| *** julim has quit IRC | 00:05 | |
| *** _nadya_ has quit IRC | 00:25 | |
| *** matsuhashi has joined #openstack-ceilometer | 00:27 | |
| *** nosnos has joined #openstack-ceilometer | 00:32 | |
| *** ildikov_ has joined #openstack-ceilometer | 00:48 | |
| *** ildikov has quit IRC | 00:51 | |
| *** liusheng has quit IRC | 00:51 | |
| *** liusheng has joined #openstack-ceilometer | 00:52 | |
| *** _nadya_ has joined #openstack-ceilometer | 00:52 | |
| *** skelpter has joined #openstack-ceilometer | 00:57 | |
| *** dmsimard is now known as dmsimard_away | 01:08 | |
| *** zqfan has joined #openstack-ceilometer | 01:10 | |
| *** nati_ueno has quit IRC | 01:30 | |
| *** fnaval has joined #openstack-ceilometer | 01:41 | |
| *** rwsu has quit IRC | 01:42 | |
| *** _nadya_ has quit IRC | 01:50 | |
| *** _cjones_ has quit IRC | 01:55 | |
| *** _cjones_ has joined #openstack-ceilometer | 01:56 | |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/pycadf: Bump hacking to 0.9.x series https://review.openstack.org/99542 | 01:57 |
|---|---|---|
| *** thomasem has joined #openstack-ceilometer | 01:59 | |
| *** thomasem has quit IRC | 01:59 | |
| *** _cjones_ has quit IRC | 02:00 | |
| *** skelpter has quit IRC | 02:11 | |
| *** _nadya_ has joined #openstack-ceilometer | 02:17 | |
| *** _cjones_ has joined #openstack-ceilometer | 02:27 | |
| *** _cjones_ has quit IRC | 02:34 | |
| *** fnaval has quit IRC | 03:01 | |
| *** zhikunliu has joined #openstack-ceilometer | 03:05 | |
| *** Longgeek has joined #openstack-ceilometer | 03:21 | |
| *** _cjones_ has joined #openstack-ceilometer | 03:29 | |
| *** _cjones_ has quit IRC | 03:34 | |
| *** nosnos has quit IRC | 03:44 | |
| *** r3pl4y has joined #openstack-ceilometer | 04:07 | |
| *** matsuhashi has quit IRC | 04:11 | |
| *** harlowja is now known as harlowja_away | 04:16 | |
| *** noorul has joined #openstack-ceilometer | 04:23 | |
| openstackgerrit | A change was merged to openstack/python-ceilometerclient: Remove © and remove unnecessary encoding lines https://review.openstack.org/95880 | 04:27 |
| *** r3pl4y has quit IRC | 04:28 | |
| *** _cjones_ has joined #openstack-ceilometer | 04:30 | |
| *** _cjones_ has quit IRC | 04:35 | |
| *** yatin has joined #openstack-ceilometer | 04:38 | |
| *** yatin_ has joined #openstack-ceilometer | 04:46 | |
| *** yatin_ has joined #openstack-ceilometer | 04:47 | |
| *** matsuhashi has joined #openstack-ceilometer | 04:48 | |
| *** yatin has quit IRC | 04:48 | |
| *** nosnos has joined #openstack-ceilometer | 04:48 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ceilometer: Updated from global requirements https://review.openstack.org/96819 | 04:53 |
| *** psharma has joined #openstack-ceilometer | 04:57 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-ceilometerclient: Updated from global requirements https://review.openstack.org/99567 | 04:59 |
| *** Longgeek_ has joined #openstack-ceilometer | 05:21 | |
| *** Longgeek has quit IRC | 05:24 | |
| *** _nadya_ has quit IRC | 05:51 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ceilometer: Imported Translations from Transifex https://review.openstack.org/88506 | 06:01 |
| *** _nadya_ has joined #openstack-ceilometer | 06:17 | |
| *** ildikov_ has quit IRC | 06:20 | |
| *** eglynn has joined #openstack-ceilometer | 06:35 | |
| *** ildikov has joined #openstack-ceilometer | 06:38 | |
| *** r3pl4y has joined #openstack-ceilometer | 06:46 | |
| *** inc0 has joined #openstack-ceilometer | 06:50 | |
| *** _nadya_ has quit IRC | 06:51 | |
| *** manoj has joined #openstack-ceilometer | 06:52 | |
| *** _nadya_ has joined #openstack-ceilometer | 06:52 | |
| manoj | hi all | 06:52 |
| manoj | is the instance level analytics available in openstack ceilometer | 06:54 |
| eglynn | manoj: good morning | 07:04 |
| eglynn | manoj: did you mean instance *metrics*? | 07:05 |
| manoj | eglynn: good morning | 07:05 |
