| vishy | adam_g: which signatures are you referring to? | 00:00 |
|---|---|---|
| comstud | crap | 00:02 |
| comstud | I didn't get the meeting onto my calendar | 00:03 |
| comstud | :( | 00:03 |
| *** dwalleck has quit IRC | 00:03 | |
| * comstud fired. | 00:03 | |
| vishy | comstud: its ok we made all of the decisions without you | 00:03 |
| comstud | it's cool.. | 00:03 |
| *** misheska has joined #openstack-dev | 00:05 | |
| *** dragondm has quit IRC | 00:10 | |
| *** jdg has quit IRC | 00:10 | |
| *** zns has joined #openstack-dev | 00:17 | |
| openstackgerrit | Brad Hall proposed a change to openstack-dev/devstack: Add support for specifying an alternate Quantum host/port https://review.openstack.org/2270 | 00:24 |
| adam_g | vishy: looking at it a bit closer, im running into mismatches between the v2 aws signatures euca2ools/boto attaches to an ec2 query and what keystone is generating in authenticate_ec2() | 00:27 |
| vishy | adam_g: hmm, it works fine for me | 00:27 |
| vishy | adam_g: which version of the code are you using? | 00:28 |
| adam_g | or i should say, the keystone generated signatures in its versin of the boto functions. :) even tho, string_to_sign is identical on both ends | 00:28 |
| adam_g | im using 2012.1~e2~20111209.1405-0ubuntu1 | 00:29 |
| *** nati2 has joined #openstack-dev | 00:29 | |
| * vishy is confused | 00:30 | |
| vishy | i use keystone/ec2 with boto in devstack | 00:30 |
| vishy | on trunk and stable/diablo | 00:30 |
| vishy | so I'm not sure what could be going wrong | 00:30 |
| *** zns has quit IRC | 00:31 | |
| adam_g | hmm | 00:31 |
| vishy | are you making a call using ips? | 00:31 |
| vishy | or hostname? | 00:31 |
| adam_g | this is essex/precise. im making calls from the client using hostname in EC2_URL | 00:31 |
| vishy | hmm | 00:32 |
| vishy | i wonder if that could be breaking it | 00:32 |
| adam_g | ill give it a host using an IP | 00:32 |
| vishy | you got precise working with keystone? | 00:32 |
| vishy | impressive | 00:32 |
| adam_g | ive tried both hostnames and IPs in the endpointTemplates | 00:32 |
| adam_g | vishy: yeah, it was working nicely with glance and nova up until today :P | 00:32 |
| adam_g | well, as of friday at least | 00:32 |
| vishy | adam_g: endpoint template shouldn't really matter for the ec2 portion | 00:33 |
| vishy | (it will be working with the endpoint template after the ec2 validation, but if python-novaclient works, then you know that section is working) | 00:34 |
| adam_g | right | 00:34 |
| *** davlap has joined #openstack-dev | 00:36 | |
| adam_g | yeah, looks like hostname in EC2_URL is throwing something off. getting a 401 now with matching signatures, so thats good | 00:38 |
| *** darthflatus1 has quit IRC | 00:38 | |
| *** darthflatus has joined #openstack-dev | 00:40 | |
| *** hugokuo has joined #openstack-dev | 00:46 | |
| vishy | adam_g: so if you use the same hostname in the flag it should work | 00:47 |
| *** Mandell has quit IRC | 00:48 | |
| vishy | adam_g: hmm, mb not | 00:49 |
| vishy | adam_g: looks like it is pulled from req.host | 00:49 |
| vishy | adam_g: that should probably come from a flag, considering load balancers, etc. | 00:50 |
| *** bhall_ has joined #openstack-dev | 00:50 | |
| *** misheska has quit IRC | 00:50 | |
| *** bhall has quit IRC | 00:50 | |
| *** bhall_ is now known as bhall | 00:50 | |
| vishy | adam_g: or at least use X-Forwarded-For | 00:52 |
| *** jog0 has quit IRC | 00:52 | |
| *** jog0 has joined #openstack-dev | 00:52 | |
| adam_g | vishy: use the same hostname in which flag? | 00:53 |
| vishy | adam_g: sorry there isn't a flag | 00:53 |
| *** jog0_ has joined #openstack-dev | 00:53 | |
| vishy | adam_g: host "should" be able to be overridden with a flag for sure | 00:55 |
| openstackgerrit | Jake Dahn proposed a change to openstack/horizon: Merging Floating Ips, Keypairs, and Security Groups into a single panel. https://review.openstack.org/2260 | 00:55 |
| adam_g | ah | 00:55 |
| openstackgerrit | Dean Troyer proposed a change to openstack-dev/devstack: Add open-iscsi packages for Nova compute nodes. https://review.openstack.org/2271 | 00:56 |
| *** vladimir3p has joined #openstack-dev | 00:57 | |
| *** jog0 has quit IRC | 00:57 | |
| *** jog0_ is now known as jog0 | 00:57 | |
| *** hugokuo has quit IRC | 00:58 | |
| vladimir3p | sandywalsh: ping | 01:02 |
| vladimir3p | Hi, have anybody tried schedulng multiple instances with requested_networks param? | 01:03 |
| vladimir3p | seems like it designed to work with a single instance creation only, or am I missing something ... ? | 01:03 |
| openstackgerrit | Jake Dahn proposed a change to openstack/horizon: Merging Floating Ips, Keypairs, and Security Groups into a single panel. https://review.openstack.org/2260 | 01:04 |
| sandywalsh | vladimir3p, yes, we had to switch back to single instance creation | 01:05 |
| *** spiffxp has quit IRC | 01:05 | |
| sandywalsh | vladimir3p, we were running into concurrency issues with multiple instance creation | 01:05 |
| vladimir3p | in scheduler? | 01:05 |
| sandywalsh | yup | 01:06 |
| vladimir3p | is it fixable? was it due to DB access > | 01:06 |
| *** zns has joined #openstack-dev | 01:07 | |
| vladimir3p | I've seen that you removed create_all_in_once | 01:07 |
| vladimir3p | but we create 2 at once and it seems to work for us | 01:07 |
| sandywalsh | yeah, all the zone/boot stuff was integrated into POST /servers and zone/boot removed | 01:07 |
| vladimir3p | (we scheduling them on 2 completely different compute nodes) | 01:07 |
| sandywalsh | vladimir3p, it will for trivial situations, but when you pound through fill-first requests it overallocates on a single host | 01:08 |
| *** ben_duyujie has joined #openstack-dev | 01:08 | |
| vladimir3p | I see.. but it seems to depend on filter/weight implementation logic | 01:09 |
| vladimir3p | for diablo/stable we've implemented our driver on top of abstract scheduler and creating a "build" plan with sorted list of nodes (based on available CPU/memory) | 01:09 |
| sandywalsh | vladimir3p, yes, sorry, there were many large refactorings to the scheduler to simplify the code and unify all the base classes | 01:10 |
| *** deshantm has joined #openstack-dev | 01:10 | |
| vladimir3p | ok, I see... So, do you plan to remove multiple instance creations at all? | 01:10 |
| sandywalsh | vladimir3p, for now, yes, we need to add support for cross-scheduler locking in order to do it right. | 01:11 |
| *** bhall has quit IRC | 01:11 | |
| vladimir3p | ok, thanks. | 01:12 |
| sandywalsh | np | 01:12 |
| *** jog0 has quit IRC | 01:14 | |
| *** jog0 has joined #openstack-dev | 01:14 | |
| *** aloga has quit IRC | 01:16 | |
| *** aloga has joined #openstack-dev | 01:18 | |
| *** zns has quit IRC | 01:18 | |
| *** sleepsonthefloo has quit IRC | 01:21 | |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: Fixes the syspanel instances page after instances_and_volumes merge. https://review.openstack.org/2272 | 01:26 |
| openstackgerrit | Vish Ishaya proposed a change to openstack/nova: Vm state management and error states https://review.openstack.org/1695 | 01:27 |
| *** tryggvil_ has quit IRC | 01:27 | |
| *** adjohn has quit IRC | 01:39 | |
| *** adjohn has joined #openstack-dev | 01:40 | |
| *** vladimir3p has quit IRC | 01:44 | |
| *** adjohn has quit IRC | 01:46 | |
| *** rods has quit IRC | 01:47 | |
| *** rods has joined #openstack-dev | 01:48 | |
| *** darthflatus has left #openstack-dev | 01:49 | |
| *** rods has quit IRC | 01:56 | |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: Volume name is passed to novaclient properly. Fixed bug 890908. https://review.openstack.org/2273 | 01:56 |
| uvirtbot | Launchpad bug 890908 in horizon "Use metadata rather than description for servers" [High,Confirmed] https://launchpad.net/bugs/890908 | 01:56 |
| *** lorin1 has joined #openstack-dev | 01:57 | |
| openstackgerrit | Jake Dahn proposed a change to openstack/horizon: merging https://review.openstack.org/2274 | 02:00 |
| *** rsampaio has joined #openstack-dev | 02:03 | |
| openstackgerrit | Jake Dahn proposed a change to openstack/horizon: merging https://review.openstack.org/2274 | 02:03 |
| *** davlap has quit IRC | 02:05 | |
| *** rsampaio has quit IRC | 02:07 | |
| *** jog0_ has joined #openstack-dev | 02:11 | |
| *** jog0_ has quit IRC | 02:12 | |
| *** jog0 has quit IRC | 02:14 | |
| *** jdurgin has quit IRC | 02:15 | |
| *** maplebed has quit IRC | 02:17 | |
| openstackgerrit | Neil Johnston proposed a change to openstack/horizon: blueprint https://blueprints.launchpad.net/horizon/+spec/cert-download https://review.openstack.org/2275 | 02:18 |
