC | 08:38 | |
| *** kpepple has quit IRC | 08:40 | |
| *** kpepple has joined #openstack | 08:40 | |
| *** miclorb_ has quit IRC | 08:42 | |
| *** calavera has joined #openstack | 08:42 | |
| *** citral has joined #openstack | 08:43 | |
| *** MarcMorata has joined #openstack | 08:43 | |
| *** iammartian has quit IRC | 08:47 | |
| *** iammartian has joined #openstack | 08:49 | |
| *** kpepple has quit IRC | 08:51 | |
| *** kpepple has joined #openstack | 08:52 | |
| *** Nacx has joined #openstack | 08:53 | |
| *** kpepple has quit IRC | 08:53 | |
| *** kpepple has joined #openstack | 08:54 | |
| *** hadrian has joined #openstack | 08:54 | |
| *** Seoman has joined #openstack | 08:56 | |
| *** kpepple has quit IRC | 08:58 | |
| *** MarcMorata has quit IRC | 08:58 | |
| *** hadrian has quit IRC | 08:59 | |
| *** kpepple has joined #openstack | 08:59 | |
| *** azneita has quit IRC | 09:00 | |
| *** Seoman has quit IRC | 09:02 | |
| *** miclorb has joined #openstack | 09:05 | |
| *** aimon_ has joined #openstack | 09:07 | |
| *** aimon has quit IRC | 09:08 | |
| *** aimon_ is now known as aimon | 09:08 | |
| *** befreax has joined #openstack | 09:09 | |
| *** iammartian has quit IRC | 09:11 | |
| *** kpepple has quit IRC | 09:12 | |
| *** iammartian has joined #openstack | 09:14 | |
| *** kpepple has joined #openstack | 09:14 | |
| *** kpepple has quit IRC | 09:18 | |
| *** kpepple has joined #openstack | 09:19 | |
| *** MarcMorata has joined #openstack | 09:19 | |
| *** kpepple has quit IRC | 09:21 | |
| *** kpepple has joined #openstack | 09:21 | |
| *** drico has quit IRC | 09:24 | |
| *** kpepple has quit IRC | 09:25 | |
| *** kpepple has joined #openstack | 09:26 | |
| *** Hayati has joined #openstack | 09:27 | |
| *** nic_ has quit IRC | 09:27 | |
| *** CloudChris has quit IRC | 09:27 | |
| *** kpepple has quit IRC | 09:28 | |
| *** kpepple has joined #openstack | 09:28 | |
| *** krigstask has joined #openstack | 09:33 | |
| *** kpepple has quit IRC | 09:33 | |
| *** kpepple_ has joined #openstack | 09:33 | |
| *** Cybodog has quit IRC | 09:34 | |
| *** photron has joined #openstack | 09:36 | |
| *** kpepple_ has quit IRC | 09:38 | |
| *** MarcMorata has quit IRC | 09:38 | |
| *** MarcMorata has joined #openstack | 09:38 | |
| *** kpepple has joined #openstack | 09:38 | |
| *** kpepple has quit IRC | 09:40 | |
| *** kpepple has joined #openstack | 09:40 | |
| *** iammartian has quit IRC | 09:43 | |
| *** kpepple has quit IRC | 09:45 | |
| *** iammartian has joined #openstack | 09:45 | |
| *** kpepple has joined #openstack | 09:45 | |
| krigstask | Hello everyone | 09:46 |
| krigstask | Bexar release announcement says that it's now possible to handle KVM images natively | 09:46 |
| krigstask | But docs seem to lag behind and I can't found anything there | 09:47 |
| krigstask | How should one register KVM/Qemu image for Nova to handle and run instances of it? | 09:47 |
| *** kpepple has quit IRC | 09:50 | |
| *** kpepple has joined #openstack | 09:50 | |
| *** kpepple has quit IRC | 09:52 | |
| *** kpepple has joined #openstack | 09:52 | |
| *** kpepple has quit IRC | 09:56 | |
| *** kpepple has joined #openstack | 09:57 | |
| *** kpepple has quit IRC | 10:02 | |
| *** kpepple_ has joined #openstack | 10:02 | |
| *** kpepple_ has quit IRC | 10:03 | |
| *** kpepple has joined #openstack | 10:03 | |
| *** kpepple has quit IRC | 10:04 | |
| *** kpepple has joined #openstack | 10:04 | |
| *** iammartian has quit IRC | 10:06 | |
| *** iammartian has joined #openstack | 10:08 | |
| cdbs | Can someone kindly re-approve this merge: https://code.launchpad.net/~bilalakhtar/nova/fix-708347/+merge/49029 I wrapped the line to under 79 chars | 10:09 |
| *** eikke has joined #openstack | 10:11 | |
| *** kpepple has quit IRC | 10:11 | |
| *** kpepple has joined #openstack | 10:12 | |
| rcc | heh, pep8 issues? it drives me mad :/ | 10:15 |
| *** kpepple has quit IRC | 10:17 | |
| *** kpepple has joined #openstack | 10:18 | |
| *** CloudChris has joined #openstack | 10:18 | |
| *** MarcMorata has quit IRC | 10:20 | |
| *** miclorb has quit IRC | 10:20 | |
| *** kpepple has quit IRC | 10:23 | |
| *** kpepple has joined #openstack | 10:24 | |
| *** kpepple has quit IRC | 10:25 | |
| *** kpepple has joined #openstack | 10:26 | |
| *** kpepple has quit IRC | 10:26 | |
| *** kpepple has joined #openstack | 10:27 | |
| ttx | cdbs, rcc: no more excuses: https://lists.launchpad.net/openstack/msg00250.html :) | 10:27 |
| cdbs | ttx: thanks for the link :) | 10:28 |
| cdbs | ttx: so, can you mark that approved? | 10:28 |
| ttx | cdbs: no. I'm not nova-core. | 10:29 |
| ttx | soren: ^? | 10:29 |
| rcc | haha | 10:29 |
| rcc | awesome, thx ttx | 10:30 |
| * soren looks | 10:30 | |
| soren | rcc: The backslash isn't needed, but it doesn't matter. | 10:30 |
| soren | whoops | 10:31 |
| soren | cdbs: The backslash isn't needed, but it doesn't matter. | 10:31 |
| *** kpepple has quit IRC | 10:31 | |
| * soren approves | 10:31 | |
| cdbs | soren: it matters, since there is a % | 10:31 |
| cdbs | soren: without it, a syntax error comes | 10:31 |
| *** kpepple has joined #openstack | 10:32 | |
| * cdbs tests patches before pushing them | 10:32 | |
| *** kpepple has quit IRC | 10:33 | |
| *** MarcMorata has joined #openstack | 10:33 | |
| *** kpepple has joined #openstack | 10:34 | |
| *** kpepple has quit IRC | 10:34 | |
| soren | cdbs: I don't see why that would happen. | 10:35 |
| *** kpepple has joined #openstack | 10:35 | |
| cdbs | soren: Python automatically looks at the next line if a line ends with , | 10:35 |
| soren | cdbs: As I said: It's fine. It doesn't hurt. | 10:35 |
| cdbs | else it doesn't | 10:35 |
| cdbs | soren: :) | 10:35 |
| soren | cdbs: Uh, yes it does, if you're inside a set of parentheses. | 10:35 |
| soren | cdbs: ...and you are. | 10:36 |
| cdbs | :o | 10:36 |
| *** adjohn has quit IRC | 10:37 | |
| soren | cdbs: http://pastebin.com/AKSdMm8h | 10:37 |
| soren | cdbs: You can even split strings across multiple lines if you're inside parentheses. | 10:37 |
| cdbs | soren: okay, I believe in you, and /me now joins ~sorens-target-audience | 10:38 |
| soren | *chuckle* | 10:38 |
| * soren forgot about that :) | 10:38 | |
| cdbs | :) | 10:38 |
| soren | ttx: You weren't around for that were you? | 10:39 |
| ttx | I was | 10:39 |
| soren | ttx: Nah, couldn't have been. | 10:39 |
| soren | ttx: In Mountain View? | 10:39 |
| ttx | I was around, but not *there*. Anyway, I found the badge for the team. | 10:39 |
| soren | Ah :) | 10:40 |
| *** kpepple has quit IRC | 10:41 | |
| *** kpepple has joined #openstack | 10:42 | |
| openstackhudson | Project nova build #498: SUCCESS in 1 min 32 sec: http://hudson.openstack.org/job/nova/498/ | 10:42 |
| openstackhudson | Tarmac: This branch fixes bug #708347: RunInstances: Invalid instance type gives improper error message | 10:42 |
| uvirtbot | Launchpad bug 708347 in nova "RunInstances: Invalid instance type gives improper error message" [Low,Fix committed] https://launchpad.net/bugs/708347 | 10:42 |
| *** fabiand_ has joined #openstack | 10:43 | |
| *** kpepple has quit IRC | 10:48 | |
| *** kpepple_ has joined #openstack | 10:49 | |
| *** reldan has joined #openstack | 10:49 | |
| *** iammartian has quit IRC | 10:50 | |
| *** iammartian has joined #openstack | 10:52 | |
| *** kpepple_ has quit IRC | 10:53 | |
| *** kpepple has joined #openstack | 10:54 | |
| *** kpepple has quit IRC | 10:58 | |
| *** kpepple has joined #openstack | 10:59 | |
| *** reldan has quit IRC | 11:02 | |
| *** kpepple has quit IRC | 11:03 | |
| *** kpepple has joined #openstack | 11:04 | |
| *** iammartian has quit IRC | 11:05 | |
| *** iammartian has joined #openstack | 11:06 | |
| *** kpepple has quit IRC | 11:08 | |
| *** kpepple_ has joined #openstack | 11:09 | |
| *** perra has quit IRC | 11:10 | |
| *** kpepple_ has quit IRC | 11:13 | |
| *** kpepple has joined #openstack | 11:14 | |
| *** kpepple has quit IRC | 11:18 | |
| *** kpepple has joined #openstack | 11:19 | |
| *** kpepple has quit IRC | 11:20 | |
| *** kpepple has joined #openstack | 11:21 | |
| *** iammartian has quit IRC | 11:23 | |
| *** iammartian has joined #openstack | 11:25 | |
| *** kpepple has quit IRC | 11:25 | |
| *** kpepple has joined #openstack | 11:26 | |
| *** kpepple has quit IRC | 11:31 | |
| *** kpepple_ has joined #openstack | 11:31 | |
| *** kpepple_ has quit IRC | 11:35 | |
| *** kpepple has joined #openstack | 11:36 | |
| *** omidhdl has quit IRC | 11:37 | |
| *** omidhdl has joined #openstack | 11:38 | |
| *** reldan has joined #openstack | 11:39 | |
| *** kpepple has quit IRC | 11:40 | |
| *** kpepple has joined #openstack | 11:40 | |
| *** iammartian has quit IRC | 11:42 | |
| *** kpepple has quit IRC | 11:44 | |
| *** iammartian has joined #openstack | 11:44 | |
| *** kpepple has joined #openstack | 11:45 | |
| *** kpepple has quit IRC | 11:46 | |
| *** kpepple has joined #openstack | 11:47 | |
| *** kpepple has quit IRC | 11:51 | |
| *** kpepple has joined #openstack | 11:52 | |
| *** kpepple has quit IRC | 11:56 | |
| *** kpepple has joined #openstack | 11:57 | |
| *** iammartian has quit IRC | 11:58 | |
| cjreyn | hey all. I'm having issues with Flat networking mode and the metadata service. The redirection on the compute node does not automatically get set up. | 12:00 |
| *** kpepple has quit IRC | 12:00 | |
| *** kpepple has joined #openstack | 12:01 | |
| *** iammartian has joined #openstack | 12:01 | |
| cjreyn | I added an iptables entry to do this, as defined in the metadata_foward() of klinux_net.py, but still no joy | 12:01 |
| cjreyn | *linux_net.py srry | 12:01 |
| cjreyn | is this the only iptables line that provides this redirection, or am I missing something else? | 12:03 |
| *** kpepple has quit IRC | 12:05 | |
| *** kpepple has joined #openstack | 12:06 | |
| cjreyn | oh, and my nova.conf has the --ec2_dmz_host= line | 12:06 |
| *** kpepple has quit IRC | 12:08 | |
| *** kpepple has joined #openstack | 12:08 | |
| *** perra has joined #openstack | 12:10 | |
| *** kpepple has quit IRC | 12:12 | |
| *** kpepple has joined #openstack | 12:13 | |
| *** iammartian has quit IRC | 12:18 | |
| *** iammartian has joined #openstack | 12:19 | |
| *** kpepple_ has joined #openstack | 12:20 | |
| *** kpepple has quit IRC | 12:20 | |
| *** kpepple_ has quit IRC | 12:25 | |
| *** kpepple has joined #openstack | 12:25 | |
| *** omidhdl has quit IRC | 12:27 | |
| *** vvuksan has joined #openstack | 12:29 | |
| *** kpepple_ has joined #openstack | 12:30 | |
| *** kpepple has quit IRC | 12:30 | |
| *** iammartian has quit IRC | 12:32 | |
| *** kpepple_ has quit IRC | 12:34 | |
| *** iammartian has joined #openstack | 12:34 | |
| *** kpepple has joined #openstack | 12:35 | |
| *** kpepple has joined #openstack | 12:36 | |
| *** citral1 has joined #openstack | 12:37 | |
| *** kpepple has quit IRC | 12:39 | |
| *** kpepple has joined #openstack | 12:40 | |
| *** citral has quit IRC | 12:40 | |
| *** Hayati has quit IRC | 12:41 | |
| sandywalsh_ | which branch should I be merging from now for new development? Still lp:nova? | 12:42 |
| zykes- | has anyone started to look at openvswitch ? | 12:43 |
| ttx | sandywalsh_: yes, lp:nova points to cactus now | 12:43 |
| sandywalsh_ | thx ttx | 12:44 |
| *** reldan has quit IRC | 12:44 | |
| *** ctennis has quit IRC | 12:46 | |
| *** citral1 has quit IRC | 12:47 | |
| zykes- | ttx: you got a clue on that ? | 12:48 |
| *** kpepple has quit IRC | 12:49 | |
| ttx | zykes-: I looked superficially into it (read: I looked at the docs but did not deploy it) | 12:49 |
| *** kpepple has joined #openstack | 12:49 | |
| *** stewart_ has joined #openstack | 12:49 | |
| zykes- | ok | 12:49 |
| zykes- | it's too bad that it's not supported by libvirt yet | 12:50 |
| *** stewart has quit IRC | 12:52 | |
| *** kpepple has quit IRC | 12:54 | |
| *** kpepple has joined #openstack | 12:55 | |
| *** vvuksan has quit IRC | 12:57 | |
| *** kpepple has quit IRC | 12:59 | |
| *** kpepple has joined #openstack | 13:00 | |
| *** kpepple has quit IRC | 13:04 | |
| *** kpepple_ has joined #openstack | 13:04 | |
| *** iammartian has quit IRC | 13:06 | |
| *** iammartian has joined #openstack | 13:08 | |
| *** kpepple_ has quit IRC | 13:08 | |
| *** kpepple has joined #openstack | 13:09 | |
| *** kpepple has quit IRC | 13:13 | |