| manoj | eglynn: yes | 07:05 |
| eglynn | manoj: (... as opposed to *analytics*) | 07:05 |
| *** _nadya_ has quit IRC | 07:05 | |
| eglynn | manoj: here are the instance-level metrics that ceilometer gathers ... http://docs.openstack.org/developer/ceilometer/measurements.html#compute-nova | 07:06 |
| eglynn | manoj: ... does that answer your question? | 07:06 |
| manoj | eglynn: yes, i need to know what is the instance duration? | 07:07 |
| *** ildikov has quit IRC | 07:07 | |
| *** safchain has joined #openstack-ceilometer | 07:08 | |
| eglynn | manoj: ... I guess you can derive that from the first and last sample timestamp associated with the resource | 07:09 |
| eglynn | manoj: http://docs.openstack.org/developer/ceilometer/webapi/v2.html#resources | 07:09 |
| eglynn | manoj: (though that would not accout for any "downtime" when the instance was suspended) | 07:09 |
| eglynn | bbiab | 07:09 |
| *** eglynn is now known as eglynn-commute | 07:10 | |
| zqfan | hi, the bot uploaded patch https://review.openstack.org/#/c/91238/14 has been reviewed for nearly 6 weeks, is it good to fly now? | 07:10 |
| *** zhikunliu1 has joined #openstack-ceilometer | 07:21 | |
| *** zhikunliu has quit IRC | 07:23 | |
| *** ildikov has joined #openstack-ceilometer | 07:24 | |
| *** zhikunliu1 has quit IRC | 07:26 | |
| *** zhikunliu has joined #openstack-ceilometer | 07:31 | |
| *** _nadya_ has joined #openstack-ceilometer | 07:38 | |
| *** eglynn-commute has quit IRC | 07:38 | |
| *** _cjones_ has joined #openstack-ceilometer | 07:43 | |
| *** _cjones_ has quit IRC | 07:47 | |
| *** _nadya_ has quit IRC | 08:03 | |
| *** eglynn-commute has joined #openstack-ceilometer | 08:07 | |
| *** asvechnikov has quit IRC | 08:16 | |
| *** vrovachev has quit IRC | 08:16 | |
| *** ilyashakhat has quit IRC | 08:17 | |
| *** ityaptin has quit IRC | 08:17 | |
| openstackgerrit | ZhiQiang Fan proposed a change to openstack/ceilometer: Improve api_paste_config file searching https://review.openstack.org/99599 | 08:20 |
| *** eglynn-commute is now known as eglynn | 08:34 | |
| *** zhikunliu has quit IRC | 08:38 | |
| llu | zqfan: thanks for the patch | 08:42 |
| *** _cjones_ has joined #openstack-ceilometer | 08:43 | |
| zqfan | hi, llu, the bot uploaded patch https://review.openstack.org/#/c/91238/14 has been reviewed for nearly 6 weeks, is it good to fly now? | 08:46 |
| *** pnascimento has joined #openstack-ceilometer | 08:48 | |
| zqfan | llu, and since I upload a workaround for that patch (accidently modify the topic), jenkins uploads another patch which does same job.... | 08:48 |
| *** _cjones_ has quit IRC | 08:48 | |
| *** Alexei_987 has quit IRC | 08:53 | |
| *** manoj has quit IRC | 08:56 | |
| llu | I think the previous patch sets didn't get merged in because of failure requirement tests, I think the infra has reverted the patch which caused that problem | 08:56 |
| llu | zqfan: sileht just approved that :) | 08:57 |
| zqfan | llu, sileht, thanks | 08:58 |
| sileht | llu, zqfan, but gate have a tone of patch that waiting for merging, we needs to wait at least 1 days before it get merged :( | 08:59 |
| zqfan | sileht, that's ok to me | 09:01 |
| *** zhikunliu has joined #openstack-ceilometer | 09:01 | |
| *** zhikunliu has quit IRC | 09:01 | |
| llu | sileht: i've got much more patience after I step into the openstack world :) | 09:02 |
| *** zhikunliu has joined #openstack-ceilometer | 09:02 | |
| *** _nadya_ has joined #openstack-ceilometer | 09:14 | |
| eglynn | liusheng: o/ | 09:16 |
| liusheng | eglynn, hi! | 09:17 |
| eglynn | liusheng: good morning / evening! | 09:18 |
| eglynn | liusheng: ... just wondering if you saw this thread on the ML http://lists.openstack.org/pipermail/openstack-dev/2014-June/037304.html | 09:18 |
| liusheng | eglynn, yes I just saw that | 09:19 |
| eglynn | liusheng: yeah ... so the usage of the nova floating_ip API all_tenants flag has been reverted on the ceilometer side | 09:19 |