| openstackgerrit | Neil Johnston proposed a change to openstack/horizon: blueprint https://blueprints.launchpad.net/horizon/+spec/cert-download https://review.openstack.org/2276 | 02:18 |
| *** novas0x2a|laptop has quit IRC | 02:26 | |
| *** zigo has joined #openstack-dev | 02:33 | |
| openstackgerrit | Brian Waldon proposed a change to openstack/tempest: General test cleanup https://review.openstack.org/2277 | 02:41 |
| openstackgerrit | gholt proposed a change to openstack/swift: TempURL middleware. https://review.openstack.org/2246 | 02:44 |
| *** adjohn has joined #openstack-dev | 02:45 | |
| *** zns has joined #openstack-dev | 02:47 | |
| *** andrewbogott has quit IRC | 03:00 | |
| *** andrewbogott has joined #openstack-dev | 03:00 | |
| *** adjohn has quit IRC | 03:04 | |
| *** adjohn has joined #openstack-dev | 03:06 | |
| openstackjenkins | Project nova-docs build #1,763: SUCCESS in 3 min 3 sec: https://jenkins.openstack.org/job/nova-docs/1763/ | 03:18 |
| openstackjenkins | Vishvananda Ishaya: Vm state management and error states | 03:18 |
| *** ben_duyujie has quit IRC | 03:28 | |
| *** pixelbeat has quit IRC | 03:29 | |
| *** scottjg has joined #openstack-dev | 03:31 | |
| *** adjohn has quit IRC | 03:35 | |
| *** rohitk has joined #openstack-dev | 03:41 | |
| *** zigo-_- has joined #openstack-dev | 03:47 | |
| *** zigo has quit IRC | 03:48 | |
| *** rmkdroid has joined #openstack-dev | 03:48 | |
| *** lorin1 has quit IRC | 03:55 | |
| *** adjohn has joined #openstack-dev | 03:55 | |
| *** bcwaldon has quit IRC | 04:04 | |
| *** adjohn has quit IRC | 04:10 | |
| openstackgerrit | Yun Mao proposed a change to openstack/nova: fixed typos. removed an unused import https://review.openstack.org/2268 | 04:12 |
| *** mjfork has quit IRC | 04:29 | |
| *** rmkdroid has quit IRC | 04:32 | |
| *** deshantm has quit IRC | 04:47 | |
| *** bhall has joined #openstack-dev | 04:49 | |
| *** bhall has joined #openstack-dev | 04:49 | |
| *** sleepsonthefloo has joined #openstack-dev | 04:51 | |
| openstackgerrit | Brad Hall proposed a change to openstack/nova: Fix for bug 902175 https://review.openstack.org/2259 | 05:00 |
| uvirtbot | Launchpad bug 902175 in quantum "nova-manage network delete fails with QuantumManager" [High,In progress] https://launchpad.net/bugs/902175 | 05:00 |
| *** camm has quit IRC | 05:02 | |
| *** camm has joined #openstack-dev | 05:02 | |
| *** zigo-_- is now known as zigo | 05:30 | |
| openstackgerrit | gholt proposed a change to openstack/swift: TempURL and FormPost middleware. https://review.openstack.org/2246 | 05:41 |
| *** novas0x2a|laptop has joined #openstack-dev | 05:44 | |
| openstackgerrit | Naveed Massjouni proposed a change to openstack/nova: Renaming instance_actions.instance_id column to instance_uuid. blueprint: internal-uuids https://review.openstack.org/2244 | 05:50 |
| *** misheska has joined #openstack-dev | 06:01 | |
| *** novas0x2a|laptop has quit IRC | 06:06 | |
| *** ben_duyujie has joined #openstack-dev | 06:14 | |
| *** bhall has quit IRC | 06:17 | |
| *** novas0x2a|laptop has joined #openstack-dev | 06:26 | |
| openstackgerrit | A change to openstack/nova has been rejected: XenAPI: destroy VM if the post-boot operations fail. https://review.openstack.org/1909 | 06:29 |
| openstackgerrit | Kei Masumoto proposed a change to openstack/nova: First commit "deleting-compute-node-table" https://review.openstack.org/1917 | 07:09 |
| *** sanderj has quit IRC | 07:18 | |
| *** apevec has joined #openstack-dev | 07:25 | |
| *** nati2 has quit IRC | 07:27 | |
| *** lloydde has quit IRC | 07:31 | |
| openstackgerrit | dan wendlandt proposed a change to openstack/quantum: bug 903580: remove invalid extensions path from quantum.conf https://review.openstack.org/2278 | 07:43 |
| uvirtbot | Launchpad bug 903580 in quantum "avoid erroneous error about extensions path on boot" [Low,New] https://launchpad.net/bugs/903580 | 07:43 |
| *** zns has quit IRC | 07:44 | |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: Implementation of the BP services-insertion-wrapper inside the Cisco Plugin https://review.openstack.org/2279 | 07:47 |
| openstackgerrit | dan wendlandt proposed a change to openstack/quantum: bug 903581: remove etc/quantum.conf.sample as it is invalid https://review.openstack.org/2280 | 07:50 |
| uvirtbot | Launchpad bug 903581 in quantum "etc/quantum.conf.sample is not valid" [Undecided,New] https://launchpad.net/bugs/903581 | 07:50 |
| *** ben_duyujie has quit IRC | 07:56 | |
| *** ben_duyujie has joined #openstack-dev | 07:56 | |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: Implementation of the BP services-insertion-wrapper inside the Cisco Plugin https://review.openstack.org/2279 | 08:04 |
| *** vladimir3p has joined #openstack-dev | 08:09 | |
| *** pixelbeat has joined #openstack-dev | 08:13 | |
| *** mnour has joined #openstack-dev | 08:13 | |
| *** rohitk has quit IRC | 08:13 | |
| *** scottjg has quit IRC | 08:15 | |
| *** reidrac has joined #openstack-dev | 08:18 | |
| *** apevec has quit IRC | 08:19 | |
| *** lloydde has joined #openstack-dev | 08:26 | |
| openstackgerrit | Ghe Rivero proposed a change to openstack/quantum: Adds sqlalchemy support for ovs_quantum_plugin. https://review.openstack.org/1677 | 08:30 |
| *** lloydde has quit IRC | 08:30 | |
| *** pixelbeat has quit IRC | 08:32 | |
| *** vladimir3p has quit IRC | 08:33 | |
| *** ben_duyujie has quit IRC | 08:35 | |
| *** apevec has joined #openstack-dev | 08:39 | |
| *** pixelbeat has joined #openstack-dev | 08:40 | |
| openstackgerrit | Kei Masumoto proposed a change to openstack/nova: First commit "deleting-compute-node-table" https://review.openstack.org/1917 | 08:52 |
| *** zaitcev has joined #openstack-dev | 08:59 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/nova: floating-ip: return UUID of instance rather than ID https://review.openstack.org/2255 | 09:02 |
| andrea_ | hi I've got a question about SmokeStack, someone is online and can help me? | 09:33 |
| soren | What's the question? | 09:41 |
| *** darraghb has joined #openstack-dev | 10:09 | |
| openstackgerrit | dan wendlandt proposed a change to openstack/quantum: blueprint operational-status-ovs-plugin https://review.openstack.org/2281 | 10:32 |
| openstackgerrit | dan wendlandt proposed a change to openstack/quantum: blueprint api-operational-status https://review.openstack.org/2097 | 10:32 |
| openstackgerrit | A change to openstack/quantum has been rejected: blueprint operational-status-ovs-plugin https://review.openstack.org/2281 | 10:33 |
| andrea_ | Hi soren for this change:https://review.openstack.org/1503 the XenServer Failed | 10:38 |
| andrea_ | but I cannot understand the failure and how I can reproduce it and try to fix | 10:38 |
| *** tryggvil_ has joined #openstack-dev | 10:38 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: blueprint api-operational-status https://review.openstack.org/2097 | 10:44 |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: blueprint api-operational-status https://review.openstack.org/2097 | 10:52 |
| *** rods has joined #openstack-dev | 11:09 | |
| soren | andrea_: https://lists.launchpad.net/openstack/msg05858.html | 11:10 |
| *** zigo has quit IRC | 11:16 | |
| *** Oneiroi has joined #openstack-dev | 11:28 | |
| *** Oneiroi has quit IRC | 11:32 | |
| *** Oneiroi has joined #openstack-dev | 11:32 | |
| andrea_ | @soren thanks so much | 11:45 |
| larissa | andrea_: Error: "soren" is not a valid command. | 11:45 |
| andrea_ | soren thanks so much | 11:45 |
| *** zaitcev has quit IRC | 11:58 | |
| *** apevec has quit IRC | 12:02 | |
| soren | andrea_: np | 12:09 |
| *** lorin1 has joined #openstack-dev | 12:12 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: Fix for bug 803086 Moving the plugin_provider attribute into Quantum.conf, removing plugins.ini https://review.openstack.org/1415 | 12:20 |
| uvirtbot | Launchpad bug 803086 in quantum "plugins.ini should be collapsed into quantum.conf to prevent configuraitn"sprawl"" [Medium,In progress] https://launchpad.net/bugs/803086 | 12:20 |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: Bug 903684: functions defined twice in utils.py https://review.openstack.org/2282 | 12:24 |
| uvirtbot | Launchpad bug 903684 in quantum "import_class and import_object defined twice in utils.py" [Undecided,New] https://launchpad.net/bugs/903684 | 12:24 |
| *** zigo has joined #openstack-dev | 13:18 | |
| *** lorin1 has quit IRC | 13:29 | |
| *** bsza has joined #openstack-dev | 13:32 | |
| *** rsampaio has joined #openstack-dev | 13:32 | |
| *** apevec has joined #openstack-dev | 13:37 | |
| *** markvoelker has joined #openstack-dev | 13:37 | |
| openstackjenkins | Project nova-docs build #1,764: SUCCESS in 3 min 18 sec: https://jenkins.openstack.org/job/nova-docs/1764/ | 13:49 |