| sandywalsh_ | is there any code in nova/swift for safely storing credentials in a decrypt-able fashion? | 13:13 |
| *** kpepple has joined #openstack | 13:14 | |
| *** stewart__ has joined #openstack | 13:15 | |
| *** stewart_ has quit IRC | 13:17 | |
| *** hggdh_ has quit IRC | 13:18 | |
| *** iammartian has quit IRC | 13:20 | |
| *** iammartian has joined #openstack | 13:22 | |
| *** kpepple has quit IRC | 13:24 | |
| *** kpepple has joined #openstack | 13:25 | |
| sandywalsh_ | can we take this out of the code now? | 13:26 |
| sandywalsh_ | # vim: tabstop=4 shiftwidth=4 softtabstop=4 | 13:26 |
| jaypipes | sandywalsh_: why? | 13:28 |
| sandywalsh_ | makes the python yucky :) | 13:28 |
| *** kpepple has quit IRC | 13:29 | |
| *** kpepple has joined #openstack | 13:31 | |
| *** lionel__ is now known as lionel | 13:31 | |
| *** kpepple has quit IRC | 13:32 | |
| *** kpepple has joined #openstack | 13:32 | |
| *** jfluhmann has joined #openstack | 13:34 | |
| *** kpepple has quit IRC | 13:37 | |
| *** kpepple has joined #openstack | 13:37 | |
| *** kpepple has quit IRC | 13:39 | |
| *** kpepple has joined #openstack | 13:40 | |
| *** zul has quit IRC | 13:40 | |
| *** zul has joined #openstack | 13:40 | |
| uvirtbot | New bug: #716414 in nova "Traffic to floating IP does not work from network worker" [Undecided,New] https://launchpad.net/bugs/716414 | 13:42 |
| dabo | sandywalsh_: +1 | 13:42 |
| *** hggdh has joined #openstack | 13:43 | |
| *** kpepple has quit IRC | 13:44 | |
| *** kpepple has joined #openstack | 13:45 | |
| *** dprince has joined #openstack | 13:46 | |
| *** Hayati has joined #openstack | 13:47 | |
| soren | dabo, sandywalsh_: What are you guys' settings? | 13:51 |
| soren | Do any of you guys have an opinion on https://bugs.launchpad.net/nova/+bug/716427 ? | 13:51 |
| uvirtbot | Launchpad bug 716427 in nova "RPC concurrency problem" [Undecided,New] | 13:51 |
| soren | Specifically the last paragraph. | 13:52 |
| *** kpepple has quit IRC | 13:52 | |
| dabo | soren: I think the idea is that anyone should be able to edit the files the way they like. Edtior-specific markup really doesn't belong in code | 13:52 |
| *** kpepple has joined #openstack | 13:52 | |
| jaypipes | sandywalsh_, dabo: seems pretty petty to me :) keeps the tabs out of the code. I suppose if pep8 does that, it's not needed anymore, but I wouldn't exactly call removing modelines a top priority ;) | 13:52 |
| soren | dabo: I guess. As long as the stuff that lands in the files lines up with our coding standard, it's cool. | 13:53 |
| dabo | jaypipes: it only keeps tabs out if you're using vim :) | 13:53 |
| jaypipes | dabo: and if you're not, you should ignore it. | 13:53 |
| dabo | that's not the point | 13:53 |
| dabo | a) it's ugly | 13:53 |
| jaypipes | dabo: it's like emacs modelines... I just ignore em. I don't remove them fro source code just because I don't use emacs.... | 13:54 |
| dabo | b) it doesn't belong in code | 13:54 |
| dabo | c) it's editor-specific | 13:54 |
| jaypipes | dabo: who cares? really? this is something we need to focus on? :) | 13:54 |
| jaypipes | every open source project I've worked on has stuff like that in it... | 13:54 |
| dabo | not focus, no. sandywalsh_ only asked if it was ok to remove those lines | 13:54 |
| jaypipes | it's not a world-shaker | 13:54 |
| sandywalsh_ | it | 13:55 |
| sandywalsh_ | it's not a big deal ... just askin' | 13:55 |
| dabo | it also seems silly to have stuff like that but not important stuff like: # -*- coding: utf-8 -*- | 13:55 |
| dabo | you know, stuff that actually affects code | 13:55 |
| jaypipes | dabo: I really don't care to be honest. I tend to focus on the code, not stuff like this... | 13:55 |
| dabo | hmmm... if you don't care, why didn't you simply respond with "sure, take 'em out. NBD" | 13:56 |
| dabo | :) | 13:56 |
| uvirtbot | New bug: #716427 in nova "RPC concurrency problem" [Undecided,New] https://launchpad.net/bugs/716427 | 13:56 |
| sandywalsh_ | soren, that's an interesting bug ... thinking | 13:57 |
| sandywalsh_ | soren, where is the calling originating from? | 13:57 |
| sandywalsh_ | soren, if it's from API there shouldn't be a concurrency issue | 13:58 |
| soren | api something. | 13:58 |
| soren | How do you figure that? | 13:58 |
| dabo | soren: is there a significant price for calling connection(True)? | 13:58 |
| sandywalsh_ | soren, each request should be via the reactor, no? | 13:58 |
| dabo | oops instance(True) | 13:58 |
| soren | sandywalsh_: The what now? | 13:58 |
| *** hggdh has quit IRC | 13:59 | |
| sandywalsh_ | soren, eventlet, so multiple clients should be queued in the same thread | 13:59 |
| soren | dabo: Dunno. It creates a new connection to rabbit. Not sure how expensive that is. | 13:59 |
| soren | sandywalsh_: And what is supposed to manage that queue? | 13:59 |
| sandywalsh_ | soren, the api message pump (it *is* eventlet, right?) | 14:00 |
| soren | sandywalsh_: This is eventlet, yes. | 14:00 |
| dabo | soren: yeah, me neither. Unless it proved to be too slow (premature optimization, anyone?), why not go with the safer choice? | 14:00 |
| soren | sandywalsh_: Where is this "pump"? | 14:00 |
| *** hggdh has joined #openstack | 14:00 | |
| sandywalsh_ | soren, sec | 14:00 |
| *** kpepple has quit IRC | 14:01 | |
| *** kpepple has joined #openstack | 14:01 | |
| sandywalsh_ | soren, bin/nova-api wsgi.Server.start() | 14:01 |
| soren | sandywalsh_: You lost me. | 14:02 |
| *** ppetraki has joined #openstack | 14:02 | |
| sandywalsh_ | soren, unless perhaps the wsgi server has multiple threads running? I assumed it was just one thread handling all the http requests | 14:02 |
| soren | sandywalsh_: Oh, god no. | 14:02 |
| soren | IT does eventlety sorts of things. Not sure what that is, but there's *definitely* multiple threads at play. | 14:03 |
| sandywalsh_ | soren, or, multiple threads handling the requests and dumping the request into eventlet for processing | 14:03 |
| soren | sandywalsh_: I don't even understand what that means :) | 14:03 |
| *** adjohn has joined #openstack | 14:03 | |
| *** westmaas has joined #openstack | 14:04 | |
| *** kpepple has quit IRC | 14:05 | |
| *** reldan has joined #openstack | 14:05 | |
| *** kpepple has joined #openstack | 14:05 | |
| soren | sandywalsh_: eventlet's wsgi created a pool of threads to handle requests. | 14:05 |
| soren | EAch of them can use rpc. There's only one connection for rpc.cast => fail. | 14:06 |
| sandywalsh_ | soren, are you seeing my side channel summarization? | 14:08 |
| *** kpepple has quit IRC | 14:09 | |
| sandywalsh_ | soren, and to your point, I'm sure there are situations where using a shared connection is fine. But we'd have to be pretty careful delineating them. | 14:09 |
| wedjat | hi swift people | 14:10 |
| *** kpepple has joined #openstack | 14:10 | |
| *** smaresca has quit IRC | 14:11 | |
| wedjat | could someone please update the documentation : the module python netifaces is now needed in 1.2.0 and it is not yet documented in the installation process... | 14:11 |
| *** ctennis has joined #openstack | 14:11 | |
| *** MarkAtwood has joined #openstack | 14:13 | |
| *** hggdh has quit IRC | 14:15 | |
| *** kpepple has quit IRC | 14:15 | |
| *** kpepple has joined #openstack | 14:16 | |
| *** kpepple has quit IRC | 14:17 | |
| *** kpepple has joined #openstack | 14:17 | |
| *** kpepple has quit IRC | 14:18 | |
| *** hggdh has joined #openstack | 14:19 | |
| *** kpepple has joined #openstack | 14:19 | |
| *** DanEllis has joined #openstack | 14:21 | |
| uvirtbot | New bug: #716441 in openstack-manuals "Search problems - indexed keywords not appearing" [Undecided,New] https://launchpad.net/bugs/716441 | 14:21 |
| *** admboom has joined #openstack | 14:22 | |
| *** admboom has left #openstack | 14:22 | |
| sandywalsh_ | soren, carrot 2.2 has more eventlet support http://groups.google.com/group/celery-users/browse_thread/thread/8cec538a398ff249?pli=1 | 14:23 |
| *** kpepple has quit IRC | 14:24 | |
| *** kpepple has joined #openstack | 14:24 | |
| *** kpepple has quit IRC | 14:28 | |
| soren | sandywalsh_: You mean celery, right? | 14:29 |
| sandywalsh_ | I think we're using carrot in rpc.py ? | 14:29 |
| sandywalsh_ | <checking> | 14:29 |
| *** kpepple has joined #openstack | 14:29 | |
| sandywalsh_ | soren, yup, rpc.py uses carrot | 14:30 |
| *** rpedde has joined #openstack | 14:32 | |
| *** adjohn has quit IRC | 14:33 | |
| *** kpepple has quit IRC | 14:33 | |
| sandywalsh_ | so, is the model now that api should never manipulate the db itself? There should always be a service to do the work? | 14:35 |
| *** kpepple has joined #openstack | 14:35 | |
| sandywalsh_ | I have some very simple operations (CRUD zones) ... and it seems like overkill to make another service for it. | 14:35 |
| sandywalsh_ | soren, sorry, I'm an idiot. Yes, celery 2.2 ... my brain broke. | 14:37 |
| *** kpepple has quit IRC | 14:38 | |
| *** clopez has joined #openstack | 14:39 | |
| *** kpepple has joined #openstack | 14:39 | |
| *** adjohn has joined #openstack | 14:40 | |
| *** adjohn has quit IRC | 14:41 | |
| *** adjohn has joined #openstack | 14:43 | |
| soren | sandywalsh_: Heh :) | 14:44 |
| *** hadrian has joined #openstack | 14:45 | |
| *** kpepple_ has joined #openstack | 14:46 | |
| *** kpepple has quit IRC | 14:46 | |
| *** kpepple_ has quit IRC | 14:47 | |
| *** kpepple has joined #openstack | 14:48 | |
| jaypipes | dabo: because I don't think it's something to waste your time on. That's why I didn't respond that way. | 14:50 |
| *** mattinator has joined #openstack | 14:51 | |
| dabo | jaypipes: I'm not wasting my time on it. sandywalsh_ asked a question, and I agreed that editor-specific directives could be removed. I've spent a lot more time explaining that simple chain of events. | 14:51 |
| jaypipes | dabo: heh, fair enough. | 14:52 |
| *** kpepple has quit IRC | 14:52 | |
| *** kpepple_ has joined #openstack | 14:52 | |
| *** gondoi has joined #openstack | 14:52 | |
| *** BK_man has joined #openstack | 14:55 | |
| *** kpepple_ has quit IRC | 14:56 | |
| *** adjohn has quit IRC | 14:57 | |
| *** kpepple has joined #openstack | 14:57 | |
| sandywalsh_ | jaypipes, thoughts on my db-calls-in-api question above? | 14:58 |
| jaypipes | sandywalsh_: went to breakfast... back-reading now. | 14:58 |
| sandywalsh_ | jaypipes, take your time :) | 14:58 |
| * jaypipes notes he should get into discussions with sandywalsh_ and dabo as grumpy-jay before going to breakfast... | 14:58 | |
| jaypipes | should NOT, should NOT. | 14:58 |
| sandywalsh_ | hahah | 14:59 |
| sandywalsh_ | jaypipes, you'd have to do a *lot* more than that to offend me. | 14:59 |
| westmaas | sandywalsh_: are you talking about the python api? | 14:59 |
| sandywalsh_ | westmaas, kinda. I have to add some zone operations to the os api, but I recall the "nova best practice" is to do all work in a service and not in API | 15:00 |
| jaypipes | sandywalsh_: ya, indeed, but I don't like coming across as a poop. and reading up, I think I came off as a poop. which is, well, poopy. | 15:00 |
| *** vvuksan has joined #openstack | 15:01 | |
| sandywalsh_ | jaypipes, meh | 15:01 |
| *** f4m8 is now known as f4m8_ | 15:01 | |
| westmaas | in a service, via an asynchronous call? | 15:01 |
| jaypipes | sandywalsh_: I was a tad surprised from your email that you were going to do the python-novatools *first*. Is that because you want to get a feel for what it *should* look like to a client first? | 15:01 |
| westmaas | a cast, I guess I should say | 15:01 |
| *** adiantum has joined #openstack | 15:02 | |
| sandywalsh_ | jaypipes, yes, and so I can drive the api easier | 15:02 |
| sandywalsh_ | westmaas, yes, the normal process is api -> queue -> worker | 15:03 |
| *** CloudChris has quit IRC | 15:03 | |
| sandywalsh_ | westmaas, but these are some really trivial CRUD operations. | 15:03 |
| *** krigstask has left #openstack | 15:03 | |
| westmaas | yeah, there are just synchronous calls via queues and asynchronous casts, so I wanted to make sure we were talking about the same thing | 15:04 |
| *** msassak has joined #openstack | 15:04 | |
| jaypipes | sandywalsh_: gotcha. ok, now I'll really read back in the log.... one sec ;) | 15:04 |
| westmaas | sandywalsh_: are these crud options really the only thing that this service would do if you created it? | 15:05 |
| sandywalsh_ | westmaas, that I can think of. pretty sure, yup. | 15:05 |