| *** _nadya_ has quit IRC | 09:19 | |
| eglynn | liusheng: the suggestion on that thread is that we could possibly derive this information from notifications instead | 09:20 |
| eglynn | liusheng: ... does that sound reasonable to you? | 09:20 |
| eglynn | liusheng: ... in any case, could you chime in on the thread? | 09:20 |
| liusheng | eglynn, I have found that, and I am working on the patch https://review.openstack.org/99251 | 09:20 |
| liusheng | eglynn, yeah absolutely:) | 09:21 |
| eglynn | liusheng: cool, so definitely the underlying bug needs to be fixed ... | 09:22 |
| eglynn | liusheng: ... so thanks for proposing https://review.openstack.org/99251 | 09:22 |
| eglynn | liusheng: the other aspect is whether ceilometer can acquire the same info via notifications, this putting less load on nova-api | 09:23 |
| liusheng | eglynn, it is my fault for the effection :9 | 09:23 |
| liusheng | eglynn, aggree ! | 09:23 |
| eglynn | liusheng: cool, so replying on the ML thread with your thoughts and a reference to your nova fix would be great! | 09:24 |
| eglynn | liusheng: thank you sir! | 09:24 |
| liusheng | eglynn, previously, I found the the floatingip pollster is designed in ceilometer, but not avaliable | 09:24 |
| liusheng | eglynn, ok, thanks | 09:25 |
| *** KurtRao has quit IRC | 09:38 | |
| *** fc__ has quit IRC | 09:39 | |
| *** fc__ has joined #openstack-ceilometer | 09:40 | |
| *** Alexei_987 has joined #openstack-ceilometer | 09:52 | |
| sileht | eglynn, we continue to have the TestRealNotification that fail sometimes: https://github.com/enovance/ceilometer/commit/c12274ca2dfd270919608001e5bc5ab55ccfde55 doesn't fix the issue | 10:01 |
| sileht | latest gate issue: http://logs.openstack.org/40/99140/1/gate/gate-ceilometer-python26/df56e26/console.html | 10:01 |
| eglynn | sileht: darn! :( | 10:01 |
| openstackgerrit | Chinmaya Bharadwaj proposed a change to openstack/ceilometer: VMware:Update the ceilometer doc with VMware opts https://review.openstack.org/98132 | 10:01 |
| * eglynn looks ... | 10:01 | |
| sileht | eglynn, I guess the 'atexit' error come from a unrelease semaphore | 10:01 |
| openstackgerrit | Chinmaya Bharadwaj proposed a change to openstack/ceilometer: Fixes ceilometer-compute service start failure https://review.openstack.org/97831 | 10:02 |
| eglynn | sileht: i.e. the semaphore introduced in that initial "fix"? | 10:03 |
| eglynn | sileht: would it make sense to revert c12274ca? | 10:03 |
| eglynn | sileht: ... i..e pending a more correct fix | 10:04 |
| sileht | eglynn, I will try the find an other solution to ensure notification are published | 10:05 |
| * sileht needs to resetup its py26 VM | 10:05 | |
| eglynn | sileht: thank you sir! | 10:05 |
| *** zhikunliu has quit IRC | 10:05 | |
| * eglynn feels guilty every time someone mentions py26 ... | 10:05 | |
| sileht | this error occurs only in py26... | 10:06 |
| *** asvechnikov has joined #openstack-ceilometer | 10:11 | |
| *** ityaptin has joined #openstack-ceilometer | 10:12 | |
| *** idegtiarov has joined #openstack-ceilometer | 10:18 | |
| *** yatin__ has joined #openstack-ceilometer | 10:19 | |
| *** matsuhashi has quit IRC | 10:20 | |
| *** matsuhashi has joined #openstack-ceilometer | 10:20 | |
| *** yatin_ has quit IRC | 10:23 | |
| *** matsuhashi has quit IRC | 10:25 | |
| *** zqfan is now known as zqfan_afk | 10:34 | |
| *** underyx is now known as Underyx | 10:34 | |
| *** zqfan_afk is now known as zqfan_bot | 10:34 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Revert "Fix the floatingip pollster" https://review.openstack.org/99212 | 10:43 |
| *** _cjones_ has joined #openstack-ceilometer | 10:45 | |
| openstackgerrit | Yatin Kumbhare proposed a change to openstack/ceilometer-specs: Specs for email alarm notification https://review.openstack.org/99019 | 10:49 |
| *** _cjones_ has quit IRC | 10:50 | |