| openstackjenkins | Yun Mao: fixed typos. removed an unused import | 13:49 |
| *** hugokuo has joined #openstack-dev | 13:53 | |
| *** rsampaio has quit IRC | 14:10 | |
| *** zns has joined #openstack-dev | 14:13 | |
| *** mattray has joined #openstack-dev | 14:36 | |
| openstackgerrit | A change to openstack/nova has been rejected: First commit "deleting-compute-node-table" https://review.openstack.org/1917 | 14:43 |
| *** bcwaldon has joined #openstack-dev | 14:44 | |
| *** lorin1 has joined #openstack-dev | 14:56 | |
| openstackgerrit | Thierry Carrez proposed a change to openstack/nova: Sanitize EC2 manifests and image tarballs https://review.openstack.org/2283 | 15:00 |
| *** AlanClark has quit IRC | 15:07 | |
| *** zns1 has joined #openstack-dev | 15:09 | |
| *** zns has quit IRC | 15:09 | |
| *** joesavak has joined #openstack-dev | 15:09 | |
| *** dprince has joined #openstack-dev | 15:10 | |
| *** AlanClark has joined #openstack-dev | 15:10 | |
| openstackjenkins | Project nova-docs build #1,765: SUCCESS in 3 min 25 sec: https://jenkins.openstack.org/job/nova-docs/1765/ | 15:19 |
| openstackjenkins | Thierry Carrez: Sanitize EC2 manifests and image tarballs | 15:19 |
| *** mnour has quit IRC | 15:22 | |
| *** zaitcev has joined #openstack-dev | 15:22 | |
| *** dubsquared has joined #openstack-dev | 15:25 | |
| *** GheRivero_ has joined #openstack-dev | 15:26 | |
| *** joesavak has quit IRC | 15:28 | |
| *** deshantm has joined #openstack-dev | 15:33 | |
| *** code_franco has joined #openstack-dev | 15:43 | |
| *** dwalleck has joined #openstack-dev | 15:43 | |
| *** deshantm has quit IRC | 15:47 | |
| *** GheRivero_ has quit IRC | 15:47 | |
| *** nati2 has joined #openstack-dev | 15:48 | |
| *** cp16net has joined #openstack-dev | 15:49 | |
| *** rsampaio has joined #openstack-dev | 15:51 | |
| *** dragondm has joined #openstack-dev | 15:55 | |
| *** rnirmal has joined #openstack-dev | 15:58 | |
| *** deshantm has joined #openstack-dev | 16:01 | |
| openstackgerrit | dan wendlandt proposed a change to openstack/quantum: blueprint operational-status-ovs-plugin https://review.openstack.org/2281 | 16:01 |
| *** adjohn has joined #openstack-dev | 16:02 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/python-novaclient: Add list() method to ManagerWithFind https://review.openstack.org/2286 | 16:03 |
| openstackgerrit | Julien Danjou proposed a change to openstack/python-novaclient: utils.find_resource fixes + fix for volumes https://review.openstack.org/2287 | 16:03 |
| openstackgerrit | Julien Danjou proposed a change to openstack/python-novaclient: utils.find_resource fixes + fix for volumes https://review.openstack.org/2287 | 16:06 |
| openstackgerrit | Julien Danjou proposed a change to openstack/python-novaclient: Add list() method to ManagerWithFind https://review.openstack.org/2286 | 16:06 |
| *** nati2 has quit IRC | 16:06 | |
| openstackgerrit | Brian Waldon proposed a change to openstack/tempest: General test cleanup https://review.openstack.org/2277 | 16:06 |
| *** apeeved has joined #openstack-dev | 16:10 | |
| *** bsza has quit IRC | 16:12 | |
| *** apevec has quit IRC | 16:13 | |
| *** sleepsonthefloo has quit IRC | 16:14 | |
| *** reidrac has quit IRC | 16:15 | |
| *** maplebed has joined #openstack-dev | 16:16 | |
| *** lloydde has joined #openstack-dev | 16:19 | |
| *** eugenes has joined #openstack-dev | 16:26 | |
| eugenes | Not sure if openstack-dev is the right place for this question, but... I've been running strace on nova-compute in an effort to capture information in a log file. When I run strace on nova-compute and start creating instances, I'm presented with the following error: | 16:28 |
| *** dwalleck_ has joined #openstack-dev | 16:28 | |
| eugenes | TRACE: Command: sudo iptables-save -t filter | 16:29 |
| eugenes | TRACE: Exit code: 1 | 16:29 |
| eugenes | TRACE: Stdout: '' | 16:29 |
| eugenes | TRACE: Stderr: 'sudo: must be setuid root\n' | 16:29 |
| eugenes | TRACE: | 16:29 |
| *** dwalleck has quit IRC | 16:32 | |
| zaitcev | I dimly recall that I hit that before, in the sense that suid binaries cannot be strace-ed. | 16:32 |
| eugenes | any ideas for a workaround then zaitcev? | 16:34 |
| openstackgerrit | Julien Danjou proposed a change to openstack/python-novaclient: Fix PEP8 error https://review.openstack.org/2289 | 16:36 |
| openstackgerrit | Julien Danjou proposed a change to openstack/python-novaclient: Fix PEP8 errors https://review.openstack.org/2290 | 16:36 |
| *** hugokuo has left #openstack-dev | 16:38 | |
| zaitcev | eugenes: well, you can make strace suid itself | 16:40 |
| eugenes | ah okay. I will give that a go. Thanks zaitcev. | 16:41 |
| *** adjohn has quit IRC | 16:42 | |
| *** dwalleck has joined #openstack-dev | 16:44 | |
| *** dwalleck_ has quit IRC | 16:44 | |
| *** dragondm has quit IRC | 16:48 | |
| *** davlap has joined #openstack-dev | 16:52 | |
| vishy | smoser: there is a patch metadata by todd that makes it way faster in general | 16:55 |
| vishy | smoser, zul: I will try to get him to update it | 16:56 |
| smoser | thats the one that has the review up? | 16:57 |
| smoser | i can test that | 16:58 |
| smoser | i admit to never having don eit. | 16:58 |
| openstackgerrit | Rohit Karajgi proposed a change to openstack/tempest: Fixes lp:903466 https://review.openstack.org/2291 | 16:59 |
| zul | vishy: i can test that as well | 17:01 |
| openstackgerrit | Brad Hall proposed a change to openstack/nova: Fix some errors found by pychecker https://review.openstack.org/2265 | 17:02 |
| *** f-yang has quit IRC | 17:05 | |
| *** f-yang has joined #openstack-dev | 17:05 | |
| *** darraghb has quit IRC | 17:06 | |
| *** novas0x2a|laptop has quit IRC | 17:06 | |
| openstackgerrit | A change to openstack/nova has been rejected: models: make IP addresses unique https://review.openstack.org/1106 | 17:07 |
| *** darraghb has joined #openstack-dev | 17:07 | |
| vishy | smoser: it is really old at this point | 17:07 |
| vishy | smoser: but it prevents metadata from hitting the db every single time | 17:08 |
| vishy | smoser: do you get an error log in nova-api saying what the problem is btw? | 17:08 |
| vishy | smoser: when you get the 7/10 500s | 17:08 |
| *** dwalleck_ has joined #openstack-dev | 17:11 | |
| *** dwalleck has quit IRC | 17:11 | |
| *** openstackjenkins has quit IRC | 17:12 | |
| jeblair | i'm restarting jenkins to add a plugin; it usually takes 5-10 minutes to restart | 17:13 |
| *** openstackjenkins has joined #openstack-dev | 17:13 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: Bug #890028 https://review.openstack.org/1576 | 17:15 |
| uvirtbot | Launchpad bug 890028 in quantum "Use run_tests or pylint and pep8 validation" [Low,In progress] https://launchpad.net/bugs/890028 | 17:15 |
| *** lorin1 has quit IRC | 17:15 | |
| *** joesavak has joined #openstack-dev | 17:17 | |
| *** bhall has joined #openstack-dev | 17:18 | |
| *** bhall has joined #openstack-dev | 17:18 | |
| openstackgerrit | A change to openstack/nova has been rejected: models: make IP addresses unique https://review.openstack.org/1106 | 17:18 |
| *** maplebed has quit IRC | 17:19 | |
| *** bhall has quit IRC | 17:19 | |
| vishy | nova-core: review help here: https://review.openstack.org/#change,1202 | 17:20 |
| openstackgerrit | Julien Danjou proposed a change to openstack/python-novaclient: Allow to not specify image if block_device_mapping is set https://review.openstack.org/2292 | 17:21 |
| *** bhall has joined #openstack-dev | 17:21 | |
| dprince | Davey/zul: any time to merge this today? https://code.launchpad.net/~dan-prince/glance/ubuntu-glance-split/+merge/85059 | 17:29 |
| smoser | vishy, i cannot reproduce it now... i don't know what might have been causing it. possibly i had deleted the ami-id for the instance that was running | 17:31 |
| vishy | smoser: ok then | 17:31 |
| smoser | but even that wouldnt explain *some* success | 17:31 |
| smoser | it was strange. | 17:31 |
| vishy | smoser: I will bug todd about getting his fixes in | 17:31 |
| openstackgerrit | Tyler Smith proposed a change to openstack/quantum: Fixing bug/903829 Making setup_server.py not try to install quantum.conf.sample https://review.openstack.org/2293 | 17:31 |
| vishy | smoser: there is a weird heisenbug relating to sqlalchemy joinedloads | 17:31 |
| vishy | smoser: I suspect that is it | 17:32 |
| smoser | i'll comment in that bug and if i see it again i will record any log messages. | 17:32 |
| *** code_franco has quit IRC | 17:36 | |
| openstackgerrit | Stuart McLaren proposed a change to openstack/glance: Adds missing unit test for authentication plugins. https://review.openstack.org/2235 | 17:37 |
| *** code_franco has joined #openstack-dev | 17:37 | |