| sandywalsh_ | westmaas, the scheduler will consume the db entries | 15:06 |
| wedjat | could someone help me with swift 1.2.0 ? i have some weird log when i launch "swift-proxy-server conffile" | 15:06 |
| wedjat | http://pastebin.com/H30Vv6mA | 15:06 |
| *** dubsquared has joined #openstack | 15:08 | |
| *** kpepple has quit IRC | 15:11 | |
| jaypipes | soren, sandywalsh_: there are not multiple threads at play. there are multiple GreenThreads at play ;) | 15:11 |
| *** littleidea has joined #openstack | 15:11 | |
| *** kpepple has joined #openstack | 15:11 | |
| sandywalsh_ | jaypipes, yes, the problem seems to be eventlet doesn't know that carrot is caching the connection. | 15:12 |
| jaypipes | sandywalsh_: as for the zone stuff, I was thinking you'd implement zone CRUD stuff the same way the other admin operations are implemented? | 15:12 |
| jaypipes | sandywalsh_: Ah, I see (re: carrot/eventlet) | 15:13 |
| jaypipes | sandywalsh_: does eventlet have a "greened" carrot? | 15:13 |
| jaypipes | sandywalsh_: or rather, a greened pyamqplib? | 15:13 |
| sandywalsh_ | jaypipes, from what I understand celery uses kombu, which is eventlet friendly | 15:13 |
| soren | jaypipes: Whatever the crap it is, it screws up amqp :) | 15:14 |
| jaypipes | soren: :) | 15:14 |
| *** msassak_ has joined #openstack | 15:14 | |
| jaypipes | sandywalsh_: kombu is simply a fork of carrot, IIRC. If it's greened, then we should try using it instead of carrot directly, like we do now. | 15:14 |
| *** msassak has quit IRC | 15:14 | |
| *** grapex has joined #openstack | 15:15 | |
| jaypipes | soren: which bug is this discussion for? | 15:15 |
| jaypipes | I see... 716427 | 15:15 |
| sandywalsh_ | jaypipes, all the existing admin cmds are compute functions. This doesn't smell like a compute function. | 15:18 |
| jaypipes | sandywalsh_: hmm, good point. | 15:18 |
| wedjat | hi, there are no instructions on installing the last Swift release 1.2.0 from the .tar.gz file. I thought it was by executing "python setup.py install" but seems to be incomplete, right ? | 15:18 |
| jaypipes | sandywalsh_: how are EC2 Regions implemented? | 15:18 |
| sandywalsh_ | jaypipes, oh, good point ... checking | 15:19 |
| jaypipes | wedjat: if it is incomplete, please do report a bug. what do you feel is incomplete (besides translation message catalogs, which we are workin gon...) | 15:19 |
| wedjat | jaypipes: ok thanks. i was just wondering how to execute the last stable version | 15:20 |
| *** jsgotangco has joined #openstack | 15:20 | |
| sandywalsh_ | jaypipes, regions are gflags, but it appears the ec2 api does lots of db work directly | 15:21 |
| *** dendrobates is now known as dendro-afk | 15:23 | |
| jaypipes | wedjat: execute or install? | 15:23 |
| *** adiantum has quit IRC | 15:23 | |
| jaypipes | sandywalsh_: hmm, ok. perhaps they are gflags because they're readonly... I can't remember. | 15:23 |
| jaypipes | sandywalsh_: maybe, then, EC2 SecurityGroups are a better fit? just wondering out loud here... | 15:23 |
| *** adiantum has joined #openstack | 15:23 | |
| jaypipes | sandywalsh_: trying to find a good fitting example that you can say "ok, I'll try to emulate this on the OS API side..." | 15:24 |
| *** fabiand_ has quit IRC | 15:24 | |
| *** dendro-afk is now known as dendrobates | 15:24 | |
| cjreyn | hey all. I'm having issues with Flat networking mode and the metadata service. The redirection on the compute node does not automatically get set up. | 15:27 |
| cjreyn | I added an iptables entry to do this, as defined in the metadata_foward() of klinux_net.py, but still no joy | 15:27 |
| *** msassak_ has quit IRC | 15:27 | |
| cjreyn | is this the only iptables line that provides this redirection, or am I missing something else? | 15:27 |
| *** littleidea has quit IRC | 15:27 | |
| *** henrichrubin has joined #openstack | 15:28 | |
| jaypipes | soren: any idea on above? ^^ for cjreyn | 15:28 |
| * jaypipes needs to learn more about the user-data injection stuff... | 15:28 | |
| cjreyn | its the last (yeah right) of my issues! | 15:28 |
| jaypipes | cjreyn: ah, give it a few more minutes...I'm sure we can find more ;) | 15:29 |
| *** reldan has quit IRC | 15:29 | |
| *** kpepple has quit IRC | 15:29 | |
| henrichrubin | can anyone help with this error? all euca-* commands result in it. i am trying the manual multi-node setup described on the wiki. thanks. http://paste.org/pastebin/view/28696 | 15:29 |
| *** sebastianstadil has joined #openstack | 15:29 | |
| *** kpepple_ has joined #openstack | 15:29 | |
| cjreyn | jaypipes, I hope so, its the only thing holding me back now! | 15:29 |
| *** that__guy has joined #openstack | 15:31 | |
| jaypipes | henrichrubin: no idea, since euca-authorize is hiding the error. could you possibly put a print statement for the raw string returned by putting "print ex" on line 162 of euca-authorize? | 15:31 |
| sandywalsh_ | jaypipes, yes, security groups and a number of other cloud.py operations read and write to the db | 15:31 |
| jaypipes | sandywalsh_: cool. so that could be a good example to work from? | 15:31 |
| soren | cjreyn: It's possible that you need to add the 169.254.169.254 address to one of you hosts's interfaces. | 15:32 |
| sandywalsh_ | jaypipes, it's not the example I'm worried about, I thought "nova architecture" deemed no db operations in api. | 15:32 |
| *** msassak has joined #openstack | 15:32 | |
| sandywalsh_ | jaypipes, I just assumed ec2 never got around to being refactored yet. | 15:33 |
| cjreyn | soren, I thought that address just gets mapped to the CC address? | 15:33 |
| soren | cjreyn: It does. | 15:33 |
| soren | cjreyn: But.. | 15:33 |
| henrichrubin | jaypipes: thanks. this is the error "<unknown>:1:0: syntax error" | 15:33 |
| sandywalsh_ | jaypipes, if it's ok to do these trivial db operations in api ... I'm good to go. | 15:33 |
| that__guy | Hi all, I just upgraded to the our cluster to the launchpad maverick trunk ppa. I'm pretty happy with it, with the exception of euca-describe-instances returning 0 sporadically when it should be returning the user's instance list. | 15:33 |
| sandywalsh_ | jaypipes, I was just wondering if I needed to create yet-another-service for this (which seems like overkill) | 15:33 |
| *** kpepple_ has quit IRC | 15:34 | |
| sandywalsh_ | that__guy, heh ... good handle | 15:34 |
| soren | cjreyn: In order for the packet to ever leave the guest, it needs to either know that it needs to be routed somewhere or it needs to think that there's a host on the LAN that has that IP. Otherwise, it'll never send it (because it doesn't know which MAC to send it to). | 15:34 |
| that__guy | should I file a bug report, or are there plans to take care of this already? | 15:34 |
| soren | cjreyn: It depends on the guest, really. | 15:34 |
| that__guy | sandywalsh_: thanks ;) | 15:34 |
| *** kpepple has joined #openstack | 15:34 | |
| jaypipes | sandywalsh_: understood. | 15:35 |
| *** vvuksan has quit IRC | 15:35 | |
| cjreyn | soren, ok the guest is a UEC image. It gives a url error [[Errno 113] No route to host] | 15:35 |
| soren | cjreyn: Hang on, did you say FlatManager? | 15:35 |
| cjreyn | soren, yes Flatmanager | 15:35 |
| soren | cjreyn: Yeah, that's not going to work. You need a dhcp server. | 15:35 |
| soren | cjreyn: FlatDhcpManager, for instance, will give you that. | 15:36 |
| cjreyn | soren, running where? | 15:36 |
| soren | Um. | 15:36 |
| cjreyn | soren, and why DHCP needed for this? | 15:36 |
| *** reldan has joined #openstack | 15:36 | |
| soren | cjreyn: Because that's how the images get their network config? | 15:36 |
| soren | s/images/instances/ | 15:36 |
| cjreyn | soren, I thought with Flatmamanger they're injected when instantiated | 15:37 |
| soren | Oh, yeah, I guess we still do that. | 15:37 |
| *** sateesh has quit IRC | 15:37 | |
| soren | I haven't tested that in ages. | 15:37 |
| soren | It's a Bad Idea[tm]. | 15:37 |
| sandywalsh_ | jaypipes, heh, so it's ok to do db ops in api? | 15:37 |
| cjreyn | soren, but in the manual ;-) | 15:38 |
| *** kpepple has quit IRC | 15:39 | |
| cjreyn | soren, is there a way to solve this for flat networking, the mapping for the metadata service? | 15:39 |
| *** kpepple has joined #openstack | 15:39 | |
| jaypipes | sandywalsh_: no, I was saying I understand your conundrum ;) | 15:39 |
| sandywalsh_ | jaypipes, haha | 15:39 |
| cjreyn | soren, or is FlatManager considered unusable with this feature? | 15:39 |
| jaypipes | sandywalsh_: but for this specifically, I don't see a problem communicating with the db directly from the API. These zone commands are going to be very infrequently accessed/modified. Thje only other place would be the scheduler worker? | 15:40 |
| soren | cjreyn: No, it should work. | 15:40 |
| sandywalsh_ | jaypipes, exactly | 15:40 |
| soren | cjreyn: Try something like "sudo ip addr add 169.254.169.254 dev eth0" on your host. | 15:41 |
| jaypipes | sandywalsh_: API or Scheduler, that is the question! :) | 15:41 |
| wedjat | <jaypipes> wedjat: execute or install? > just right after install, when i launch "swift-init proxy start" i get an Uncaught exception | 15:41 |
| cjreyn | soren, hold on, in the guest, or underlying compute node? | 15:41 |
| wedjat | jaypipes: the auth-server seems to work | 15:41 |
| sandywalsh_ | jaypipes, infrequent operations in API. Bulk of the reads in Scheduler (to do the work) | 15:41 |
| jaypipes | wedjat: and what is that exceptioN? | 15:41 |
| soren | cjreyn: compute node, for instance. Not the guest. | 15:41 |
| *** rnirmal has joined #openstack | 15:42 | |
| wedjat | jaypipes: http://pastebin.com/H30Vv6mA | 15:42 |
| soren | cjreyn: It's not of much consequence, really. Something just needs to respond to arp requests. | 15:42 |
| wedjat | jaypipes: can't know. I thought it was a missing python module or misplaced libs, but doesn't seem | 15:42 |
| soren | cjreyn: The actual request will get DNAT'ed. | 15:42 |
| jaypipes | wedjat: looking... | 15:43 |
| henrichrubin | jaypipes: thanks. this is the error "<unknown>:1:0: syntax error" | 15:43 |
| jaypipes | wedjat: weird. it says uncaught exception, but doesn't seem to be showing an exception being thrown.... creiht, any ideas? | 15:44 |
| jaypipes | henrichrubin: well that's not very helpful is it? :) any chance you can grab the Nova logs? We need to see what is in the request coming from euca-authorize. | 15:46 |
| cjreyn | soren, works! Heroics! This requires me assigning an IP to an eth that I didn't want to use though. Is there way round this? | 15:46 |
| soren | cjreyn: IT's something we need to fix in Nova. | 15:46 |
| jaypipes | soren: you're my hero. | 15:46 |
| soren | cjreyn: I mean... We'd do the same as you just did manually. | 15:46 |
| soren | cjreyn: You just wouldn't have to worry about it. | 15:47 |
| soren | jaypipes: :) | 15:47 |
| *** hub_cap has joined #openstack | 15:47 | |
| cjreyn | soren, but the dependency is a free eth interface? | 15:47 |
| *** kpepple has quit IRC | 15:48 | |
| wedjat | jaypipes: thanks | 15:48 |
| *** kpepple has joined #openstack | 15:49 | |
| *** kpepple has quit IRC | 15:50 | |
| *** kpepple has joined #openstack | 15:50 | |
| *** kpepple has quit IRC | 15:51 | |
| *** kpepple has joined #openstack | 15:51 | |
| *** kpepple has quit IRC | 15:53 | |
| *** kpepple has joined #openstack | 15:53 | |
| creiht | wedjat: re: netifaces, I thought I got all the doc areas, can you point me to which one doesn't have it? | 15:56 |
| *** photron has quit IRC | 15:57 | |
| *** kpepple has quit IRC | 15:58 | |
| creiht | wedjat: Re: installing, if you are on ubuntu, there are ubuntu ppas available, otherwise the setup.py install should work as well | 15:58 |
| *** kpepple has joined #openstack | 15:58 | |
| *** Hayati has quit IRC | 15:58 | |
| *** photron has joined #openstack | 15:59 | |
| jarrod | i think i need to downgrade to 10.04 from 10.10 | 16:00 |
| creiht | wedjat: Can you post your proxy config file? | 16:00 |
| wedjat | creiht: for netifaces, seems to be missing here : http://swift.openstack.org/debian_package_guide.html | 16:01 |
| wedjat | creiht: are swift 1.2.0 ppas really available ? i didn't found them | 16:01 |
| *** guigui has quit IRC | 16:02 | |
| *** ccustine has joined #openstack | 16:02 | |
| creiht | wedjat: I thought they were... looking | 16:03 |
| *** kpepple has quit IRC | 16:03 | |