| *** _nadya_ has joined #openstack-ceilometer | 10:56 | |
| *** pnascimento has quit IRC | 10:57 | |
| *** _nadya_ has quit IRC | 11:05 | |
| EmilienM | eglynn: hey, just to let you know that ceilometer gating has now grenade jobs | 11:08 |
| EmilienM | eglynn: and also the fix for grenade upgrade is here: https://review.openstack.org/#/c/99412/ | 11:08 |
| EmilienM | thanks to sileht for the highlight | 11:09 |
| sileht | EmilienM, your welcome | 11:10 |
| EmilienM | sileht: :) | 11:10 |
| eglynn | EmilienM: excellent! :) | 11:12 |
| eglynn | EmilienM: ... so the lack of an dbsync was the cause of the ceilo-api silent failure after the upgrade? | 11:12 |
| EmilienM | eglynn: we think so yes | 11:12 |
| eglynn | EmilienM: cool, makes sense ... good catch guys! | 11:12 |
| EmilienM | all the credits to sileht | 11:13 |
| *** nosnos has quit IRC | 11:13 | |
| EmilienM | eglynn: so now, you should have gating for each ceilometer patch | 11:13 |
| eglynn | kudos to sileht's keen eyes! :) | 11:13 |
| sileht | eglynn, EmilienM even it does fix the ceilo-api, the dbsync is missing :p | 11:13 |
| EmilienM | and I'm scared until the dbsync stuff is not merged, all the ceilometer patches are blocked | 11:13 |
| eglynn | EmilienM: a-ha so it's currently voting? | 11:14 |
| eglynn | EmilienM: ... so we need to get https://review.openstack.org/99412 landed asap in that case | 11:14 |
| EmilienM | eglynn: i think yes, it's voting | 11:15 |
| EmilienM | if I understand correctly | 11:15 |
| eglynn | EmilienM: ... sdague has +2'd, I've just added dtroyer as an additional reviewer | 11:15 |
| eglynn | EmilienM: ... let's see how it plays out for a day or two | 11:16 |
| eglynn | EmilienM: ... if it's problematic, would it be easy to revert to non-voting for a bedding-in period? | 11:16 |
| EmilienM | eglynn: of course | 11:16 |
| openstackgerrit | Chinmaya Bharadwaj proposed a change to openstack/ceilometer: Fixes ceilometer-compute service start failure https://review.openstack.org/97831 | 11:17 |
| *** shakayumi has joined #openstack-ceilometer | 11:23 | |
| *** shakayumi has quit IRC | 11:23 | |
| *** dmsimard_away is now known as dmsimard | 11:23 | |
| *** shakayumi has joined #openstack-ceilometer | 11:24 | |
| *** r3pl4y has quit IRC | 11:27 | |
| *** r3pl4y has joined #openstack-ceilometer | 11:28 | |
| *** kun_huang has joined #openstack-ceilometer | 11:33 | |
| openstackgerrit | Dirk Mueller proposed a change to openstack/ceilometer: Downgrade publisher logging to debug level again https://review.openstack.org/96706 | 11:38 |
| *** yatin__ is now known as yatin | 11:39 | |
| *** yatin has quit IRC | 11:45 | |
| *** yatin has joined #openstack-ceilometer | 11:45 | |
| *** _cjones_ has joined #openstack-ceilometer | 11:46 | |
| *** kun_huang has quit IRC | 11:48 | |
| *** _cjones_ has quit IRC | 11:50 | |
| *** shakayumi has quit IRC | 11:50 | |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack/pycadf: Bump hacking to 0.9.x series https://review.openstack.org/99542 | 11:52 |
| yatin | hello folks | 11:58 |
| yatin | i found that ceilometer zeromq support in havana is broken.... | 11:59 |
| yatin | I want to fix that, but don;t know how do i go about it? | 12:00 |
| inc0 | yatin, what exactly do you mean? | 12:01 |
| yatin | inc0: create_worker and join_consumer method implementation not present for impl_zmq | 12:02 |
| yatin | inc0: hence ceilometer will not capture notificaton-info queue data | 12:02 |
| sileht | yatin, all rpc stuffs with zeromq is broken | 12:02 |
| inc0 | yatin, you could fix it in master | 12:02 |
| sileht | yatin, nobody use/maintain zmq driver of openstack | 12:02 |
| EmilienM | eglynn: sileht : so I confirm that grenade is now part of ceilometer gate: http://is.gd/CoaIH0 | 12:02 |
| inc0 | then it could be merged to stable/havaba but I doubt it | 12:03 |
| EmilienM | and is votin | 12:03 |
| EmilienM | and is voting | 12:03 |