| *** dragondm has joined #openstack-dev | 17:38 | |
| vishy | dprince: how do i rekick smokestack for this one? https://review.openstack.org/#change,1202 | 17:38 |
| *** lorin1 has joined #openstack-dev | 17:39 | |
| *** maplebed has joined #openstack-dev | 17:40 | |
| *** jdurgin has joined #openstack-dev | 17:42 | |
| dprince | vishy: anyone on titan can click a button to re-run tests. | 17:42 |
| dprince | vishy: or you can post a new patchset. | 17:42 |
| *** jog0 has joined #openstack-dev | 17:42 | |
| vishy | dprince: mind rekicking that one? | 17:42 |
| vishy | i think it was a false negative | 17:42 |
| dprince | vishy: or I can give you an account and you can click the button yourself too. | 17:42 |
| dprince | vishy: Sure. | 17:42 |
| vishy | hmm, I had one at one point | 17:43 |
| dprince | vishy: I think I re-installed at some point.... | 17:43 |
| dprince | vishy: One thing to keep in mind is that it will only post results once per patchset. | 17:44 |
| dprince | vishy: at least that is how I coded it up yesterday as a first go. | 17:44 |
| vishy | dprince: that is cool | 17:44 |
| vishy | mtaylor: any progress on openvz? | 17:44 |
| dprince | vishy: to handle the false negatives we can just add a comment (manually) for now. | 17:45 |
| vishy | dprince, yeah i commented, but I want to actaully make sure it is a false negative :) | 17:45 |
| dprince | vishy: I'd rather go after the root causes of false negatives rather than spin to much on fancy re-posting results, etc. | 17:45 |
| zul | dprince: merged | 17:45 |
| dprince | zul: thanks. appreciated. | 17:46 |
| vishy | dprince: also, why do some of the results only have 1 or 2 of the test sections | 17:47 |
| vishy | (on reviewday) | 17:47 |
| dprince | vishy: For those I think it is where previous false negatives have been deleted. | 17:48 |
| dprince | vishy: those w/ accounts can do that... and we can re-run them. | 17:48 |
| jeblair | stable/diablo people: be aware of bug 903839 | 17:48 |
| uvirtbot | Launchpad bug 903839 in openstack-ci "nova stable/diablo unit test gate job fails due to wrong virtualenv" [Critical,In progress] https://launchpad.net/bugs/903839 | 17:48 |
| vishy | ah ok | 17:48 |
| *** spiffxp has joined #openstack-dev | 17:50 | |
| dprince | jaypipes: Just hit the glance unit tests failure again: http://paste.openstack.org/show/3777/ | 17:52 |
| dprince | jaypipes: I put in an explicit delete for /tmp/glance-tests* so my guess is that this is happening once we already start running tests as opposed to it being a cleanup issue between the test suite runs. | 17:53 |
| *** winston-d has quit IRC | 18:01 | |
| *** zns1 is now known as zns | 18:03 | |
| *** bhall has quit IRC | 18:04 | |
| openstackjenkins | Project nova-docs build #1,766: SUCCESS in 3 min 23 sec: https://jenkins.openstack.org/job/nova-docs/1766/ | 18:04 |
| openstackjenkins | Brad Hall: Fix network forwarding rule initialization in QuantumManager | 18:04 |
| *** dwalleck_ has quit IRC | 18:06 | |
| *** danishman has joined #openstack-dev | 18:07 | |
| *** adjohn has joined #openstack-dev | 18:08 | |
| *** andrewbogott has joined #openstack-dev | 18:09 | |
| *** bengrue has quit IRC | 18:10 | |
| *** bhall has joined #openstack-dev | 18:11 | |
| *** novas0x2a|laptop has joined #openstack-dev | 18:15 | |
| *** zaitcev has quit IRC | 18:18 | |
| *** pixelbeat has quit IRC | 18:19 | |
| *** maplebed has quit IRC | 18:19 | |
| openstackjenkins | Project nova-docs build #1,767: SUCCESS in 3 min 37 sec: https://jenkins.openstack.org/job/nova-docs/1767/ | 18:20 |
| openstackjenkins | Brad Hall: Fix for bug 902175 | 18:20 |
| uvirtbot | Launchpad bug 902175 in quantum "nova-manage network delete fails with QuantumManager" [High,In progress] https://launchpad.net/bugs/902175 | 18:20 |
| andrewbogott | ok... I have created a nova extension api, and it sort of works. But I /still/ don't understand how URLs are correlated to methods on the controller. Can anyone help me out? | 18:20 |
| *** apeeved has quit IRC | 18:21 | |
| andrewbogott | http://pastebin.com/tCY07DL3 | 18:21 |
| bcwaldon | andrewbogott: I can later today, busy atm | 18:21 |
| andrewbogott | ok | 18:22 |
| *** Oneiroi has quit IRC | 18:22 | |
| *** zigo has quit IRC | 18:34 | |
| *** darraghb has quit IRC | 18:37 | |
| *** jsavak has joined #openstack-dev | 18:38 | |
| *** Oneiroi has joined #openstack-dev | 18:41 | |
| *** Oneiroi has joined #openstack-dev | 18:41 | |
| openstackgerrit | Brian Waldon proposed a change to openstack/glance: Remove extra swift delete_object call https://review.openstack.org/2294 | 18:41 |
| *** Vek has quit IRC | 18:41 | |
| mtaylor | vishy: no. waiting on feedback from reddwarf | 18:41 |
| *** _cerberus_ has quit IRC | 18:41 | |
| *** joesavak has quit IRC | 18:42 | |
| openstackgerrit | Tyler Smith proposed a change to openstack/quantum: Fix for bug/893663 Making Cisco CLI usable from installed packages https://review.openstack.org/2295 | 18:42 |
| *** Oneiroi has quit IRC | 18:42 | |
| *** bengrue has joined #openstack-dev | 18:43 | |
| *** scottjg has joined #openstack-dev | 18:44 | |
| openstackgerrit | Chuck Short proposed a change to openstack/nova: Fix tgtadm off by one error. Fixes bug #871278 https://review.openstack.org/2296 | 18:47 |
| uvirtbot | Launchpad bug 871278 in nova "Cannot attach volumes to instances if tgt is used" [High,In progress] https://launchpad.net/bugs/871278 | 18:47 |
| vishy | mtaylor: :( | 18:47 |
| mtaylor | vishy: I'll do some poking, see if I can get some feedback | 18:48 |
| *** gyee has quit IRC | 18:48 | |
| *** joesavak has joined #openstack-dev | 18:49 | |
| *** gyee has joined #openstack-dev | 18:50 | |
| *** jsavak has quit IRC | 18:51 | |
| *** _cerberus_ has joined #openstack-dev | 18:55 | |
| *** GheRivero_ has joined #openstack-dev | 19:00 | |
| *** anotherjesse has joined #openstack-dev | 19:01 | |
| *** lorin1 has quit IRC | 19:02 | |
| *** andrewbogott has quit IRC | 19:06 | |
| *** andrewbogott has joined #openstack-dev | 19:08 | |
| *** scottjg has quit IRC | 19:10 | |
| *** dwalleck has joined #openstack-dev | 19:12 | |
| *** blamar has quit IRC | 19:33 | |
| *** blamar has joined #openstack-dev | 19:33 | |
| openstackgerrit | Brian Waldon proposed a change to openstack/glance: Ensure functional tests clean up their images https://review.openstack.org/2297 | 19:34 |
| *** danishman has quit IRC | 19:45 | |
| *** dwalleck_ has joined #openstack-dev | 19:54 | |
| openstackgerrit | Neil Johnston proposed a change to openstack/horizon: blueprint Download RC Files https://review.openstack.org/2298 | 19:55 |
| *** dwalleck has quit IRC | 19:57 | |
| *** davlap has quit IRC | 19:59 | |
| *** stuntmachine has joined #openstack-dev | 20:04 | |
| *** deshantm has quit IRC | 20:04 | |
| openstackgerrit | Brad Hall proposed a change to openstack/nova: Fix some errors found by pychecker https://review.openstack.org/2265 | 20:10 |
| *** stuntmachine has left #openstack-dev | 20:12 | |
| ttx | vishy: would be good to fix bug 903405 for E2 | 20:14 |
| uvirtbot | Launchpad bug 903405 in nova "ec2 metadata service extremely unreliable" [Critical,Confirmed] https://launchpad.net/bugs/903405 | 20:14 |
| vishy | ttx: it would be | 20:14 |
| vishy | unfortunately smoser can't repro or give me an error message | 20:14 |
| smoser | :) | 20:15 |
| smoser | i suspect i'll see it again | 20:15 |
| ttx | vishy: you mean if you run it, it works ? | 20:15 |
| smoser | if i run it, it works. | 20:15 |
| smoser | its random. | 20:15 |
| vishy | ttx: I expect a heisenbug that is fixed with a very odd change to sqlalchemy lazy loading | 20:15 |
| ttx | smoser: sounds like an important characteristic of this bug, missing from the deescription | 20:16 |
| ttx | will downgrade to "High" then | 20:16 |
| vishy | fixed ip model change to: network = relationship(Network, backref=backref('fixed_ips'), lazy="joined") | 20:16 |
| vishy | apparently fixes the heisenbug, so I wouldn't be surprised if this is the same | 20:16 |
| ttx | smoser: edited bug | 20:17 |
| smoser | i updated with comment. | 20:17 |
| openstackgerrit | Brian Waldon proposed a change to openstack/tempest: Fixing revert/confirm resize tests https://review.openstack.org/2299 | 20:17 |
| ttx | vishy, smoser: while you're both around -- what do we do with bug 901594 | 20:18 |
| uvirtbot | Launchpad bug 901594 in nova "Change in hostname exposition renders EC2 API unusable" [High,In progress] https://launchpad.net/bugs/901594 | 20:18 |
| *** jpipes has joined #openstack-dev | 20:19 | |
| ttx | I still think my patch makes it convenient a middle ground between how it used to be and how it should be | 20:19 |
| smoser | ttx, your patch was to return IP in private, is that correct ? | 20:20 |