| *** kpepple has joined #openstack | 16:03 | |
| creiht | mtaylor: did you ever get around to making swift packages? :) | 16:04 |
| henrichrubin | jaypipes: here is nova-api.conf on the cloud. http://paste.org/pastebin/view/28698 | 16:04 |
| wedjat | creiht: http://pastebin.com/Y1zNrqUV >> my proxy-server conf file | 16:04 |
| wedjat | creiht: thank you :) | 16:04 |
| creiht | wedjat: ahh.. doh about the packaging doc, I'll fix that | 16:05 |
| *** blueadept has joined #openstack | 16:05 | |
| creiht | hrm | 16:05 |
| *** dragondm has joined #openstack | 16:06 | |
| creiht | wedjat: can you run the proxy server manually, and let me know if it shows a different error? | 16:06 |
| creiht | run: swift-proxy-server /etc/swift/proxy-server.conf | 16:06 |
| *** kashyapc has joined #openstack | 16:06 | |
| *** kpepple has quit IRC | 16:07 | |
| *** kpepple has joined #openstack | 16:07 | |
| wedjat | creiht: does the same :( | 16:08 |
| creiht | wedjat: ok I will do some digging | 16:11 |
| *** kpepple has quit IRC | 16:12 | |
| soren | cjreyn: No no, not a free one. | 16:12 |
| soren | cjreyn: We can add that address to any interface. | 16:12 |
| soren | cjreyn: We're not replacing the existing address, just adding another. | 16:12 |
| *** kpepple has joined #openstack | 16:12 | |
| cjreyn | soren, Ok, i just assigned it to a free eth interface. When I add an address as you suggested, it doesn't show in ifconfig | 16:14 |
| wedjat | creiht: thanks. it sounds really weird and i'm wondering if it is not my fault ? i've buildeda debian package quickly using 'checkinstall python setup.py install' and wanted to start it | 16:15 |
| *** kpepple has quit IRC | 16:17 | |
| creiht | wedjat: I just verified that the tar.gz works on my dev instance with setup.py install | 16:17 |
| * creiht goes to look at the traceback a bit closer | 16:17 | |
| *** kpepple has joined #openstack | 16:18 | |
| wedjat | creiht: what is your instance ? i'm on an ubuntu 10.10 | 16:20 |
| creiht | 10.04 | 16:21 |
| wedjat | ok, i will test it right now on an 10.04 | 16:21 |
| openstackhudson | Project nova build #499: SUCCESS in 1 min 33 sec: http://hudson.openstack.org/job/nova/499/ | 16:22 |
| openstackhudson | Tarmac: Get a fresh connection in rpc.cast rather than using a recycled one. | 16:22 |
| *** kpepple_ has joined #openstack | 16:22 | |
| *** kpepple has quit IRC | 16:22 | |
| *** blueadept has quit IRC | 16:24 | |
| *** perra has quit IRC | 16:24 | |
| jarrod | i can never get a freaking server past the pending state | 16:26 |
| jarrod | even using the nova.sh deployment. friggin a | 16:26 |
| *** eikke has quit IRC | 16:28 | |
| *** blueadept has joined #openstack | 16:28 | |
| *** Ryan_Lane|away is now known as Ryan_Lane | 16:29 | |
| wedjat | creiht: seems to work correctly on 10.04. i will recheck something on 10.10 | 16:30 |
| creiht | odd | 16:30 |
| wedjat | hum, cannot explain that. I'll stay on 10.04 for now i think | 16:32 |
| * creiht adds testing on 10.10 to his list | 16:32 | |
| *** bcwaldon has quit IRC | 16:32 | |
| wedjat | thanks for all creiht ;) | 16:32 |
| creiht | wedjat: I wonder if there is some package oddity | 16:33 |
| wedjat | maybe | 16:34 |
| wedjat | i repeated exactly the same commands on the two systems | 16:34 |
| *** bcwaldon has joined #openstack | 16:35 | |
| wedjat | at some time, the 10.04 system complained about not finding python module netifaces, then pastedeploy, but right after installing them it worked | 16:35 |
| creiht | I'm wondering if paste.deploy changed in 10.10 | 16:35 |
| creiht | heh... the package is the same in both | 16:36 |
| *** kpepple_ has quit IRC | 16:36 | |
| *** vvuksan has joined #openstack | 16:37 | |
| *** kpepple has joined #openstack | 16:37 | |
| creiht | wedjat: an interesting experiment on your 10.10 machine would be to edit swift/common/utils.py | 16:38 |
| creiht | search for the capture_stdio function | 16:39 |
| creiht | And insert a return right after the definition of the function to bypass the functionality | 16:39 |
| creiht | then rerun the setup.py install | 16:39 |
| creiht | and try starting again to see if you get any other sort of error | 16:39 |
| *** DanEllis has quit IRC | 16:41 | |
| *** calavera has quit IRC | 16:42 | |
| wedjat | i can comment all the function definition ? | 16:42 |
| jaypipes | sounds like a David Bowie song... "Package Oddity" | 16:43 |
| creiht | Well the function will still be called, it is easier to just insert a return at the beginning | 16:43 |
| *** Hayati has joined #openstack | 16:43 | |
| creiht | thus bypassing that functionality | 16:43 |
| creiht | jaypipes: heh | 16:43 |
| creiht | :) | 16:43 |
| *** berendt has quit IRC | 16:43 | |
| creiht | wedjat: I'm just trying to get a more verbose message as to what is going wrong | 16:44 |
| *** kpepple has quit IRC | 16:44 | |
| *** GasbaKid has joined #openstack | 16:44 | |
| *** kpepple has joined #openstack | 16:44 | |
| *** bcwaldon has quit IRC | 16:44 | |
| *** bcwaldon has joined #openstack | 16:45 | |
| *** kpepple has quit IRC | 16:46 | |
| *** kpepple has joined #openstack | 16:46 | |
| wedjat | creiht: does the same :( | 16:48 |
| creiht | heh | 16:48 |
| creiht | is the error message at the top of the traceback the same? | 16:49 |
| wedjat | yes | 16:49 |
| creiht | UNCAUGHT EXCEPTION? | 16:49 |
| wedjat | yes | 16:49 |
| creiht | hrm | 16:49 |
| creiht | did you run setup.py install again? | 16:50 |
| wedjat | yes | 16:50 |
| creiht | can you paste what that function looks like? | 16:50 |
| wedjat | http://pastebin.com/caZYDjLQ | 16:54 |
| creiht | wedjat: ahh ok... make it look like this: | 16:55 |
| creiht | http://pastebin.com/eNyJqsAp | 16:55 |
| *** kpepple has quit IRC | 16:55 | |
| *** kpepple has joined #openstack | 16:56 | |
| wedjat | ok | 16:57 |
| *** kpepple has quit IRC | 17:00 | |
| *** kpepple has joined #openstack | 17:01 | |
| jarrod | wow i have a lot to learn | 17:03 |
| wedjat | creiht: ok, found the problem | 17:03 |
| *** bcwaldon has quit IRC | 17:04 | |
| wedjat | creiht: the proxy server was complaining for not finding .gz at the right places | 17:04 |
| *** bcwaldon has joined #openstack | 17:04 | |
| wedjat | you know, builder.gz files | 17:04 |
| *** thimble has quit IRC | 17:04 | |
| *** dprince has quit IRC | 17:05 | |
| *** kpepple has quit IRC | 17:05 | |
| wedjat | creiht: it should work correctly now | 17:06 |
| *** kpepple has joined #openstack | 17:06 | |
| creiht | wedjat: ahh interesting | 17:07 |
| creiht | cool | 17:07 |
| iammartian | after going through the bexar install docs on the wiki, is there an actual "startup this nova monster" command? my next step is uploading an image using euca-upload-bundle, and i get this: "socket.error: [Errno 111] Connection refused" | 17:07 |
| iammartian | as if some process that nees to be started has not been | 17:08 |
| *** ramkrsna has quit IRC | 17:09 | |
| wedjat | creiht: thanks :) | 17:09 |
| jarrod | iammartian: i would checkout the nova.sh run category | 17:11 |
| *** stewart__ has quit IRC | 17:12 | |
| iammartian | jarrod: you mean using the all in one script vs. the manual process? | 17:12 |
| jarrod | no, just referencing the script to see what is necessary to start it | 17:12 |
| iammartian | i see, thanks | 17:13 |
| iammartian | thats in contrib, correct? | 17:13 |
| *** kpepple has quit IRC | 17:13 | |
| * jaypipes contemplating walking to the coffee shop for more coffee... hmm frostbite or espresso. which takes the bigger priority here... | 17:14 | |
| *** kpepple has joined #openstack | 17:14 | |
| jaypipes | iammartian: yep | 17:15 |
| *** kpepple has quit IRC | 17:16 | |
| openstackhudson | Yippie, build fixed! | 17:16 |
| openstackhudson | Project swift build #193: FIXED in 27 sec: http://hudson.openstack.org/job/swift/193/ | 17:16 |
| openstackhudson | Tarmac: Updates the debian packaging docs with the python-netifaces dependency | 17:16 |
| *** kpepple has joined #openstack | 17:16 | |
| *** m_3 has quit IRC | 17:19 | |
| wedjat | lol, quick update | 17:19 |
| wedjat | :) | 17:19 |
| *** kashyapc has quit IRC | 17:19 | |
| creiht | :) | 17:20 |
| *** kpepple has quit IRC | 17:21 | |
| *** hadrian has quit IRC | 17:21 | |
| *** kpepple has joined #openstack | 17:21 | |
| *** kpepple has joined #openstack | 17:22 | |
| *** dprince has joined #openstack | 17:22 | |
| *** GasbaKid has quit IRC | 17:24 | |
| *** eikke has joined #openstack | 17:24 | |
| *** GasbaKid has joined #openstack | 17:25 | |
| *** stewart has joined #openstack | 17:25 | |
| *** m_3 has joined #openstack | 17:25 | |
| *** kpepple has quit IRC | 17:27 | |
| *** dprince has quit IRC | 17:27 | |
| *** kpepple has joined #openstack | 17:27 | |
| *** dprince has joined #openstack | 17:29 | |
| *** Hayati has quit IRC | 17:32 | |
| justinsb | Hi - I'd like to implement a feature whereby you can ensure/ask that you get a volume allocated 'near' to your compute node. I'm thinking of sub-zone behavior - same cabinet. I was trying to think of how to do this. | 17:32 |
| justinsb | One way would be to support a hint on the 'create server' call - something like 'near volume-000001' | 17:33 |
| justinsb | The same hints could also be used for e.g. HIPAA compliant hardware, or 'I want a RAID-5 volume' | 17:34 |
| justinsb | I was thinking that tags / metadata would be the way to do this | 17:34 |
| justinsb | What do people think? | 17:34 |
| *** nelson has joined #openstack | 17:34 | |
| nelson | creiht: dumb question: if you don't create the cert_file, don't point to it in the .conf file, how do the clients (e.g. proxy) know not to connect using ssl? | 17:35 |
| aimon | justinb,, the model would work with the URI based naming scheme proposal | 17:35 |
| nelson | creiht: I don't see where anything looks at cert_file except the server side when it sets up the ssl wrapping. | 17:35 |
| justinsb | aimon: is that a blueprint? Is there a link I can read? | 17:36 |
| aimon | but 'near' is very very hard thing to reliably determine | 17:36 |
| aimon | ther is a thread in maillist | 17:36 |
| aimon | about naming | 17:36 |
| justinsb | aimon: Thanks, I'll find the thread | 17:37 |
| aimon | justinb | 17:37 |
| aimon | here are topics: | 17:37 |
| aimon | two avtually same idea mentioned | 17:37 |
| aimon | "[Openstack] Proposed OpenStack Service Requirements" | 17:37 |
| aimon | "[Openstack] Multi Clusters in a Region ..." | 17:38 |
| aimon | both of those propose URI naming scheme | 17:38 |
| justinsb | aimon: thanks - digging them out now :-) | 17:39 |
| *** maplebed has joined #openstack | 17:39 | |
| *** kpepple has quit IRC | 17:44 | |
| *** kbringard has joined #openstack | 17:44 | |
| *** kpepple has joined #openstack | 17:44 | |
| jaypipes | justinsb: yeah, definitely want to catch up on those ML threads and talk with sandywalsh_ and eday, too :) | 17:45 |
| kbringard | I think I saw a blueprint for this... but I was wondering if there was any status on Open vSwitch integration, or if anyone has had any luck using the brutils compatability stuff to make it work | 17:46 |
| *** kpepple has quit IRC | 17:49 | |
| *** kpepple has joined #openstack | 17:49 | |
| *** joearnold has joined #openstack | 17:50 | |
| *** berendt has joined #openstack | 17:51 | |
| *** kpepple has quit IRC | 17:51 | |
| henrichrubin | jaypipes: here is nova-api.conf on the cloud. http://paste.org/pastebin/view/28698 | 17:52 |
| *** kpepple has joined #openstack | 17:52 | |
| jaypipes | henrichrubin: sorry for delay... looking now. thx for nagging :) | 17:53 |
| henrichrubin | jaypipes: no problem. i'm trying to restore another server in the meantime using mondoresuce | 17:53 |
| *** Nacx has quit IRC | 17:54 | |
| jaypipes | k | 17:54 |
| comstud | hrm | 17:54 |
| comstud | anyone else noticing the openstack mailing list being ~30 minutes delayed ? | 17:55 |
| comstud | :-/ | 17:55 |
| comstud | tho a number of things just flushed out at once | 17:55 |
| *** eikke has quit IRC | 17:56 | |
| kbringard | comstud: I've been noticing a delay in delivering stuff to gmail in general... if your email address is a gmail addy, perhaps that's the issue? | 17:57 |
| comstud | hmm, actually, yeah, i'm subscribed under my google apps address | 17:58 |
| comstud | not my rackspace one | 17:58 |
| kbringard | no idea if that's the issue, it's been intermittent, but like I've had people send me stuff from other gmail addresses, and it took like, 45 mins to an hour to show up | 17:59 |