| yatin | inc0: ok. i will check master then... | 12:03 |
| yatin | inc0: just thought ceilometer is collecting lots of data, hence zmq will be good fit for faster messaging | 12:04 |
| inc0 | yatin, well, it only listens on messages really | 12:04 |
| inc0 | so you would have to set up your whole openstack on zeromq | 12:05 |
| yatin | inc0: that's right | 12:05 |
| inc0 | and that might be...well...full of surprises and unwanted fun | 12:05 |
| sileht | yatin, inc0 zeromq in openstack have 0 unittests ... | 12:05 |
| inc0 | I'd strongly reccomend to stay with rabbit | 12:05 |
| sileht | or qpid | 12:05 |
| yatin | sileht: i'm new contributer, and read docs that openstack supports zeromq, qpid and rabbtmq hence, i guess, all should be well supported | 12:06 |
| sileht | yatin, do you have a link ? (this must be fixed :p) | 12:07 |
| yatin | sileht: i checkout havana ceilometer code, and those two methods, not fould implemented | 12:08 |
| yatin | sileht: in havana, collecter/service.py calls create_worker and join_consumer | 12:08 |
| yatin | sileht: from initialize_service_hook method | 12:09 |
| sileht | yatin, I known, I just say the doc must be fixed zeromq don't work with any openstack componments | 12:10 |
| EmilienM | eglynn: i think we should manage to have this patch merged | 12:10 |
| sileht | yatin, I will be happy to change that If I have a link :) | 12:10 |
| yatin | sileht: yes. you are right. docs does says that.. i remember seen this in docs | 12:10 |
| sileht | yatin, the first step to get zmq works with openstack is to write unittests in oslo.messaging project (and fix the tone of issue) and then impls the TODO and then perhaps looks in other project | 12:12 |
| yatin | sileht: sure. understood. | 12:15 |
| yatin | sileht: is that true, the fixes can't happen on havana. | 12:17 |
| yatin | sileht: ? | 12:17 |
| *** jdob has joined #openstack-ceilometer | 12:17 | |
| eglynn | EmilienM: ... thanks for the confirmation | 12:17 |
| sileht | yatin, I guess, this is not just bug fix, zmq driver doesn't implements some features, so backport will be not possible | 12:19 |
| yatin | sileht: yes. backport not possible. | 12:20 |
| EmilienM | eglynn: the patch is going to be merged, it has been approved by sdague | 12:21 |
| eglynn | EmilienM: cool :) | 12:22 |
| yatin | sileht: in that case, what you said earlier is the only choice | 12:22 |
| *** gordc has joined #openstack-ceilometer | 12:29 | |
| *** psharma has quit IRC | 12:41 | |
| *** openstackgerrit has quit IRC | 12:46 | |
| *** _cjones_ has joined #openstack-ceilometer | 12:46 | |
| *** openstackgerrit has joined #openstack-ceilometer | 12:47 | |
| *** rbowen has joined #openstack-ceilometer | 12:48 | |
| *** _cjones_ has quit IRC | 12:51 | |
| *** prad has joined #openstack-ceilometer | 12:56 | |
| *** julim has joined #openstack-ceilometer | 13:01 | |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/ceilometer: [TEST] try fix 1328679 https://review.openstack.org/99654 | 13:04 |
| *** thomasem has joined #openstack-ceilometer | 13:08 | |
| *** thomasem has quit IRC | 13:12 | |
| *** thomasem has joined #openstack-ceilometer | 13:13 | |
| inc0 | damn you ceilometer havana/stable, I've added ttl to samples (sqlalchemy impl) and I was happy | 13:19 |
| inc0 | but now I've looked up this db | 13:19 |
| inc0 | and figured out that there are no cascade on deletes between samples and sourceassoc and metadatas | 13:19 |
| inc0 | which means, nothing got deleted because of constrains >< | 13:20 |
| inc0 | and I have 20m records to delete manually | 13:20 |
| *** KurtRao has joined #openstack-ceilometer | 13:31 | |
| *** jergerber has joined #openstack-ceilometer | 13:32 | |
| *** sileht has quit IRC | 13:43 | |
| *** yatin has quit IRC | 13:43 | |
| *** zhikunliu has joined #openstack-ceilometer | 13:44 | |
| *** yatin has joined #openstack-ceilometer | 13:44 | |