| openstackgerrit | Verification of a change to openstack/glance failed: Fix for bug 901609, when using v2 auth should use /v2.0/tokens path. https://review.openstack.org/2235 | 20:20 |
| uvirtbot | Launchpad bug 901609 in glance "With keystone v2 auth does not use v2 path (/v2.0/tokens)" [Undecided,New] https://launchpad.net/bugs/901609 | 20:20 |
| *** adjohn has quit IRC | 20:21 | |
| *** adjohn has joined #openstack-dev | 20:21 | |
| ttx | smoser: used to be: privateDnsName = fixed_ip, dnsName = floating_ip or fixed_ip | 20:21 |
| ttx | smoser: currently is: privateDnsName = server_name, dnsName = floating_ip or server_name | 20:21 |
| ttx | smoser: my proposal: privateDnsName = server_name, dnsName = floating_ip or fixed_ip | 20:22 |
| ttx | i.e. the patch restores the old semantic for dnsName | 20:22 |
| ttx | so that at least one of the default fields gives you the IP | 20:23 |
| smoser | good explanation. i'm trying to think if it would make more sense to do dnsName or privateDnsName | 20:24 |
| smoser | with ip | 20:24 |
| ttx | smoser: I tried to align with how it used to be | 20:25 |
| smoser | well, in our diablo based nova, i get ips in both spots. | 20:25 |
| smoser | ah. yeah. which make sense with what you said. | 20:25 |
| ttx | right, and the bug was to expose server_name | 20:25 |
| smoser | but you could just as well chose: | 20:25 |
| smoser | privateDnsName = fixed_ip, dnsName = floating_ip or server_name | 20:26 |
| smoser | i was trying to think of any reason why your proposal would be superior/inferor to that option | 20:26 |
| ttx | smoser: right. I just thought "floating_ip or server_name" was kinda buggy anyway | 20:26 |
| smoser | i'm willing to defer to you there. | 20:27 |
| ttx | it's weird enough to have IPs in dns fields, at least be consistent :) | 20:27 |
| smoser | hopefully 12.04 will have some sort of '--ipv4' option to euca tools. | 20:27 |
| *** adjohn has quit IRC | 20:28 | |
| openstackgerrit | A change to openstack/tempest has been rejected: Fixing revert/confirm resize tests https://review.openstack.org/2299 | 20:29 |
| openstackgerrit | Brian Waldon proposed a change to openstack/tempest: Fixing revert/confirm resize tests https://review.openstack.org/2299 | 20:31 |
| bcwaldon | jeblair: can we stop the openstackgerrit bot from posting -2 reviews here? | 20:32 |
| markwash_ | vishy: ping | 20:35 |
| *** cp16net has quit IRC | 20:35 | |
| *** Vek has joined #openstack-dev | 20:37 | |
| *** eugenes has left #openstack-dev | 20:37 | |
| vishy | markwash_: pong | 20:41 |
| markwash_ | vishy: so I'm looking at the issue of trying to delete a server when it is in a bad state, specifically during a hung resize | 20:42 |
| markwash_ | and its looking like the simplest and most reliable approach to this involves kind of a big change and I want to get some initial feedback from you | 20:43 |
| markwash_ | basically, I'm thinking we should have a messaging queue per compute instance | 20:43 |
| markwash_ | well, technically, per (compute node, instance) pair | 20:43 |
| markwash_ | hmm | 20:43 |
| markwash_ | well details to be determined there | 20:43 |
| openstackgerrit | Aaron Lee proposed a change to openstack/nova: Make sure the rxtx_cap is used to set qos info https://review.openstack.org/2300 | 20:44 |
| markwash_ | I guess I'm wondering if you have any initial concerns about this approach? | 20:45 |
| notmyname | mtaylor: new git hotness http://johnkary.net/git-1-7-8-changes-for-the-everyday-developer/. I especially like the `git merge --edit` feature, and I think we could gain a _lot_ of benefit from it with gerrit | 20:47 |
| *** jog0 has quit IRC | 20:51 | |
| *** jog0 has joined #openstack-dev | 20:52 | |
| *** dwalleck_ has quit IRC | 20:53 | |
| *** anotherjesse has quit IRC | 20:53 | |
| *** vladimir3p has joined #openstack-dev | 20:54 | |
| openstackgerrit | Dan Prince proposed a change to openstack/glance: Update glance caching middleware so doesn't try to process calls to subresources. Fixes LP bug #889209. https://review.openstack.org/2301 | 20:54 |
| uvirtbot | Launchpad bug 889209 in glance "glance image-members causes httplib.BadStatusLine" [Medium,New] https://launchpad.net/bugs/889209 | 20:54 |
| openstackgerrit | dan wendlandt proposed a change to openstack/quantum: blueprint operational-status-ovs-plugin https://review.openstack.org/2281 | 20:55 |
| *** zns has quit IRC | 20:58 | |
| openstackgerrit | Ghe Rivero proposed a change to openstack/quantum: Improved VlanMap. https://review.openstack.org/2302 | 20:59 |
| openstackgerrit | Alex Meade proposed a change to openstack/nova: Expose instance faults in the api https://review.openstack.org/2303 | 20:59 |
| *** zns has joined #openstack-dev | 21:01 | |
| *** dprince has quit IRC | 21:01 | |
| vishy | smoser: I have a workaround for you btw | 21:05 |
| vishy | smoser: for your issue with uploading to glance and launching via euca | 21:05 |
| *** troytoman-away is now known as troytoman | 21:06 | |
| smoser | oh? | 21:07 |
| vishy | smoser: name | 21:07 |
| vishy | smoser: you can specify ami-<name> when you launch as well | 21:08 |
| smoser | i'm not terribly concerned about a work around, but I consider that necessary for essex. | 21:08 |
| smoser | funny... ami-name. | 21:08 |
| *** mjfork has joined #openstack-dev | 21:08 | |
| smoser | i figured i could use name to convert glance to euca ami | 21:08 |
| vishy | i used to do that with tty | 21:08 |
| vishy | register with the name tty then launch with ami-tty | 21:08 |
| smoser | but thats not a real solution | 21:08 |
| smoser | as there could be multipel names i suppose ? | 21:09 |
| vishy | smoser: yeah I don't know the real solution | 21:09 |
| smoser | or is name glablally unique? | 21:09 |
| smoser | globally even. | 21:09 |
| vishy | no it isn't | 21:09 |
| vishy | we need to involve jaypipes and mb bcwaldon to find the right solution | 21:09 |
| bcwaldon | yes? | 21:10 |
| smoser | vishy, bug 903927 | 21:10 |
| uvirtbot | Launchpad bug 903927 in nova "libvirt lxc failure path does not clean up mounts" [Undecided,New] https://launchpad.net/bugs/903927 | 21:10 |
| smoser | how should something leave the 'error' state ? | 21:10 |
| vishy | bcwaldon: so the issue is there is no longer non-uuid returned from glance | 21:10 |
| vishy | bcwaldon: so there is no way to figure out what the ec2 id will be when you register an image | 21:11 |
| *** markvoelker has quit IRC | 21:11 | |
| *** anotherjesse has joined #openstack-dev | 21:11 | |
| soren | *sigh* | 21:11 |
| bcwaldon | vishy: yep | 21:12 |
| soren | EC2 API not a second class citizen my <body part>. | 21:12 |
| bcwaldon | soren: well, we could have just stuck with EC2 API | 21:12 |
| bcwaldon | soren: would that be better? | 21:12 |
| smoser | well, soren, you're only in the predicament of having to do that because you can't upload an image via the ec2 api | 21:12 |
| smoser | if you could publish an image with ec2 api, then there'd really be no need to convert glance -> ami-id. | 21:13 |
| soren | smoser: Need I repeat myself or can you just imagine me doing it? | 21:13 |
| soren | bcwaldon: Honestly? | 21:13 |
| bcwaldon | soren: yeah, do you think that would be a better situation | 21:14 |
| soren | bcwaldon: Yes. | 21:14 |
| bcwaldon | soren: can you explain why? | 21:14 |
| bcwaldon | soren: I'm genuinely interested, not being a dink :) | 21:14 |
| vishy | shut up, you dink | 21:14 |
| vishy | :p | 21:15 |
| *** Ryan_Lane has joined #openstack-dev | 21:15 | |
| soren | I have eleventy billion reasons why. | 21:15 |
| soren | Primary one: The existing ecosystem. | 21:15 |
| bcwaldon | smoser: can you use nova-objectstore? | 21:15 |
| vishy | bcwaldon: yes, if it still worked with keystone | 21:15 |
| bcwaldon | damn | 21:15 |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: blueprint Download RC Files https://review.openstack.org/2298 | 21:15 |
| bcwaldon | soren: how could we expose our own functionality? | 21:16 |
| soren | bcwaldon: We could extend the EC2 API. Just like we've extended the RAckspace API. | 21:16 |
| bcwaldon | soren: well, the Compute API is no longer the Rackspace API, but that's definitely where it came from | 21:17 |
| soren | bcwaldon: It's not rocket science. If we actually had some traction, people would use it, and Amazon would have to choose between alienating users of these new features, or adopting our extensions. | 21:17 |
| bcwaldon | soren: Amazon doesn't give a crap | 21:17 |
| bcwaldon | soren: but I do agree with the existing ecosystem point | 21:18 |
| soren | bcwaldon: Because they don't care about the extra functionality we have. | 21:18 |