| jaypipes | comstud: pretty typical | 17:59 |
| jaypipes | comstud: LP mailing lists are not known to be speedy... | 17:59 |
| comstud | bummer | 18:00 |
| jaypipes | comstud: though they seem to go in spurts... must be some maintenance job that periodically kicks in. | 18:00 |
| jaypipes | henrichrubin: OK, I'm stumped. the logs and the exception from Boto aren't useful in the slightest. It's not showing a) what was in the request, and b) what was in the response. :( | 18:01 |
| jaypipes | <unknown>:1:0 syntax error is less than helpful. | 18:01 |
| *** that__guy has quit IRC | 18:02 | |
| openstackhudson | Project nova build #500: SUCCESS in 1 min 37 sec: http://hudson.openstack.org/job/nova/500/ | 18:02 |
| openstackhudson | Tarmac: Use a NullPool for sqlite connections. | 18:02 |
| henrichrubin | jaypipes: i know. it isn't making it to nova-api? maybe something w/ boto? i restarted all the services. the same error occurs if i run the same command on the cloud node too. | 18:03 |
| henrichrubin | i presume this is right ... Starting /usr/bin/nova-api on 0.0.0.0:8773 | 18:03 |
| jaypipes | henrichrubin: https://bugs.launchpad.net/ubuntu/+source/eucalyptus/+bug/480048 | 18:04 |
| uvirtbot | Launchpad bug 480048 in eucalyptus "euca-register fails with syntax error (dup-of: 503180)" [Low,Fix released] | 18:04 |
| eday | henrichrubin: yes, 0.0.0.0 == INADDR_ANY | 18:04 |
| uvirtbot | Launchpad bug 503180 in eucalyptus "[SRU] eucalyptus-cloud doesn't reply to requests (eucalyptus doesn't work after reboot or services restart issues due to upstart networking behavior)" [High,Fix released] | 18:04 |
| jarrod | hmm | 18:06 |
| jaypipes | henrichrubin: those bugs seem to suggest you need to restart the cloud controller... | 18:06 |
| jaypipes | henrichrubin: also, what version of boto are you using? | 18:07 |
| jaypipes | henrichrubin: apparently you need python-boto 1.9b-4. (see here: http://open.eucalyptus.com/forum/failed-parse-error-message-aws-unknown10-syntax-error) | 18:07 |
| jarrod | yea that was a pain | 18:07 |
| *** GasbaKid has quit IRC | 18:09 | |
| justinsb | I've read the links aimon sent me; I like both proposals (hierarchical zones and hierarchical URI-based naming). I don't see any proposal to actually _use_ the capabilities/metadata though (yet?) | 18:09 |
| justinsb | .I'm thinking that we could use the metadata tag support for machines to say 'openstack:near=vol-000001' or something like that | 18:09 |
| jaypipes | justinsb: yup. | 18:09 |
| *** dprince has quit IRC | 18:09 | |
| justinsb | jaypipes: Do you think I should open a blueprint for this? I don't see an existing one | 18:10 |
| henrichrubin | jaypipes: root@island108:/var/log/nova# python | 18:11 |
| henrichrubin | >>> import boto | 18:11 |
| henrichrubin | >>> boto.Version | 18:11 |
| henrichrubin | '1.9b' | 18:11 |
| jaypipes | justinsb: though I think that would be a function of the request packet sent to the volume controller. In other words, the set of required (or desired) attributes that the volume should have is sent to from the scheduler to the volume controller. One of those attributes can be {'near': 'openstack-volume://zone1.volume.com/'} etc.. | 18:11 |
| henrichrubin | jaypipes: i restarted all the services on the cloud. i'll try again | 18:12 |
| *** dprince has joined #openstack | 18:12 | |
| jaypipes | henrichrubin: I think you may need to install 1.9b4. | 18:12 |
| aimon | doh! | 18:12 |
| aimon | ^^ deym too many chat windows | 18:12 |
| uvirtbot | aimon: Error: "^" is not a valid command. | 18:12 |
| henrichrubin | jaypipes: i'll try that. i wonder why that isn't the default on 10.04. | 18:13 |
| jaypipes | justinsb: not sure about existing blueprint or another one. I think perhaps safest bet is to add notes to the existing multi-cluster blueprint and, if needed in the future, break it out into a separate one for clients of the new zone API commands? | 18:13 |
| jaypipes | justinsb: with the various Volume managers being those "clients" of the zone functionality... | 18:13 |
| justinsb | jaypipes: I'm not entirely sure yet how the supported tags will bubble down through the various services. A request for a RAID-5 volume would definitely hit the volume service; a request for a compute node near a volume _might_ not | 18:14 |
| justinsb | jaypipes: I think it's sort of orthogonal to the hierarchical zones & hierarchical naming | 18:14 |
| jaypipes | justinsb: it would definitely hit the volume service, of course, but only after going through the scheduler service. | 18:14 |
| jaypipes | justinsb: agreed, mostly :) | 18:14 |
| justinsb | jaypipes: It's more about consuming the capabilities that those blueprints provide | 18:15 |
| jaypipes | justinsb: sure, agreed. | 18:15 |
| justinsb | jaypipes: I might propose a new blueprint, not least because it | 18:15 |
| justinsb | jaypipes: it'll be contentious in terms of how the client provides the metadata | 18:15 |
| jaypipes | justinsb: I'd support you on that. feel free to propose away! :) discussing these things early and often is the best approach. | 18:15 |
| *** photron has quit IRC | 18:16 | |
| henrichrubin | jaypipes: i only see a 1.9b version. any idea where i can find 1.9b4 for ubuntu. http://code.google.com/p/boto/downloads/list | 18:18 |
| *** dendrobates is now known as dendro-afk | 18:19 | |
| *** dendro-afk is now known as dendrobates | 18:19 | |
| berendt | is it possible to define new INSTANCE_TYPES on the fly without modifying compute/instance_types.py? | 18:21 |
| jaypipes | henrichrubin: no. try installing the 2.0b3 tarball then :) | 18:22 |
| jaypipes | henrichrubin: I'm only going from the various website links I've been reading about your specific <unknown>:1:0 error... | 18:22 |
| berendt | jaypipes: is boto 2.0b3 working now? i had problems with this version a few weeks ago... | 18:23 |
| wedjat | should swauth work with swift 1.2.0 ? | 18:23 |
| *** kashyapc has joined #openstack | 18:23 | |
| jaypipes | berendt: gah, I don't know... just been trying to find something that works for henrichrubin :( | 18:23 |
| berendt | wedjat: yes should work | 18:23 |
| berendt | jaypipes: ok :) | 18:23 |
| jaypipes | berendt, henrichrubin: I suspect this issue has nothing to do with Nova at all, and everything to do with Boto, but it's tough to tell from euca/boto's limited error output. | 18:24 |
| jaypipes | berendt: I think you'll agree that <unknown>:1:0 syntax error isn't particularly enlightening ;) | 18:24 |
| berendt | jaypipes: ;) | 18:26 |
| berendt | i try spawning an instance from an image located on a local running object store | 18:27 |
| berendt | (nova.exception): TRACE: Command: /usr/bin/curl --fail --silent http://ares:3333/_images/ami-04404gn9/image -H 'Date: Thu, 10 Feb 2011 18:24:46 GMT' -H 'Authorization: AWS f0780055-ddb1-4216-9235-ad920f585603:ttylinux-uec:cZpOI4+imV7ded/N+9jcj4Uceto=' -o /var/lib/nova/instances/_base/ami-04404gn9 | 18:27 |
| berendt | any hints...? | 18:27 |
| jaypipes | berendt: and what is the exception? | 18:27 |
| wedjat | thanks berendt | 18:28 |
| * jaypipes thought swauth was going away now that auth-server is fully in swift? | 18:28 | |
| wedjat | isn't it devauth ? | 18:35 |
| *** dirakx1 has joined #openstack | 18:35 | |
| *** dirakx has quit IRC | 18:35 | |
| *** ianweller has quit IRC | 18:37 | |
| vvuksan | berendt: is the image available ? | 18:37 |
| openstackhudson | Project nova build #501: SUCCESS in 1 min 33 sec: http://hudson.openstack.org/job/nova/501/ | 18:37 |
| openstackhudson | Tarmac: Add forwarding rules for floating IPs to the OUTPUT chain on the network node in addition to the PREROUTING chain. | 18:37 |
| berendt | vvuksan: yes | 18:38 |
| vvuksan | berendt: ie. it's not untarring ? | 18:38 |
| berendt | ares:/var/lib/nova/images # find | 18:39 |
| berendt | . | 18:39 |
| berendt | ./ami-qqu0aarf | 18:39 |
| berendt | ./ami-04404gn9/info.json | 18:39 |
| berendt | ./ami-x3rgltwg | 18:39 |
| berendt | ./ami-x3rgltwg/image | 18:39 |
| berendt | ./ami-x3rgltwg/info.json | 18:39 |
| vvuksan | something happened | 18:40 |
| vvuksan | since the image file is migging | 18:40 |
| vvuksan | for 04404gn9 | 18:40 |
| vvuksan | missing | 18:40 |
| berendt | ./ami-04404gn9/image..? | 18:40 |
| vvuksan | yep | 18:41 |
| jaypipes | dubsquared: you just want the world, don't you? ;) | 18:41 |
| berendt | vvuksan: that's the image, or not..? | 18:41 |
| jaypipes | dubsquared: I'll have to ask soren about the PPAs. Not sure what has changed in the recent weeks about those. | 18:41 |
| vvuksan | berendt: correct | 18:42 |
| jaypipes | berendt: hey! that find . output above is different from the one you showed me in the PM! | 18:43 |
| berendt | jaypipes: open your eyes.. :) it's the same | 18:45 |
| wedjat | what does mean : "412 Precondition failed" in Swift ? | 18:46 |
| berendt | vvuksan: so that image is on the local filesystem.. you wrote the image file ami-04404gn9 is missing | 18:46 |
| jaypipes | berendt: *cough* no it's not :) | 18:47 |
| jaypipes | <berendt> ares:/var/lib/nova/images # find . | 18:47 |
| jaypipes | <berendt> . | 18:47 |
| vvuksan | berendt: ami-04404gn9/image is missing | 18:48 |
| jaypipes | berendt: something has gone awry with the kernel image (ami-04404gn9) somehow... not sure what, but something's amiss. | 18:48 |
| *** ianweller_ has joined #openstack | 18:48 | |
| berendt | you're right.. | 18:49 |
| berendt | but i don't know how you could see that in the find output.. :) | 18:49 |
| berendt | lrwxrwxrwx 1 root root 75 Feb 10 18:47 image -> /tmp/uec-publish-tarball.4slMnV/ttylinux-uec-amd64-12.1_2.6.35-22_1-vmlinuz | 18:49 |
| jaypipes | berendt: could that ./ami-04404gn9/info.json have happened after you did that symlink? | 18:49 |
| berendt | ls: cannot access /tmp/uec-publish-tarball.4slMnV/ttylinux-uec-amd64-12.1_2.6.35-22_1-vmlinuz: No such file or directory | 18:49 |
| berendt | jaypipes: i just called uec-publish-tarball | 18:50 |
| jaypipes | ok, that's officially weird. | 18:50 |
| jaypipes | it is there. no it's not. yes it is! no it's not! | 18:50 |
| berendt | symlink to tmp.. and there are files | 18:51 |
| jaypipes | berendt: does nova-objectstore have a log? not sure... | 18:51 |
| berendt | i'll remote the images first and upload them again.. | 18:51 |
| berendt | remove | 18:51 |
| jaypipes | k | 18:52 |
| jaypipes | berendt: all looking good in PM. | 18:53 |
| jarrod | any reason the ubuntu 10x uec wouldnt have ssh open | 18:53 |
| jarrod | even tho the port was allowed | 18:54 |
| wedjat | does someone already tested the swauth configuration that is on the official documentation of swift ? i can make swift work very well with devauth, but never with swauth | 18:55 |
| cjreyn | hey guys. I'm having a hard time convincing my boss that we should be using Openstack instead of Euca. He'd like to know who else is using Openstack. Can you point me to any web links illustrating success stories with Openstack? | 18:55 |
| cjreyn | and yes, he is a divvy | 18:55 |
| jarrod | General error mounting filesystems. <-- that may be why heh | 18:55 |
| *** DigitalFlux has joined #openstack | 18:56 | |
| *** reldan has quit IRC | 18:57 | |
| *** dfg has joined #openstack | 18:57 | |
| *** DigitalFlux has quit IRC | 18:58 | |
| creiht | nelson: ssl shouldn't be enabled unless you set the cert_file in the conf | 18:59 |
| nelson | ja, I'm just trying to figure out how the HTTP client figures out that it's talking to an SSL HTTP server. | 19:00 |
| nelson | cjreyn: Looks like Wikipedia is going to be using Swift. | 19:01 |
| *** dprince_ has joined #openstack | 19:01 | |
| *** dprince has quit IRC | 19:01 | |
| cjreyn | nelson, citation!? | 19:01 |
| nelson | http://wikitech.wikimedia.org/view/Media_server/Distributed_File_Storage_choices | 19:01 |
| creiht | wedjat: yes, swauth should work fine in swift 1.2 | 19:02 |
| creiht | jaypipes: we are going to be depricating devauth in preference to swauth | 19:02 |
| *** fabiand_ has joined #openstack | 19:02 | |
| cjreyn | nelson, thanks. Anything specific to Nova would be great also as this is our use case | 19:03 |
| redbo | how's that not original research? | 19:03 |
| creiht | nelson: usually if the url starts with https :) | 19:03 |
| creiht | nelson: and woot about using it at wikipedia | 19:04 |