| eglynn | folks, realistically those 3 remaining juno-1 target'd patches are not gonna land today | 13:55 |
| *** Longgeek has joined #openstack-ceilometer | 13:55 | |
| * eglynn bumps LP #1326250 #1326723 #1293337 to juno-2 | 13:55 | |
| eglynn | FYI the juno-1 tag will be dropped on https://github.com/openstack/ceilometer/commit/5aa5cfd5 | 13:55 |
| *** kun_huang has joined #openstack-ceilometer | 13:57 | |
| *** Longgeek_ has quit IRC | 13:57 | |
| *** kun_huang has quit IRC | 13:58 | |
| *** kun_huang has joined #openstack-ceilometer | 14:01 | |
| *** sileht has joined #openstack-ceilometer | 14:02 | |
| *** noorul has left #openstack-ceilometer | 14:05 | |
| *** zqfan_ has joined #openstack-ceilometer | 14:05 | |
| *** gordc has quit IRC | 14:06 | |
| *** Underyx is now known as underyx | 14:07 | |
| *** yatin has quit IRC | 14:08 | |
| *** yatin has joined #openstack-ceilometer | 14:08 | |
| sileht | launchpad mail spam has started, ceilometer j-1 will be cutted soon :) | 14:09 |
| ildikov | sileht: :) | 14:10 |
| *** yatin has quit IRC | 14:21 | |
| eglynn | sileht: yep, the juno-1 tag is out ... https://github.com/openstack/ceilometer/tree/2014.2.b1 | 14:21 |
| *** r3pl4y has quit IRC | 14:23 | |
| *** rbowen is now known as DrBacchus | 14:28 | |
| *** kun_huang has quit IRC | 14:30 | |
| *** kun_huang has joined #openstack-ceilometer | 14:34 | |
| DinaBelova | folks, btw - all Russians have holiday today - so I guess you do not see us today :) | 14:35 |
| *** chmartinez has joined #openstack-ceilometer | 14:36 | |
| ildikov | DinaBelova: tnx for the info! | 14:37 |
| DinaBelova | ildikov, I'll be on the meeting but just lurking :) | 14:38 |
| ildikov | DinaBelova: you should enjoy your vacation, but I will not say a word as I guess I'm the second most workaholic here after eglynn ;) | 14:39 |
| *** fnaval has joined #openstack-ceilometer | 14:39 | |
| DinaBelova | :D | 14:39 |
| * eglynn is more of a workaphobe ;) | 14:39 | |
| chmartinez | DinaBelova: vacations?? (btw, I'm cmart) | 14:40 |
| DinaBelova | :) - well, national holidays :) | 14:41 |
| DinaBelova | so all Russians are inactive :) | 14:41 |
| chmartinez | Oh I see.. | 14:41 |
| ildikov | eglynn: LOL :) | 14:41 |
| chmartinez | well, here in Argentina we have smth similar.. When Soccer World Cup starts, the country is frozen | 14:42 |
| chmartinez | i.e: nobody works.. Or breathe | 14:42 |
| chmartinez | :D | 14:42 |
| *** gordc has joined #openstack-ceilometer | 14:43 | |
| *** _cjones_ has joined #openstack-ceilometer | 14:48 | |
| *** admin0 has joined #openstack-ceilometer | 14:51 | |
| *** _cjones_ has quit IRC | 14:53 | |
| *** r3pl4y has joined #openstack-ceilometer | 15:03 | |
| *** zqfan_ is now known as zqfan | 15:03 | |
| *** r3pl4y has quit IRC | 15:07 | |
| *** r3pl4y has joined #openstack-ceilometer | 15:08 | |
| *** chmartinez has quit IRC | 15:12 | |
| *** fnaval has quit IRC | 15:17 | |
| *** _nadya_ has joined #openstack-ceilometer | 15:17 | |
| *** underyx is now known as Underyx|off | 15:17 | |
| *** fnaval has joined #openstack-ceilometer | 15:17 | |
| *** zhikunliu has quit IRC | 15:19 | |
| *** chmartinez has joined #openstack-ceilometer | 15:19 | |
| *** _nadya_ has quit IRC | 15:21 | |
| *** fnaval has quit IRC | 15:22 | |
| openstackgerrit | Cyril Roelandt proposed a change to openstack/ceilometer: Add an alarm service based on taskflow https://review.openstack.org/91763 | 15:23 |
| openstackgerrit | ZhiQiang Fan proposed a change to openstack/ceilometer: Improve api_paste_config file searching https://review.openstack.org/99599 | 15:25 |
| *** alexpilotti has joined #openstack-ceilometer | 15:26 | |
| Alexei_987 | jd__: so off meeting question - why? | 15:29 |
| Alexei_987 | jd__: you don't like the database? | 15:29 |
| jd__ | Alexei_987: because we have better solution available | 15:29 |
| jd__ | I like database to store data to request | 15:29 |
| jd__ | not as a queue system | 15:29 |