| smoser | as far as a solid work around goes, if i could do a 'describe-images with ami-GLANCE-UID' and get back ami-XXXXXXX, then i'd be good. it seems that if ami-NAME works, maybe ami-UUID does. | 21:18 |
| smoser | but i surely would like the real solutoin, and i'd like to get an EC2_CERT to be able to publish an image, as silly a path as that is (encrypting, splitting, uploading, registering... glance is simpler there) | 21:18 |
| bcwaldon | smoser: I think amazon requires you use either ami-<NAME> or ami-<HEX-ID> where HEX-ID isn't big enough to represent our uuid | 21:19 |
| bcwaldon | soren: so then we would have to roll our own tools since we would have an extended api, and EC2 has no concept of extension | 21:19 |
| tr3buchet | soren: i can just imagine | 21:19 |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: OpenRC file download settings panel w/ tenant selector. https://review.openstack.org/2298 | 21:19 |
| smoser | bcwaldon, well, clearly it would only be a workaround for openstack. but it would be a work around. | 21:19 |
| *** apevec has joined #openstack-dev | 21:19 | |
| bcwaldon | smoser: true | 21:19 |
| smoser | and, fwiw, amazon has increased the ami-abcdefg size before | 21:20 |
| soren | bcwaldon: It woulnd't be "extensions" in the OpenStack API sense. They'd just be new API calls that Amazon don't have. | 21:20 |
| smoser | obviously, all possible instance numbers can't fit in any fixed lenght string. | 21:20 |
| bcwaldon | soren: right, but we would have to extend the tooling to support those calls | 21:20 |
| soren | bcwaldon: Sure. | 21:20 |
| smoser | soren, bcwaldon i'm often just arguing to be arguing, but i'll point out, that you're not really going to accomplish anything here | 21:21 |
| smoser | :) | 21:21 |
| smoser | i tihnk there are 2 relevant pain points: | 21:21 |
| openstackgerrit | Dean Troyer proposed a change to openstack/nova: Fix flavor create arglist https://review.openstack.org/2304 | 21:21 |
| smoser | a.) cannot publish-image using keystone (as there is no EC2_CERT equivalent) | 21:21 |
| soren | smoser: Well, I have more ammunition the next time someone claims that the EC2 API isn't a second class citizen in Nova. | 21:21 |
| smoser | b.) cannot convert a glance uuid to an ec2 ami-id if you assume you can't do 'a', but still want to upload an image. | 21:22 |
| soren | smoser: That may be the choice we make, but we should stop claiming otherwise. | 21:22 |
| *** adjohn has joined #openstack-dev | 21:23 | |
| openstackgerrit | Dean Troyer proposed a change to openstack/nova: Fix flavor create arglist https://review.openstack.org/2304 | 21:25 |
| *** novas0x2a|laptop has quit IRC | 21:26 | |
| smoser | vishy, it it would appear that ami-NAME is broken now. | 21:30 |
| smoser | http://paste.ubuntu.com/769451/ | 21:30 |
| *** ohnoimdead has joined #openstack-dev | 21:31 | |
| vishy | oh boo | 21:32 |
| vishy | smoser: perhaps - isn't supported | 21:32 |
| vishy | will have to check | 21:32 |
| vishy | (or maybe I was using a name metadata value somehow) | 21:33 |
| *** zns1 has joined #openstack-dev | 21:33 | |
| andrewbogott | bcwaldon: Are you less swamped now? (If not, I'm patient.) | 21:33 |
| bcwaldon | once the team meeting is over, I'm free | 21:33 |
| andrewbogott | ok, thanks. Just give me a nudge. (Hopefully my questions are naive enough that they'll be quick to answer.) | 21:34 |
| *** zns has quit IRC | 21:35 | |
| vishy | andrewbogott: can i assign you on https://blueprints.launchpad.net/nova/+spec/public-and-private-dns | 21:35 |
| vishy | ? | 21:35 |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: Implementation of the BP services-insertion-wrapper inside the Cisco Plugin https://review.openstack.org/2279 | 21:35 |
| andrewbogott | vishy: yep! | 21:36 |
| *** dwalleck has joined #openstack-dev | 21:36 | |
| vishy | andrewbogott: ok, you have until e-3! | 21:36 |
| vishy | gl | 21:36 |
| andrewbogott | vishy: Is there a timeline page someplace? | 21:37 |
| * andrewbogott googles frantically | 21:38 | |
| *** mgius has joined #openstack-dev | 21:38 | |
| *** novas0x2a|laptop has joined #openstack-dev | 21:38 | |
| andrewbogott | nm, found it | 21:39 |
| *** bhall has quit IRC | 21:41 | |
| *** joesavak has quit IRC | 21:42 | |
| *** bhall has joined #openstack-dev | 21:42 | |
| *** spiffxp has quit IRC | 21:44 | |
| bcwaldon | andrewbogott: I'm here | 21:46 |
| bcwaldon | andrewbogott: what can I do you for | 21:46 |
| *** mgius has quit IRC | 21:47 | |
| *** mgius has joined #openstack-dev | 21:47 | |
| openstackjenkins | Project nova-docs build #1,768: SUCCESS in 3 min 52 sec: https://jenkins.openstack.org/job/nova-docs/1768/ | 21:50 |
| openstackjenkins | * Dan Wendlandt: bug 899767: fix vif-plugging with live migration | 21:50 |
| openstackjenkins | * Brad Hall: Fix some errors found by pychecker | 21:50 |
| uvirtbot | Launchpad bug 899767 in quantum "live_migration does not perform vif unplugging " [High,Fix committed] https://launchpad.net/bugs/899767 | 21:50 |
| openstackgerrit | Scott Moser proposed a change to openstack/nova: avoid error and trace on dom.vcpus() in lxc https://review.openstack.org/2305 | 21:51 |
| andrewbogott | bcwaldon: I am having trouble comprehending the relationship between the urls specified in python-novaclient and the methods defined in the actual nova api. | 21:52 |
| andrewbogott | I expected there to be just a big table of urls, but clearly the process is more subtle than that... | 21:52 |
| bcwaldon | andrewbogott: ok, can you give me an example case? | 21:52 |
| bcwaldon | that we can walk through | 21:52 |
| andrewbogott | Hm... sure. Lemme paste up some code. | 21:52 |
| smoser | that review of mine above should be straight forward and fixes a failure in lxc path. | 21:53 |
| andrewbogott | Here's a nova extension def: http://pastebin.com/4QpbiY1n | 21:53 |
| *** darthflatus1 has joined #openstack-dev | 21:53 | |
| * Vek has never seen any REST client that used a big table of URLs. Not that they don't exist; I've just never seen one... | 21:54 | |
| bcwaldon | andrewbogott: ok, so you're adding the top-level /os-floating-ip-dns resource | 21:54 |
| andrewbogott | I know (by luck) that I can access the 'index' method in that controller via self._list("/os-floating-ip-dns", "dns_entries"). But as soon as I try to access a second method, 'get', I'm at a loss. | 21:54 |
| *** darthflatus1 has quit IRC | 21:54 | |
| bcwaldon | andrewbogott: ok, so here's an example of a GET on an extension -> https://github.com/openstack/python-novaclient/blob/master/novaclient/v1_1/quotas.py | 21:55 |
| *** darthflatus has joined #openstack-dev | 21:55 | |
| bcwaldon | andrewbogott: what would the example URL be for a GET /us-floating-ip-dns/... call look like? | 21:56 |
| *** dwalleck has quit IRC | 21:56 | |
| *** anotherjesse has quit IRC | 21:57 | |
| *** anotherjesse has joined #openstack-dev | 21:57 | |
| andrewbogott | bcwaldon: I would've thought self._get("/os-floating-ip-dns/%s/get/%s" % (floating_ip_id, zone), "dns_entries") | 21:58 |
| *** ForeverNoobie has joined #openstack-dev | 21:58 | |
| bcwaldon | andrewbogott: I mean without any code, what would the http req url look like | 21:59 |
| vishy | bhall: ping | 21:59 |
| bcwaldon | andrewbogott: so you support /os-floating-ip-dns/ | 21:59 |
| bcwaldon | andrewbogott: /os-floating-ip-dns/1.2.3.4/zone_name | 22:00 |
| *** anotherjesse_ has joined #openstack-dev | 22:00 | |
| andrewbogott | So... that url doesn't specify a function name at all? | 22:00 |
| bcwaldon | andrewbogott: a more natural (rest-like) way to get the zones for an ip would be /os-floating-ip-dns/1.2.3.4/zones and .../zones/zone_name | 22:01 |
| bcwaldon | andrewbogott: correct, the HTTP method you use does that | 22:01 |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: OpenRC file download settings panel w/ tenant selector. https://review.openstack.org/2298 | 22:01 |
| andrewbogott | I'm pretty sure I see extensions that implement multiple methods for a single http method. Is that wrong? | 22:01 |
| bhall | vishy: hey | 22:01 |
| vishy | bhall: so here https://review.openstack.org/#change,2264 | 22:01 |
| bcwaldon | andrewbogott: so the python method 'get' in novaclient corresponds to the GET http method which corresponds to 'show' in the extension resource | 22:01 |
| bcwaldon | andrewbogott: well, the methods on the extension resources are determined by the http method AND resource | 22:02 |
| andrewbogott | right -- but why does 'get' necessarily correspond to 'show' and not to 'wakalixes'? | 22:02 |
| vishy | bhall: I don't understand why you would grab the host of the first scheduler from the services table | 22:02 |
| bcwaldon | so GET /collection and GET /collection/item will route to index and show, respectively | 22:02 |