| nelson | cjreyn: I don't know if anything has been published about Wikipedia using Nova. | 19:04 |
| wedjat | creiht: with the simple conf on the website, the command swauth-add-user hangs | 19:05 |
| nelson | redbo: pthththththththt. | 19:05 |
| wedjat | creiht: it works well with devauth howver | 19:05 |
| cjreyn | nelson, ok | 19:05 |
| jarrod | this is exciting -- i am starting to see how this works more | 19:05 |
| creiht | wedjat: I use swauth on my dev config | 19:05 |
| creiht | cjreyn: for swift, Rackspace is currently using it for cloudfiles, and there are at least 2 other public services that are using it | 19:06 |
| nelson | creiht: ahhhh! But then does the auth server return an http: url? | 19:06 |
| *** ppetraki has quit IRC | 19:06 | |
| creiht | nelson: yeah you have to set your storage url endpoint to be either http or https | 19:06 |
| cjreyn | creiht, the public services are? | 19:06 |
| nelson | creiht: ahhhh, so they're not coordinated automatically. Good to know. Should be in the documentation. | 19:07 |
| creiht | nelson: yeah | 19:07 |
| creiht | cjreyn: http://www.rackspacecloud.com/cloud_hosting_products/files/ | 19:07 |
| creiht | cjreyn: http://www.internap.com/cloud-storage/our-solution/ | 19:08 |
| creiht | cjreyn: The other one, I'm not sure if it is really public knowledge | 19:08 |
| *** dprince_ has quit IRC | 19:08 | |
| creiht | cjreyn: NASA is also deploying it along side nova | 19:08 |
| wedjat | creiht: ok. swift sends me back no error so i don't understand why the http connection is waiting | 19:09 |
| justinsb | Wrote the request-capabilities blueprint: https://blueprints.launchpad.net/nova/+spec/use-metadata-tags-for-capabilities. Should I send to the openstack mailing list? | 19:09 |
| cjreyn | creiht, how big is nasa's nova cloud? | 19:10 |
| *** photron has joined #openstack | 19:10 | |
| *** rcc has quit IRC | 19:10 | |
| creiht | cjreyn: Not entirely sure | 19:10 |
| creiht | Their first deployments I think have mostly been for testing | 19:10 |
| jarrod | probably about 10 servers | 19:10 |
| jarrod | heh ;) | 19:10 |
| henrichrubin | jaypipes: same problem with boto 2.03b | 19:10 |
| *** dprince has joined #openstack | 19:10 | |
| jarrod | dont use 2.03b | 19:10 |
| jarrod | use 1.9 with the proper patch | 19:11 |
| jarrod | atleast that worked for me | 19:11 |
| creiht | cjreyn: Though I am supposed to talk with devcamcar today about deployments :) | 19:11 |
| uvirtbot | New bug: #716601 in nova "uploaded files to ObjectStore are not available (image symlinks only points to non existing files)" [Undecided,New] https://launchpad.net/bugs/716601 | 19:12 |
| jarrod | nova-objectstore is _not_ swift, right? | 19:12 |
| creiht | jaypipes: correct | 19:12 |
| creiht | erm | 19:12 |
| creiht | jarrod: | 19:12 |
| cjreyn | creiht, the more examples I can give him the better. Unfortunately, he is completely out of touch on a technical level | 19:12 |
| *** msassak has quit IRC | 19:12 | |
| jarrod | creiht: tab problems ? ;) | 19:12 |
| creiht | hehe.. yeah completion fail | 19:13 |
| *** BK_man has quit IRC | 19:13 | |
| creiht | cjreyn: Those above are all for storage | 19:13 |
| cjreyn | creiht, yeah its nova compute I have to reason | 19:13 |
| jaypipes | creiht: thx. :) | 19:13 |
| creiht | jarrod: as I understand it, nova-objectstore is mainly for dev | 19:14 |
| jaypipes | henrichrubin: grr. ok, then please file a bug report with all the relevant information. we'll see what we can find out. | 19:14 |
| creiht | lol, did I just accidently say that jaypipes was correct? :) | 19:14 |
| jaypipes | creiht: never. | 19:14 |
| jaypipes | berendt: hey, I was thinking, perhaps the issue is because uec-publish-tarball is creating the images in a temporary directory that it unlinks as soon as it finishes the bundling? | 19:15 |
| *** dprince has quit IRC | 19:18 | |
| jaypipes | wow, CSCO and AKAM stocks both getting beaten like a rented mule today... | 19:19 |
| *** kashyapc has quit IRC | 19:19 | |
| *** mray has joined #openstack | 19:20 | |
| *** aryan has quit IRC | 19:22 | |
| *** aryan has joined #openstack | 19:23 | |
| *** reldan has joined #openstack | 19:24 | |
| *** fabiand_ has quit IRC | 19:24 | |
| *** ianweller_ is now known as ianweller | 19:26 | |
| *** ianweller has joined #openstack | 19:26 | |
| *** fabiand_ has joined #openstack | 19:27 | |
| *** kpepple has quit IRC | 19:29 | |
| wedjat | haha, i've found the problem with swauth-add-user :) | 19:31 |
| jaypipes | wedjat: was it creiht? :) | 19:32 |
| *** burris has quit IRC | 19:33 | |
| dubsquared | jaypipes: keep up that tone, and I'll go eday on your ass | 19:34 |
| jaypipes | ha! | 19:34 |
| jaypipes | dubsquared: :) nice. | 19:34 |
| jaypipes | dubsquared: on a serious note, you might have to wait until soren gets back for an answer on the PPA question... | 19:35 |
| dubsquared | jaypipes eday: i love intelligent banter, you guys are making the list active and i love it | 19:35 |
| dubsquared | cool | 19:35 |
| creiht | lol | 19:40 |
| creiht | wedjat: what was the issue? sorry busy day today | 19:40 |
| *** aryan has quit IRC | 19:48 | |
| sirp_ | soren: i'm wondering if pool.NullPool can be removed once SA's pool_recycle is fixed with https://bugs.launchpad.net/nova/+bug/706405 (merge proposed) | 19:49 |
| uvirtbot | Launchpad bug 706405 in nova "MySQL Gone Away Generated" [Medium,In progress] | 19:49 |
| *** fabiand_ has quit IRC | 19:50 | |
| *** sebastianstadil has quit IRC | 19:51 | |
| vishy | sirp_: i think those are unrelated problems | 19:51 |
| *** msassak has joined #openstack | 19:52 | |
| vishy | sirp_: the NullPool is specific to sqlite locking issues | 19:52 |
| vishy | sirp_: whereas the other is mysql connection timeouts | 19:52 |
| nelson | creiht: what do you think of the possibility of turning the proxy-server into a filter whose last step is only called to handle errors? | 19:53 |
| sirp_ | vishy: oh i see, i thought sqlite was timing out in a simlar way, nm then :) | 19:53 |
| *** ctennis has quit IRC | 19:53 | |
| nelson | cuz right now I'm having to do a gratuitious HEAD on the file to see if it already exists. If it doesn't exist I need to fetch it from the existing media store or create it by deriving it from an original. | 19:54 |
| nelson | the standard pipeline in proxy-server.conf should look like this: | 19:54 |
| nelson | pipeline = healthcheck cache auth proxy-server report-error | 19:54 |
| nelson | The proxy-server would add a header which reported the status that would be returned, so the "report-error" app would trivially return that error. | 19:56 |
| *** kpepple has joined #openstack | 19:56 | |
| *** photron has quit IRC | 20:00 | |
| *** snelly has joined #openstack | 20:01 | |
| notmyname | nelson: you could add middleware to the pipeline to add that extra header | 20:02 |
| *** kpepple has quit IRC | 20:02 | |
| nelson | No, I want to stop proxy-server from returning an error. | 20:03 |
| *** brd_from_italy has joined #openstack | 20:03 | |
| *** kpepple has joined #openstack | 20:03 | |
| *** mdomsch has joined #openstack | 20:03 | |
| nelson | basically, "If file not on our cluster, fetch it from elsewhere." | 20:03 |
| *** Ryan_Lane is now known as Ryan_Lane|away | 20:03 | |
| nelson | because the transition that we want to make is not a flag day, but instead a step-by-step "try this, does it work? Great, now try that," keeping the old service in place at every step. | 20:04 |
| nelson | People react badly to Wikipedia being down. It's like having their encyclopedia not work. | 20:05 |
| notmyname | your "report-error" middleware could do that. really, it would be something after the proxy server in the pipeline. if the response from a GET/HEAD to the proxy is a 404, then you could fetch from the old system | 20:05 |
| *** aryan has joined #openstack | 20:05 | |
| nelson | notmyname: exactly, but right now proxy-server returns the 404. | 20:06 |
| notmyname | but the client only sees it because the proxy-server is last in the pipeline and nothing in the pipeline is modifying it | 20:06 |
| nelson | I've got rewrite middleware on the front-end as well, so it REALLY looks as if we can point the cache servers directly at the Swift proxy server. | 20:07 |
| nelson | the client sees it because any entity in a pipeline can return a result and then not call the rest of the pipeline. | 20:07 |
| notmyname | actually, the "fetch-from-old" middleware doesn't have to be last, just in the pipeline somewhere. perhaps the outermost one (the first) | 20:07 |
| nelson | if it's the outermost, then I need to do recursive HEAD request on the proxy. | 20:08 |
| notmyname | pipeline = catch-error cache auth fetch-from-old-on-error proxy-server | 20:08 |
| nelson | much simpler to presume that the file exists, and fix it before the proxy-server returns an error. | 20:09 |
| notmyname | do the requests as normal (no extra HEAD). then the middleware | 20:09 |
| notmyname | nelson: I think we are sayting the same thing | 20:09 |
| notmyname | but I'm likely misunderstanding | 20:09 |
| nelson | but once the proxy-server gets control, it does one of two things: it either returns the file, or it returns an HTTP error. | 20:10 |
| nelson | I'm suggesting that instead of returning the error, it be written like a filter, and then it calls the next-in-pipeline to handle the error. | 20:10 |
| notmyname | for example, the catch-errors middleware shows how to capture the result and modify it | 20:10 |
| nelson | where is catch-error? | 20:10 |
| notmyname | catch-errors detects errors and returns a generic 503 | 20:11 |
| notmyname | swift/common/middleware | 20:11 |
| notmyname | same place as all the other middleware included with swift | 20:11 |
| nelson | In Bexar or the previous version? | 20:12 |
| notmyname | both | 20:12 |
| nelson | not in my copy. | 20:12 |
| nelson | root@alsted:/usr/share/pyshared/swift# ls common/middleware/ | 20:12 |
| nelson | acl.py auth.py healthcheck.py memcache.py ratelimit.py | 20:12 |
| notmyname | swift/common/middleware/catch_errors.pyt? | 20:12 |
| notmyname | catch-errors was added a long time ago. I thought it was there from the beginning | 20:13 |
| nelson | Seemingly not. | 20:13 |
| notmyname | what version are you running? | 20:14 |
| notmyname | $ ls *.py | 20:14 |
| notmyname | __init__.pyauth.pycname_lookup.pyhealthcheck.pyratelimit.pyswift3.py | 20:14 |
| notmyname | acl.pycatch_errors.pydomain_remap.pymemcache.pyswauth.py | 20:14 |
| nelson | The one the documentation told me to install. | 20:14 |
| sandywalsh_ | jaypipes, the other reason for doing nova-tools first is we'll need an OS API client library for zone-zone communication. | 20:14 |
| nelson | :) | 20:14 |
| *** shawn has joined #openstack | 20:15 | |
| notmyname | nelson: http://bazaar.launchpad.net/~hudson-openstack/swift/1.2/files/head:/swift/common/middleware/ | 20:15 |
| notmyname | that's bexar | 20:15 |
| notmyname | that's ??? | 20:15 |
| notmyname | s/'// | 20:15 |
| nelson | Pretty sure we're on austin. | 20:15 |
| notmyname | err. apparently I can't think. keep the ' | 20:16 |
| notmyname | nelson: http://bazaar.launchpad.net/~hudson-openstack/swift/1.1/files/head:/swift/common/middleware/ | 20:16 |
| nelson | http://swift.openstack.org/howto_installmultinode.html <=== we're running what this gets you | 20:16 |
| uvirtbot | New bug: #716652 in glance "MySQL Gone Away Error" [Medium,In progress] https://launchpad.net/bugs/716652 | 20:16 |
| *** dprince has joined #openstack | 20:18 | |
| nelson | notmyname: ahhhh, I see. the proxy-server now raises exceptions rather than emitting HTTP error messages, relying on catch-errors to finally emit the error message. | 20:19 |
| nelson | That seems a fairly clean solution to me. | 20:19 |
| jaypipes | sandywalsh_: true nuf. | 20:19 |
| notmyname | catch-errors is there to prevent errors from getting to the client if some uncaught exception is raised | 20:20 |
| *** dprince has quit IRC | 20:20 | |
| nelson | Hrm. You mean that proxy-server doesn't raise exceptions? | 20:21 |
| *** Ryan_Lane|away is now known as Ryan_Lane | 20:21 | |
| nelson | Hrm. You mean that proxy-server will usually return an HTTP error status rather than raising an exception? | 20:21 |
| notmyname | yes | 20:21 |
| nelson | oh, well, crap, that doesn't do what I need. | 20:22 |