| Alexei_987 | jd__: it's not a queue | 15:30 |
| Alexei_987 | jd__: it's a state holder | 15:30 |
| Alexei_987 | so not a queue | 15:30 |
| jd__ | right | 15:30 |
| Alexei_987 | jd__: why it's better? | 15:30 |
| jd__ | so you just reimplemented Taskflow | 15:30 |
| Alexei_987 | jd__: if taskflow is implemented in 100 lines than yes | 15:31 |
| Alexei_987 | jd__: and it doesn't require a separate service configured | 15:31 |
| jd__ | no idea of the number of lines | 15:31 |
| Alexei_987 | jd__: so you could try to check code before rejecting it | 15:31 |
| jd__ | write a taskflow driver using a db (if there's not already?) | 15:31 |
| *** aviau has quit IRC | 15:33 | |
| *** _cjones_ has joined #openstack-ceilometer | 15:44 | |
| *** aviau has joined #openstack-ceilometer | 15:47 | |
| *** llu-laptop has joined #openstack-ceilometer | 15:50 | |
| sileht | llu, o/ | 15:50 |
| llu-laptop | sileht: sorry, I still don't quite understand why the senario you mentioned would result that bug, in that senario, the test case should fail | 15:50 |
| llu-laptop | but not at some I/O failure, but the assertion | 15:51 |
| llu-laptop | after the attempt to acquire the sem, self.srv.stop() stops the notification service | 15:51 |
| sileht | llu-laptop, yes, I agree, It should fail | 15:51 |
| *** fabiog has joined #openstack-ceilometer | 15:52 | |
| *** robsparker has joined #openstack-ceilometer | 15:52 | |
| sileht | llu-laptop, in gate it doesn't fail because the tests never finish, and gate kill tox | 15:53 |
| *** KurtRao has quit IRC | 15:53 | |
| sileht | llu-laptop, this is just my guessing, perhaps I'm wrong | 15:54 |
| llu-laptop | that's why I had the timeout, to make sure the acquire() returns | 15:54 |
| llu-laptop | sileht: are you saying that because the oslo.message's internal time.sleep, the sem.acquire(timeout) will never return? | 15:55 |
| sileht | llu-laptop, yes | 15:55 |
| sileht | llu-laptop, eventlet monkey patch time.sleep, so you can't know when the things happend | 15:56 |
| sileht | llu-laptop, previous I have tried to enable the testr timeout of each tests, I have put something 120 seconds, and all tests that really use rpc stuffs was fail sometimes | 15:57 |
| llu-laptop | sileht: that's the reason I think time.sleep shouldn't break eventlet's sem.acquire, becuase it monkey patch-ed time.sleep | 15:57 |
| llu-laptop | sileht: if it deosn't monkey patch time.sleep, then we should have the problem you mentioned | 15:58 |
| *** fnaval has joined #openstack-ceilometer | 15:58 | |
| llu-laptop | sileht: even in that case, I don't think remove the timeout would help, sem.acquire() just means wait for ever | 15:59 |
| sileht | llu-laptop, if time.sleep is not monkeypatch, when oslo.msg use eventlet.spawn, some tasks are never executed | 16:00 |
| *** Ruetobas has quit IRC | 16:01 | |
| *** admin0 has quit IRC | 16:01 | |
| sileht | llu-laptop, but your right that eventlet semaphore and eventlet time should work together correcly | 16:01 |
| openstackgerrit | Cyril Roelandt proposed a change to openstack/ceilometer: Add an alarm service based on taskflow https://review.openstack.org/91763 | 16:03 |
| sileht | llu-laptop, the only things we are sure is a test never returns | 16:03 |
| *** Ruetobas has joined #openstack-ceilometer | 16:03 | |
| sileht | llu-laptop, and each time I have got this backtrace , it was due to a unreleased semphare before app exit | 16:04 |
| eglynn | llu-laptop: so are you concerned that removing the timeout on the sem.acquire() will cause the test to hang in other circumstances? | 16:05 |
| llu-laptop | eglynn: yes, that's my concern | 16:05 |
| *** gordc has quit IRC | 16:06 | |
| llu-laptop | maybe we should try to use other mechanism than eventlet.semaphore if that's where the devil hide | 16:06 |
| ildikov | do we have this timeout value set from the beginning? | 16:07 |
| llu-laptop | ildikov: sorry, what do you mean by 'beginning'? | 16:07 |