| bcwaldon | andrewbogott: because we use a library called 'routes' that does that mapping for us | 22:02 |
| bcwaldon | andrewbogott: you can use different approaches to build your url maps, and the way we chose was to use the resource approach that gives us 5 methods by default: index, show, update, create, delete | 22:03 |
| andrewbogott | hm... ok. So 'routes' defines a ready-made mapping set. Is it possible to extend or override that set, if I need more methods than are available in the default set? | 22:03 |
| *** zns has joined #openstack-dev | 22:04 | |
| *** anotherjesse has quit IRC | 22:04 | |
| *** anotherjesse_ is now known as anotherjesse | 22:04 | |
| bcwaldon | andrewbogott: if you're thinking about adding new 'methods', I think you might want to take a step back and rethink your interface | 22:04 |
| bhall | vishy: see privmsg | 22:04 |
| andrewbogott | ok, fair enough. | 22:04 |
| *** rsampaio has quit IRC | 22:04 | |
| bcwaldon | andrewbogott: can you provide me your proposed list of resources? | 22:04 |
| bcwaldon | andrewbogott: a list of methods and what resources they can be used on | 22:04 |
| bcwaldon | andrewbogott: and what that resources is supposed to represent | 22:04 |
| ForeverNoobie | openstack looks like a pretty cool technology at first glance. I might get involved in a project or two | 22:05 |
| andrewbogott | bcwaldon: I'm trying to expose the driver interface defined here: http://wiki.openstack.org/InstanceAndFloatingIPDNSForNova | 22:05 |
| mtaylor | notmyname: looking | 22:05 |
| bcwaldon | andrewbogott: ok, give me a sec to look over it | 22:06 |
| openstackjenkins | Project nova-docs build #1,769: SUCCESS in 5 min 29 sec: https://jenkins.openstack.org/job/nova-docs/1769/ | 22:07 |
| openstackjenkins | naveedm9: Renaming instance_actions.instance_id column to instance_uuid. | 22:07 |
| *** zns1 has quit IRC | 22:07 | |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: OpenRC file download settings panel w/ tenant selector. https://review.openstack.org/2298 | 22:07 |
| bcwaldon | andrewbogott: can you explain how entries, zones, and addresses all relate | 22:07 |
| *** zns has quit IRC | 22:08 | |
| andrewbogott | bcwaldon: It's possible that that interface will need to be modified slightly, since it doesn't handle 'type' in a very smart way. But, for now, I'm thinking of the back end as a simple list of items: {address, zone, name} | 22:10 |
| andrewbogott | is that answering your question? | 22:11 |
| bcwaldon | andrewbogott: ok, would you want to expose that data hierarchically (sp?) | 22:11 |
| andrewbogott | I may not understand your question, but I think the answer is no -- each query will just return a flat list of one or more entries. | 22:12 |
| bcwaldon | andrewbogott: ok, so in other words, a request like GET /os-floating-ip-dns/ would return a list of lists | 22:12 |
| andrewbogott | yes. | 22:13 |
| bcwaldon | andrewbogott: ok, so to get to that point, all you need to do is define an index method | 22:13 |
| bcwaldon | (on your extension resource) | 22:13 |
| bcwaldon | andrewbogott: and after you have that working, you can add query param filtering | 22:13 |
| markwash_ | vishy: any thoughts about queue per instance as I mentioned above? just trying to screen early for any problems with the approach | 22:13 |
| bcwaldon | markwash_: imposter! | 22:14 |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: Bug #890028 https://review.openstack.org/1576 | 22:14 |
| uvirtbot | Launchpad bug 890028 in quantum "Use run_tests or pylint and pep8 validation" [Low,In progress] https://launchpad.net/bugs/890028 | 22:14 |
| andrewbogott | bcwaldon: The code that I pasted does that, right? At least, the index query seems to be working for me. | 22:15 |
| andrewbogott | Or, is my code so wrong as to be more-or-less unreadable? | 22:15 |
| bcwaldon | andrewbogott: no, I can tell that's what you're exposing | 22:15 |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: OpenRC file download settings panel w/ tenant selector. https://review.openstack.org/2298 | 22:16 |
| bcwaldon | andrewbogott: so you might want to think out how to organize this in a hierarchical fashion | 22:16 |
| andrewbogott | I don't think I know what that means :( | 22:16 |
| bcwaldon | andrewbogott: ok, let me write something up | 22:17 |
| andrewbogott | tangentially: If I need to pass in multiple params via the query... should I be rolling both into a dictionary, or is there support for multiple args in the url? | 22:17 |
| andrewbogott | (um... the thing I said about a dictionary doesn't make sense. | 22:18 |
| bcwaldon | andrewbogott: if its a create operation, use the body. If it's a get operation, use the url | 22:18 |
| andrewbogott | Is that as simple as /os-floating-ip-dns/arg1/arg2/arg3/etc? | 22:19 |
| andrewbogott | And those will get routed to the show method as its args, like magic? | 22:19 |
| bcwaldon | andrewbogott: no, /os-floating-ip-dns?arg1=this&arg2=this | 22:19 |
| andrewbogott | ahhh, ok. | 22:19 |
| bcwaldon | andrewbogott: only use that for filtering a list of objects | 22:19 |
| andrewbogott | filtering may be the main thing that I don't understand. But possibly I can ignore that for now :) | 22:20 |
| andrewbogott | I will tinker a bit. | 22:20 |
| openstackjenkins | Project nova-docs build #1,770: SUCCESS in 4 min 3 sec: https://jenkins.openstack.org/job/nova-docs/1770/ | 22:20 |
| openstackjenkins | Brad Hall: Fix error when subnet doesn't have a cidr set | 22:20 |
| bcwaldon | andrewbogott: check this out -> http://etherpad.openstack.org/hVDeD4YUGi | 22:22 |
| *** jog0_ has joined #openstack-dev | 22:22 | |
| bcwaldon | andrewbogott: what identifies an entry? how is it unique | 22:22 |
| *** bsza has joined #openstack-dev | 22:22 | |
| andrewbogott | zone.name should be unique. | 22:23 |
| *** cp16net has joined #openstack-dev | 22:23 | |
| bcwaldon | andrewbogott: what about the entries with in a zone | 22:23 |
| *** zns has joined #openstack-dev | 22:23 | |
| *** troytoman is now known as troytoman-away | 22:23 | |
| bcwaldon | andrewbogott: arbitrary integer ids? | 22:23 |
| *** jog0 has quit IRC | 22:24 | |
| *** jog0_ is now known as jog0 | 22:24 | |
| bcwaldon | andrewbogott: brb | 22:25 |
| vishy | markwash_: scrollback errors, can you copypaste? | 22:25 |
| andrewbogott | bcwaldon: In the underlying data there won't be any way to identify an entry other than by the values it contains. | 22:26 |
| andrewbogott | bcwaldon: to the extent that I use 'id' in my code, those are nova's ids for floating ips. | 22:27 |
| *** dwalleck has joined #openstack-dev | 22:31 | |
| *** jog0 has quit IRC | 22:31 | |
| bcwaldon | andrewbogott: can you only have one entry per floating ip? | 22:31 |
| andrewbogott | not necessarily. | 22:32 |
| *** jog0 has joined #openstack-dev | 22:32 | |
| andrewbogott | a given floating ip may have n entries in m different zones. | 22:32 |
| *** tylesmit has joined #openstack-dev | 22:32 | |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: OpenRC file download settings panel w/ tenant selector. https://review.openstack.org/2298 | 22:33 |
| bcwaldon | andrewbogott: so you're planning to create an entry then associate it with a zone? | 22:33 |
| bcwaldon | andrewbogott: and possibly associate it with many zones? | 22:33 |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: Bug #890028 https://review.openstack.org/1576 | 22:33 |
| uvirtbot | Launchpad bug 890028 in quantum "Use run_tests or pylint and pep8 validation" [Low,In progress] https://launchpad.net/bugs/890028 | 22:33 |
| *** GheRivero_ has quit IRC | 22:34 | |
| andrewbogott | bcwaldon: I think I just made things sound more confusing than they are. Basically, I'd expect a series of creates: | 22:34 |
| openstackgerrit | Kevin L. Mitchell proposed a change to openstack/nova: Updates OVS rules applied to IPv4 VIFs https://review.openstack.org/2306 | 22:35 |
| andrewbogott | create(id, 'foo', 'example.org') | 22:35 |
| andrewbogott | create(id, 'bar', 'example.org') | 22:35 |
| bcwaldon | andrewbogott: can you give me some context of what those values are? | 22:35 |
| andrewbogott | create(id, 'foo', 'someotherexample.com') | 22:35 |
| *** bsza has quit IRC | 22:35 | |
| openstackgerrit | Alex Meade proposed a change to openstack/nova: Expose instance faults in the api https://review.openstack.org/2303 | 22:36 |
| andrewbogott | (I hope it is not confusing that I have been saying 'zone' to mean 'dns zone' which has nothing to do with zones in nova project management) | 22:38 |
| bcwaldon | andrewbogott: i've got that | 22:39 |
| bcwaldon | andrewbogott: I'm curious what the arguments are to your create method above | 22:40 |