| notmyname | but if you look at line 37 in catch_errors.py, you see that the result could easily be assigned a variable and inspected for response code. return unmodified if it's ok. if it needs to be redirected to your old store, do it however you want (3xx redirect or fetch inline) | 20:23 |
| *** joearnold has quit IRC | 20:24 | |
| nelson | Does the pipeline return a status and iterable which returns the contents? | 20:25 |
| nelson | Cuz if that's the case, then I can just call into the app, inspect its status, and return a different Response back up. | 20:26 |
| * nelson goes off to fiddle. | 20:26 | |
| *** jmckenty has joined #openstack | 20:27 | |
| notmyname | nelson: yes. that's exactly what the middleware is designed to do | 20:32 |
| notmyname | the response has the headers and the content (or an iterable that generates the content) that is given to the wsgi app (eventlet in our case) | 20:32 |
| nelson | I thought from the pipeline name that it was more like a Unix pipe: one-way. | 20:32 |
| notmyname | no. it's from paste.deploy | 20:32 |
| notmyname | it's like a bunch of nested dolls | 20:33 |
| asksol | sandywalsh_: right, carrot is basically just crap :) | 20:33 |
| nelson | yep. okay, now I'm much more clear on the concept. KEWL!! | 20:33 |
| asksol | due to my limited understanding of amqp at the time | 20:33 |
| notmyname | nelson: think of it like inception: go down through each layer and then back up | 20:33 |
| notmyname | it just doesn't get slower as you get deeper ;-) | 20:33 |
| asksol | kombu is restoring my conscience | 20:34 |
| sandywalsh_ | asksol, there's a movement afoot to switch to celery/kombu | 20:34 |
| asksol | I believe it also fixes some of your annoyances, like consistent error handling | 20:34 |
| sandywalsh_ | asksol, jaypipes was discussing it | 20:34 |
| asksol | at least from what I read from your code | 20:34 |
| *** miclorb has joined #openstack | 20:35 | |
| sandywalsh_ | asksol, yes, well the issue soren ran into today was getting better compliance with eventlet | 20:35 |
| sandywalsh_ | asksol, the connections were getting reused and puking with concurrent requests | 20:35 |
| *** clopez has quit IRC | 20:35 | |
| asksol | sandywalsh_: yeah, and as I told Jay, I'm available if you need any help to make that happen | 20:36 |
| asksol | sandywalsh_: that was a bug in your application I hope? :) | 20:36 |
| sandywalsh_ | asksol, yes | 20:37 |
| asksol | phew | 20:37 |
| sandywalsh_ | asksol, the whole topic is heating up and we'll be doing more queue related stuff soon. I'm sure we'll be tapping you with questions. | 20:37 |
| asksol | In Celery I use the kombu connection pool to avoid that | 20:37 |
| sandywalsh_ | asksol, so thanks for the offer of help! | 20:37 |
| asksol | but still, I have not yet mapped out the best solution to deal with this | 20:37 |
| sandywalsh_ | it's timely, load testing is slated for this iteration of openstack. So it'll get a nice workout. | 20:38 |
| asksol | at least the cosmetics, having to acquire a connection and a channl everywhere is kinda annoying | 20:38 |
| asksol | really looking forward to your feedback on that then | 20:39 |
| sandywalsh_ | we moved from Twisted to Eventlet because the code was cleaner. But the magic under the hood breaks at times | 20:39 |
| sandywalsh_ | stay tuned ... we'll have lots :) | 20:39 |
| asksol | yeah, it's kind of hard to know exactly what is going on :/ | 20:39 |
| sandywalsh_ | heh, tell me about it :) | 20:40 |
| sandywalsh_ | gotta run ... will be in touch. cheers! | 20:40 |
| asksol | later | 20:40 |
| asksol | cheers | 20:41 |
| *** BK_man has joined #openstack | 20:41 | |
| *** Ryan_Lane_ has joined #openstack | 20:46 | |
| jaypipes | sirp_: if you get a chance, I pushed a test case for bug 704854 up here: https://code.launchpad.net/~jaypipes/glance/bug704854. Feel free to pull the branch and let me know if run_tests.sh succeeds properly? Thanks! | 20:47 |
| uvirtbot | Launchpad bug 704854 in glance "Add an image to Glance with wrong type and underlying exception gets eaten" [Medium,In progress] https://launchpad.net/bugs/704854 | 20:47 |
| nelson | notmyname: is that something that only works in 1.2?' | 20:47 |
| *** berendt has quit IRC | 20:48 | |
| nelson | notmyname: it looks like calling the app is only returning an iterator. | 20:48 |
| nelson | (in 1.1) | 20:48 |
| *** Ryan_Lane has quit IRC | 20:49 | |
| *** jmckenty has left #openstack | 20:50 | |
| *** sandywalsh_ has quit IRC | 20:50 | |
| notmyname | nelson: actually that's a good thing | 20:50 |
| openstackhudson | Project nova build #502: SUCCESS in 1 min 32 sec: http://hudson.openstack.org/job/nova/502/ | 20:52 |
| openstackhudson | Tarmac: Define sql_idle_timeout flag to be an integer. | 20:52 |
| openstackhudson | This fixes bug706405 where MySQL Gone Away Errors were being generated. | 20:52 |
| *** matclayton has joined #openstack | 20:53 | |
| *** sandywalsh_ has joined #openstack | 20:53 | |
| notmyname | nelson: the s3 compatability layer shows how to mangle both the request and response http://bazaar.launchpad.net/~hudson-openstack/swift/trunk/view/head:/swift/common/middleware/swift3.py | 20:54 |
| *** sandywalsh_ is now known as sandywalsh | 20:54 | |
| *** reldan has quit IRC | 20:57 | |
| *** Ryan_Lane_ is now known as Ryan_Lane | 20:57 | |
| nelson | notmyname: will look at it later, thanks. | 20:57 |
| *** adiantum has quit IRC | 20:59 | |
| nelson | notmyname: but ... webob's Request seems ( http://pythonpaste.org/webob/reference.html ) to have a method called get_response() but I can't find it in that reference documentation. | 20:59 |
| nelson | am I just confused? Do I need to go get lunch? Low blood sugar?? | 20:59 |
| *** snelly has quit IRC | 21:05 | |
| *** CloudChris has joined #openstack | 21:06 | |
| *** CloudChris has left #openstack | 21:06 | |
| blamar | if you have pydoc you can just type 'pydoc webob.Request.get_response' | 21:07 |
| *** kpepple has quit IRC | 21:12 | |
| *** kpepple has joined #openstack | 21:13 | |
| *** pvo has joined #openstack | 21:15 | |
| *** GasbaKid has joined #openstack | 21:22 | |
| *** joearnold has joined #openstack | 21:22 | |
| dsockwell | how well does virtualbox integrate into openstack? does it mesh with vlans, etc? | 21:22 |
| vishy | dsockwell: the vbox support was in an old branch but it never made it into trunk | 21:26 |
| vishy | it supported a fairly limited set of features. | 21:26 |
| dsockwell | ah. is there any option for straight emulation as opposed to virtualization? | 21:26 |
| dsockwell | some qemu perhaps? | 21:26 |
| dsockwell | i know vmware is supposed to be forthcoming | 21:27 |
| *** miclorb has quit IRC | 21:29 | |
| vvuksan | dsockwell: you should be able to utilize qemu instead of kvm | 21:30 |
| vvuksan | i have | 21:30 |
| dsockwell | cool beans | 21:30 |
| vishy | dsockwell: yes qemu works just fine | 21:30 |
| vvuksan | i used to run whole nova inside of a KVM VM :-) | 21:30 |
| dsockwell | thanks ;) | 21:31 |
| vishy | i run nova inside of virtualbox | 21:31 |
| JordanRinke | elo | 21:31 |
| kbringard | vishy: I've got an odd one for you | 21:31 |
| kbringard | when I fire up more than like, 7 VMs in a single project, describe instances APi calls start failing | 21:32 |
| vishy | odd, what is the failure? | 21:33 |
| kbringard | sec, I'll paste it | 21:33 |
| kbringard | it looks like it's a relation thing between floating_ips and fixed_ips, but it's a lazy load session error in sqlalchemy | 21:33 |
| kbringard | http://paste.openstack.org/show/680/ | 21:34 |
| kbringard | DetachedInstanceError: Parent instance <FixedIp at 0x28469d0> is not bound to a Session; lazy load operation of attribute 'network' cannot proceed | 21:34 |
| kbringard | the odd thing is, I see this error from time to time, even when it works, but when I get more than like, 6 or 7 VMs running in a project, it starts tossing the unknown error thing and not returning | 21:35 |
| kbringard | and it only seems to happen regularly to users who have is_admin = 0 | 21:36 |
| kbringard | but it does happen from time to time to users who do not | 21:36 |
| *** mdomsch has quit IRC | 21:40 | |
| *** miclorb_ has joined #openstack | 21:40 | |
| *** reldan has joined #openstack | 21:44 | |
| *** joearnol_ has joined #openstack | 21:45 | |
| *** joearnold has quit IRC | 21:45 | |
| *** sebastianstadil has joined #openstack | 21:47 | |
| kbringard | vishy: odd, right? | 21:48 |
| vishy | yes odd | 21:50 |
| vishy | i think there is an easy fix | 21:50 |
| vishy | looks like it was an error introduced by ipv6 | 21:50 |
| *** lvaughn_ has quit IRC | 21:50 | |
| kbringard | yea, I don't recall it happening before the ip6 stuff was introduced | 21:50 |
| kbringard | want I should open a bug? | 21:51 |
| *** matclayton has left #openstack | 21:53 | |
| jarrod | im downgrading to 10.04LTS | 21:55 |
| imsplitbit | jarrod: that ain't no downgrade. maverick has been removed from all of my home lab boxes for 'flakiness under fire' | 21:56 |
| kpepple | vishy: have a moment to look at a sqlalchemy-migrate issue (we think) ? | 21:56 |
| imsplitbit | 10.04 is rock solid for me | 21:56 |
| jarrod | implsitbit: couldn't agree more | 21:56 |
| *** hub_cap has quit IRC | 21:59 | |
| kpepple | the model is in http://bazaar.launchpad.net/~ken-pepple/nova/dynamicinstancetypes/view/head:/nova/db/sqlalchemy/models.py and the migation is http://bazaar.launchpad.net/~ken-pepple/nova/dynamicinstancetypes/view/head:/nova/db/sqlalchemy/migrate_repo/versions/003_cactus.py | 21:59 |
| vishy | kbringard: http://pastie.org/1550608 | 21:59 |
| vishy | kbringard: that should fix your issue | 21:59 |
| kpepple | and it does have to just be vishy ... anyone is welcome to take a crack at those files | 21:59 |
| kbringard | vishy: you are teh man | 22:00 |
| jk0 | kpepple and I are stumped on this one | 22:00 |
| * jk0 is about ready to pull his hair out :) | 22:00 | |
| jk0 | vishy: if it helps, for some reason, nova-compute is ignoring the sql flag in the flagfile, so it goes to create an sqlite db instead of using existing mysql | 22:01 |
| jk0 | nova-api, network and scheduler all work as expected | 22:01 |
| *** vvuksan has quit IRC | 22:01 | |
| *** pvo has quit IRC | 22:02 | |
| *** miclorb_ has quit IRC | 22:04 | |
| *** reldan has quit IRC | 22:05 | |
| jarrod | hmm | 22:05 |
| vishy | jk0: how does that relate to migrations? | 22:05 |
| *** reldan has joined #openstack | 22:06 | |
| jk0 | vishy: it's the only thing we can think of in that branch | 22:07 |
| jk0 | something's gotta be misfiring upstream | 22:07 |
| *** mdomsch has joined #openstack | 22:07 | |
| vishy | so, there was an issue with this | 22:07 |
| jk0 | but, I dunno, we've been digging at this for two days, so we could just be overlooking something else by this point | 22:08 |
| *** dendrobates is now known as dendro-afk | 22:08 | |
| vishy | where sqlalchemy was being imported | 22:08 |
| vishy | before flags are parsed | 22:08 |
| vishy | sounds like one of the imports in compute | 22:08 |
| jk0 | ah yes, you had to do that late load | 22:08 |
| vishy | but that was back when models were being created automatically | 22:08 |
| vishy | are you on an old version of the code? | 22:09 |
| *** dendro-afk is now known as dendrobates | 22:09 | |
| vishy | because the db_sync changes stopped that problem | 22:09 |
| kpepple | this is based on pre-bexar release | 22:09 |
| jk0 | the branch is a couple weeks old, but we've tested this with current trunk as well | 22:09 |
| vishy | yeah so that is your problem | 22:09 |
| jk0 | in trunk, the exception disappears, but the problem still exists | 22:09 |
| vishy | something is importing sqlalchemy before flags are parsed | 22:09 |
| vishy | and the models are created with the default flag | 22:10 |
| vishy | kbringard: can you report a bug on your issue and let me know if that patch fixes it? I'll propose a merge if it works. | 22:11 |
| kbringard | OK, I just patched | 22:11 |
| *** burris has joined #openstack | 22:12 | |
| kbringard | vishy: https://bugs.launchpad.net/nova/+bug/701864 | 22:12 |
| uvirtbot | Launchpad bug 701864 in nova "nova.tests.test_cloud.CloudTestCase.test_associate_disassociate_address fails under python 2.7" [Medium,Fix released] | 22:12 |
| vishy | typo? | 22:13 |
| kbringard | not sure... that bug lists the same changes you made, but not in as many places | 22:13 |
| *** pvo has joined #openstack | 22:13 | |