| andrewbogott | you mean, the 'id' arg? | 22:40 |
| *** anotherjesse has quit IRC | 22:40 | |
| bcwaldon | that, foo, and example.org | 22:40 |
| *** rnirmal has quit IRC | 22:40 | |
| openstackgerrit | A change to openstack/glance has been rejected: Update glance caching middleware so doesn't try to process calls to subresources. Fixes LP bug #889209. https://review.openstack.org/2301 | 22:41 |
| uvirtbot | Launchpad bug 889209 in glance "glance image-members causes httplib.BadStatusLine" [Medium,New] https://launchpad.net/bugs/889209 | 22:41 |
| openstackgerrit | A change to openstack/glance has been rejected: Update glance caching middleware so doesn't try to process calls to subresources. Fixes LP bug #889209. https://review.openstack.org/2301 | 22:41 |
| *** blakehaggerty has joined #openstack-dev | 22:41 | |
| andrewbogott | 'id' is the id used to identify a floating ip. So it'll be an arbitrary integer, in my experience. | 22:41 |
| openstackgerrit | A change to openstack/glance has been rejected: Update glance caching middleware so doesn't try to process calls to subresources. Fixes LP bug #889209. https://review.openstack.org/2301 | 22:41 |
| *** blakehaggerty has left #openstack-dev | 22:41 | |
| *** tylesmit has left #openstack-dev | 22:42 | |
| andrewbogott | The others are for a dns record, so they really will be strings like that in most cases. (In my case, things like 'test1' and 'labs.wikimedia.org') | 22:42 |
| bcwaldon | andrewbogott: ok, and what entity are you creating with that 'create' method? | 22:42 |
| andrewbogott | bcwaldon: It'll all depend on the driver, which is why I keep using the vague term 'entry'. In the reference driver I'm using now (minidns) it's just a line in a text file. | 22:43 |
| andrewbogott | space delimited, like "10.10.10.10 A foo example.org" | 22:44 |
| bcwaldon | andrewbogott: ok, so what is a 'zone' in the context of this api | 22:44 |
| openstackgerrit | Alex Meade proposed a change to openstack/nova: Expose instance faults in the api https://review.openstack.org/2303 | 22:45 |
| bcwaldon | I know its a dns zone, but what data will that have | 22:45 |
| andrewbogott | Just a string. | 22:45 |
| bcwaldon | andrewbogott: is a zone a logical grouping of entires? | 22:46 |
| andrewbogott | Um... zone == domain | 22:46 |
| bcwaldon | andrewbogott: ah | 22:46 |
| bcwaldon | andrewbogott: I'm not a dns expert ;) | 22:46 |
| andrewbogott | Sorry -- should've said that long ago :) | 22:46 |
| andrewbogott | I believe that there are pedantic differences, but for our purposes they're the same thing. | 22:46 |
| bcwaldon | andrewbogott: so we're really only working with entries here | 22:46 |
| andrewbogott | yep. Very simple. | 22:46 |
| *** dwalleck has quit IRC | 22:47 | |
| openstackgerrit | Gabriel Hurley proposed a change to openstack/horizon: OpenRC file download settings panel w/ tenant selector. https://review.openstack.org/2298 | 22:48 |
| *** jakedahn has joined #openstack-dev | 22:48 | |
| bcwaldon | andrewbogott: ok, go look at the etherpad again http://etherpad.openstack.org/hVDeD4YUGi | 22:49 |
| bcwaldon | andrewbogott: does that cover what you want? | 22:49 |
| *** ForeverNoobie has quit IRC | 22:52 | |
| andrewbogott | Hm... mostly. I'm not positive that we won't need to distinguish between a rename and a delete followed by an add. | 22:52 |
| bcwaldon | andrewbogott: ok, you can support a PUT operation as well that allows you to modify an entry in place | 22:52 |
| andrewbogott | PUT is different from POST, I take it? | 22:53 |
| andrewbogott | Also... in theory we need to be able to get by address/zone and also by name/zone | 22:53 |
| andrewbogott | Which, I'm not clear how we can overload get to do that. | 22:53 |
| bcwaldon | andrewbogott: yep, look at the pad again for an example PUT operation | 22:54 |
| bcwaldon | andrewbogott: I've got to head out for the day. Catch up with me tomorrow if you want to discuss this any more | 22:56 |
| *** jpipes has quit IRC | 22:57 | |
| andrewbogott | bcwaldon: Sure -- hopefully I'll have more-sensible code for you to look at by then. | 22:57 |
| andrewbogott | Thank you for your help! | 22:57 |
| bcwaldon | andrewbogott: I'm looking forward to it. You're welcome! | 22:57 |
| *** bcwaldon has quit IRC | 22:57 | |
| *** jakedahn_ has joined #openstack-dev | 23:05 | |
| *** jakedahn has quit IRC | 23:09 | |
| *** jakedahn_ is now known as jakedahn | 23:09 | |
| *** dtroyer has joined #openstack-dev | 23:15 | |
| *** stuntmachine has joined #openstack-dev | 23:19 | |
| *** andrewsben has joined #openstack-dev | 23:19 | |
| *** code_franco has quit IRC | 23:19 | |
| *** dubsquared has quit IRC | 23:19 | |
| openstackgerrit | Jake Dahn proposed a change to openstack/horizon: Clicking the logo now redirects you to the proper dashboard https://review.openstack.org/2307 | 23:22 |
| bengrue | notmyname ping | 23:29 |
| notmyname | bengrue: ack | 23:33 |
| vishy | bhall: are you aware that you can't receive pms? | 23:35 |
| vishy | bhall: says you are offline and refuses to send them | 23:35 |
| bhall | thats bizarre | 23:35 |
| bhall | this is what I get for switching to a graphical client.. back to irssi fo rme | 23:36 |
| *** jakedahn has quit IRC | 23:36 | |
| bengrue | notmyname: just curious as to what, specifically, memcache is used for in swift | 23:36 |
| notmyname | bengrue: mostly for caching auth tokens, container metadata (eg mostly existence), and account metadata (eg mostly existence) | 23:37 |
| *** bhall has quit IRC | 23:37 | |
| *** bhall has joined #openstack-dev | 23:37 | |
| notmyname | bengrue: if you don't have memcache in your pipeline at all, most everything should work. off the top of my head, I know that ratelimiting will fail and tempauth will fail | 23:38 |
| bengrue | ooooh. | 23:39 |
| bengrue | I will remove it from the pipeline and try it out. | 23:39 |
| openstackgerrit | Johannes Erdfelt proposed a change to openstack/nova: Refactor vm_state and task_state checking https://review.openstack.org/2308 | 23:39 |
| vishy | bhall: yeah weird. FYI it is node_availability_zone | 23:39 |
| notmyname | bengrue: it's supposed to work without memcached. just be slower. there are a few pieces of middleware that require it, but most things should work | 23:40 |
| bengrue | My question though is: if we do want memcache, should each swift instance in it's cluster have it's own private memcache? Seems like that's the only way to avoid adding a single point of failure. | 23:40 |
| openstackgerrit | nataraj ayyansamy proposed a change to openstack/tempest: fix for 903988 https://review.openstack.org/2309 | 23:40 |
| bengrue | Does swift assume that, should memache be there, it's a globally accessible resource? | 23:40 |
| *** zns has quit IRC | 23:40 | |
| bengrue | "globally" | 23:41 |
| notmyname | bengrue: no, eg we have memcache running on each proxy, so the MemcacheRing has reference to all of them | 23:41 |
| bengrue | excellent. | 23:41 |
| notmyname | depends what you mean by "globally" | 23:41 |
| notmyname | bengrue: if it's in the pipeline, it should be accessible to that process | 23:41 |
| bengrue | gotcha. | 23:41 |
| bengrue | That is what I was hoping, actually. Makes my work easier. | 23:42 |
| notmyname | the cache middleware adds the reference to the request env | 23:42 |
| bengrue | Thanks! | 23:42 |
| notmyname | sure | 23:42 |
| *** vlan-os has joined #openstack-dev | 23:42 | |
| jeblair | sleepsonthefloor: do you have a second to help with a devstack issue? | 23:42 |
| jeblair | https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/163/consoleFull | 23:42 |
| jeblair | i'd like to know why that failed; and if we're missing some output that would be helpful in determining that | 23:43 |
| openstackgerrit | Brad Hall proposed a change to openstack/nova: Pass additional information from nova to Quantum https://review.openstack.org/2264 | 23:43 |
| *** AlanClark has quit IRC | 23:43 | |
| jeblair | sleepsonthefloor: the vm still exists, i can add your key if needed | 23:44 |
| jeblair | sleepsonthefloor: and it's worth noting that the next run of the same change suceeded: https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/164/consoleFull | 23:45 |
| jeblair | so it's non-deterministic | 23:45 |
| vlan-os | i have a multi node OS setup using VlanManager. I have 2 subnets and am able to launch vm's in each one. However, i'm able to ping the vm on one subnet to the other vm (this happens only one way)...this is my nat iptables http://paste.openstack.org/show/3782/ | 23:46 |
| *** apevec has quit IRC | 23:46 | |
| vlan-os | any idea if i need to add another nat rule ? | 23:46 |
| *** davlap has joined #openstack-dev | 23:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!