| nelson | Are there packages on launchpad.net for Bexar yet? | 22:13 |
| nelson | I don't see them in the obvious place. | 22:14 |
| vishy | kbringard: different fix, this one is joinedloding network | 22:14 |
| creiht | nelson: not yet for swift :/ | 22:14 |
| creiht | mtaylor: ???? | 22:14 |
| vishy | *joinedloading | 22:14 |
| * nelson strikes a match and sticks it underneath creiht's chair. | 22:14 | |
| creiht | heh | 22:14 |
| nelson | heh | 22:14 |
| jk0 | vishy: are the migration versions somehow loaded automatically in compute when it's fired up? | 22:14 |
| vishy | kbringard: i may have put it in more places than strictly necessary | 22:15 |
| vishy | shouldn't be | 22:15 |
| * creiht pushes nelson next to mtaylor's chair | 22:15 | |
| jk0 | that's the only place I can see that might be a problem | 22:15 |
| jk0 | out of all the changes | 22:15 |
| vishy | you aren't importing sqlalchemy anywhere | 22:15 |
| vishy | or calling register_models? | 22:15 |
| vishy | gtg, send me a link to the branch and i'll take a look | 22:16 |
| jk0 | I don't think so, kpepple: you recall? | 22:16 |
| * nelson lights a fire under mtaylor 's butt instead. | 22:16 | |
| jk0 | vishy: http://bazaar.launchpad.net/~ken-pepple/nova/dynamicinstancetypes | 22:16 |
| *** grapex has left #openstack | 22:16 | |
| kpepple | vishy: we have the new table included in models.py:553 register_models ... but we don't call that anywhere | 22:17 |
| kbringard | vishy: yea, it looks like it worked, I'll open a bug and reference your diff | 22:17 |
| *** jaypipes has quit IRC | 22:18 | |
| *** pvo has quit IRC | 22:19 | |
| jarrod | i cant decide whether i should roll kvm or xen | 22:23 |
| jarrod | rackspace and ec2 use xen | 22:23 |
| jarrod | but kvm seems so much easier :) | 22:23 |
| kbringard | vishy: 716732 | 22:24 |
| *** spectorclan has joined #openstack | 22:24 | |
| aimon | Advantage and disadvantage of XEN is the kernel | 22:24 |
| kbringard | https://bugs.launchpad.net/nova/+bug/716732 | 22:24 |
| uvirtbot | Launchpad bug 716732 in nova "API returns UnknownError: An unknown error has occurred. Please try your request again. when describing instances" [Undecided,New] | 22:24 |
| kpepple | vishy , jk0 : i did add some sqlalchemy imports ... http://paste.openstack.org/show/682/ | 22:24 |
| aimon | Xen kernel has better abstraction from host OS and better security through kernel limits | 22:25 |
| aimon | ^^ thats why EC2 doesnt let people upload their own Kernls | 22:25 |
| uvirtbot | aimon: Error: "^" is not a valid command. | 22:25 |
| aimon | ...thats why EC2 doesnt let people upload their own Kernls | 22:25 |
| *** that__guy has joined #openstack | 22:26 | |
| that__guy | anyone up for answering a system admin question? | 22:27 |
| kbringard | depends on the question, and how long it'll take to answer | 22:28 |
| that__guy | I've got a whole slew of instances stuck in 'terminating' state... what's the most elegant way of destroying them | 22:28 |
| kbringard | I need to leave in about 10 mins | 22:28 |
| kbringard | this may not be the most elegant, but check the compute instance they're on to make sure the instance directory is gone | 22:28 |
| kbringard | and then remove them from the DB | 22:28 |
| that__guy | ok, I'll check it out | 22:29 |
| kbringard | that's how I've always done it | 22:29 |
| kbringard | but someone who has more intimate knowledge of the system may have a more elegant solution | 22:29 |
| nelson | creiht: notmyname was helping me earlier, but I think he was telling me how 1.2 works, but I'm running 1.1. | 22:29 |
| *** Sandys has joined #openstack | 22:29 | |
| nelson | creiht: specifically, it looks like the pipeline is returning an iterable that generates the content, and that the headers have already been generated. | 22:30 |
| jk0 | kpepple: looks like just the test though right? | 22:30 |
| *** joearnold has joined #openstack | 22:30 | |
| nelson | creiht: I'm trying to write a 404 handler for Swift without having to modify swift itself. | 22:31 |
| *** Sandys has quit IRC | 22:31 | |
| uvirtbot | New bug: #716732 in nova "API returns UnknownError: An unknown error has occurred. Please try your request again. when describing instances" [Undecided,New] https://launchpad.net/bugs/716732 | 22:31 |
| *** joearnol_ has quit IRC | 22:31 | |
| kpepple | jk0: i think it's just those two tests | 22:31 |
| openstackhudson | Project swift build #194: SUCCESS in 30 sec: http://hudson.openstack.org/job/swift/194/ | 22:31 |
| openstackhudson | Tarmac: cleaned up logging and added new optional proxy access_logger with it's own configuration options that default to default proxy logger's settings. | 22:31 |
| *** kbringard has quit IRC | 22:39 | |
| that__guy | for the devs out there: is there no other way of getting rid of instances? | 22:39 |
| *** allsystemsarego has quit IRC | 22:40 | |
| that__guy | besides rm'ing the instance dir and removing the instance row from the db? | 22:41 |
| *** GasbaKid has quit IRC | 22:41 | |
| *** GasbaKid has joined #openstack | 22:42 | |
| kpepple | that__guy: what do mean "get rid of" ? did your euca-terminate-instance hang ? | 22:50 |
| that__guy | kpepple: oh yes... on about 250 instances | 22:50 |
| kpepple | that__guy: if you just blow them out of the instances table, you will still have some networking to clean up | 22:52 |
| kpepple | that__guy: and dnsmasque addresses | 22:53 |
| that__guy | kpepple: oh yeah... great | 22:53 |
| kpepple | that__guy: i did that today .... it's quite a mess. i ended up dropping the db and starting over. this may not be a "best practice" tho. | 22:54 |
| that__guy | kpepple: yeah... I'm going to have to write some scripts to handle this. Have you tested any of your releases at scale? We're currently running maverick ppa trunk | 22:55 |
| *** brd_from_italy has quit IRC | 22:55 | |
| kpepple | that__guy: not at scale -- i have a three machine setup. i have heard rumors that the rackspace people are putting together a large deployment ... but you'd have to ask them about it | 22:56 |
| *** ctennis has joined #openstack | 22:59 | |
| *** ctennis has joined #openstack | 22:59 | |
| *** hub_cap has joined #openstack | 23:00 | |
| Ryan_Lane | hmm, this doesn't look positive... I rebooted an instance, and it looks like it went down, and isn't coming back up | 23:01 |
| nelson | ow | 23:02 |
| Ryan_Lane | are you not supposed to reboot an instance from inside of it? | 23:02 |
| nelson | I have trouble imagining that that shouldn't work. | 23:03 |
| jarrod | what does describe-instances show | 23:03 |
| Ryan_Lane | right | 23:03 |
| Ryan_Lane | running | 23:03 |
| Ryan_Lane | though, I did upgrade qemu while it was running... | 23:03 |
| nelson | creiht: ping | 23:03 |
| Ryan_Lane | with the natty version :) | 23:03 |
| *** GasbaKid has quit IRC | 23:04 | |
| Ryan_Lane | maybe I should reboot the host. heh | 23:04 |
| jarrod | no | 23:04 |
| creiht | nelson: howdy | 23:04 |
| Ryan_Lane | virsh shutdown doesn't work either :) | 23:05 |
| nelson | hi. still trying to figure out how I can get the proxy to fetch something from the cluster, and give me control if it fails. | 23:05 |
| *** reldan has left #openstack | 23:06 | |
| creiht | k | 23:06 |
| nelson | creiht: it kinda looks like the proxy server is calling start_response before it returns. | 23:06 |
| nelson | basically, I want a 404 handler. | 23:06 |
| creiht | that is possible | 23:06 |
| creiht | what do you want to do when a 404 occurs? | 23:06 |
| nelson | fetch the file from elsewhere and return it as if it was in the cluster. | 23:06 |
| nelson | OR, fetch it from a different name in the cluster, modify it, and return it. | 23:07 |
| creiht | well, seems like you could either handle that in the layer above (your squid proxy) | 23:07 |
| creiht | or write middleware | 23:07 |
| nelson | notmyname suggested that I do it similar to catch-errors | 23:08 |
| creiht | that would take the return iterable, and check if it starts with a 404 | 23:08 |
| nelson | but the return iterable is for the body | 23:08 |
| creiht | if so, then create a new response with whatever you want | 23:08 |
| nelson | and by that time, the 404 header has already been returned. | 23:08 |
| creiht | hrm | 23:09 |
| *** mdomsch has quit IRC | 23:09 | |
| creiht | I was pretty sure that it is the response data | 23:09 |
| * creiht looks | 23:09 | |
| nelson | (I could be wrong, of course) | 23:09 |
| creiht | I think the first line should be something like 200 OK, or 404 blah, etc. | 23:09 |
| *** spectorclan has quit IRC | 23:10 | |
| *** littleidea has joined #openstack | 23:10 | |
| *** rnirmal has quit IRC | 23:11 | |
| nelson | I was hoping that it would be returning a webob Response, but that's not what it's doing. | 23:11 |
| creiht | yeah | 23:11 |
| jarrod | weeeeeeee | 23:12 |
| creiht | nelson: In the future, I think I may need to modify how we handle that | 23:14 |
| nelson | yeah, I think that if the application in the pipeline returns a webob, then applications previously invoked in the pipeline can make changes. | 23:14 |
| creiht | yeah it would make it easier | 23:15 |
| nelson | they could be wrappers rather than prefixes. | 23:15 |
| creiht | Well they are still wrappers, just not as easy to implement | 23:15 |
| that__guy | yeeha! for the record: all I had to do was restart nova-compute on the hypervisors. the 'terminating' instances came back as 'running', and I just euca-term'd them again | 23:15 |
| creiht | nelson: you can send the app your own start_response function that allows you to get the data you are looking for, and then recreate the response | 23:16 |
| *** hub_cap has quit IRC | 23:16 | |
| that__guy | thanks for the help kbringard, kpepple! | 23:16 |
| creiht | FUJITA did that for the S3 compatibility middleware | 23:16 |
| nelson | yeah, that's what I was going to try next. | 23:16 |
| kpepple | that__guy: good to know | 23:16 |
| creiht | http://bazaar.launchpad.net/~hudson-openstack/swift/trunk/view/head:/swift/common/middleware/swift3.py | 23:16 |
| nelson | just save away the values until I figure out what I want to do. | 23:16 |
| creiht | yeah | 23:16 |
| *** mray has quit IRC | 23:17 | |
| *** kpepple has quit IRC | 23:17 | |
| *** kpepple has joined #openstack | 23:18 | |
| *** gondoi has quit IRC | 23:19 | |
| *** msassak has quit IRC | 23:21 | |
| *** MarkAtwood has quit IRC | 23:24 | |
| *** sebastianstadil has quit IRC | 23:26 | |
| *** vvuksan has joined #openstack | 23:28 | |
| *** msassak has joined #openstack | 23:33 | |
| *** vvuksan has quit IRC | 23:34 | |
| *** that__guy has quit IRC | 23:34 | |
| *** kpepple has quit IRC | 23:35 | |
| *** kpepple has joined #openstack | 23:35 | |
| *** clauden_ has joined #openstack | 23:38 | |
| *** littleidea has quit IRC | 23:38 | |
| *** kpepple has quit IRC | 23:41 | |
| *** kpepple has joined #openstack | 23:41 | |
| nelson | creiht: lines 425 and 429 of swift3.py interact strangely. | 23:42 |
| nelson | and even not, shouldn't the return at 429 be immediately after 425? | 23:44 |
| nelson | ^even if | 23:44 |
| uvirtbot | nelson: Error: "even" is not a valid command. | 23:44 |
| nelson | and even IF not, ... | 23:44 |
| iammartian | is this: https://github.com/vishvananda/novascript.git still the "master" location on nova.sh? it seems to be hosed for me | 23:44 |
| *** BK_man has quit IRC | 23:46 | |
| creiht | nelson: yeah now that you mention it, that does look a little strange | 23:47 |
| creiht | unless it is a side effect of how he is using the response | 23:48 |
| creiht | The odd thing is that when we tested the middleware, it worked :) | 23:48 |
| *** msassak has quit IRC | 23:49 | |
| nelson | It looks like he's throwing webob.Response()'s around and then calling them as if they had __call__ which they don't. | 23:50 |
| nelson | Or she, of course. | 23:50 |
| nelson | Oh, oops, yes a Response() does indeed have a __call__. | 23:51 |
| creiht | hehe | 23:51 |
| creiht | gotta run | 23:52 |
| *** BK_man has joined #openstack | 23:52 | |
| nelson | It's tough to figure out how webobs, paste.deploy, wsgi, and eventlets interact. | 23:52 |
| nelson | creiht: laters! | 23:52 |
| nelson | oh gosh, it's late there. | 23:52 |
| *** miclorb has joined #openstack | 23:54 | |
| *** miclorb has joined #openstack | 23:55 | |
| *** rpedde has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