| ildikov | llu-laptop: since that test case was created | 16:08 |
| * eglynn is being simplistic here ... but would a really long timeout, like say 600s, make any sense? | 16:08 | |
| llu-laptop | ildikov: no, this was introduced in https://review.openstack.org/#/c/98953/ to fix another gate bug | 16:08 |
| *** Ruetobas has quit IRC | 16:08 | |
| eglynn | sileht: ... or is it the case that setting *any* timeout is problematic? | 16:09 |
| ildikov | llu-laptop: a-ha, ok | 16:09 |
| sileht | eglynn, the worse case is the the first runned test started by testr is the latest to finish | 16:10 |
| ildikov | llu-laptop: and did you have a specific reason to set that timeout value or it was for preparing to anything that goes wrong? | 16:10 |
| sileht | because testr run tests in concurrency manner | 16:10 |
| *** idegtiarov has quit IRC | 16:11 | |
| eglynn | sileht: ... would that worst-case scenario cause any failures, or just an unexpected ordering of test completion? | 16:11 |
| llu-laptop | ildikov: I set the timeout in case some error occurs, so the test case has a chance to report the error and exit, not just wait there for ever | 16:11 |
| ildikov | llu-laptop: ok, I see | 16:13 |
| *** Ruetobas has joined #openstack-ceilometer | 16:13 | |
| sileht | eglynn, because of this worst-case we cannot determine how long a test needs to run | 16:14 |
| ildikov | eglynn: I guess that worst case scenarion would cause the same failure in case of even the longer timeout value | 16:14 |
| sileht | eglynn, only tests that really use oslo messaging are impacted | 16:15 |
| sileht | the semaphore just hide the real issue | 16:15 |
| ildikov | eglynn: ... and if we set a that long timeout that it could not happen, then it's more clear to not use timeout at all | 16:15 |
| sileht | the fact is we have at least one test that sometime never returns and gate timeout | 16:16 |
| eglynn | sileht: what's the real issue being hidden? ... an oslo-messaging bug, or expected oslo-messaging behavior? | 16:16 |
| sileht | I don't know yet | 16:17 |
| eglynn | sileht: a-ha, ok ... do you think that removing the timeout will help unmask the real failure? | 16:18 |
| eglynn | sileht: i.e. maybe give us better clues/evidence to suggest the real root cuase? | 16:18 |
| ildikov | what about using the semaphore without timeout, if that really works, until we know what's the real issue? | 16:18 |
| llu-laptop | if there is a real issue between timeout sempahore and oslo.messaging's internal time.sleep, do you guys think using active polling/yield would help: | 16:23 |
| llu-laptop | time_start = get start time | 16:23 |
| llu-laptop | while now() - time_start < certain time: | 16:23 |
| llu-laptop | poll | 16:23 |
| llu-laptop | yield | 16:23 |
| eglynn | llu-laptop: "poll" == "check if expected notification was received"? | 16:25 |
| llu-laptop | poll=='check if expected number of notification was received' | 16:26 |
| eglynn | yeah | 16:26 |
| eglynn | sileht: ^^^ does llu-laptop's suggestion sound like a plausible approach to you? | 16:28 |
| *** Longgeek has quit IRC | 16:29 | |
| llu-laptop | b.t.w does this bug happen in the check queue? | 16:29 |
| llu-laptop | or only in the gate-queue? | 16:29 |
| *** inc0 has quit IRC | 16:29 | |
| llu-laptop | I see, it happens in the check queue too | 16:30 |
| eglynn | llu-laptop: whay would that make a difference? (... i.e. check versus gate) | 16:30 |
| eglynn | *why | 16:30 |
| llu-laptop | llu-laptop: don't know. that's why i'm asking | 16:31 |
| sileht | eglynn, yes but this add a timeout too, but why not | 16:31 |
| eglynn | llu-laptop: a-ha, k | 16:31 |
| llu-laptop | sileht: by saying timeout, you are referring to yield? | 16:31 |
| *** _nadya_ has joined #openstack-ceilometer | 16:32 | |
| eglynn | sileht: yeah, but not the "bad" sort of timeout with the atexit side effect, right? | 16:32 |
| llu-laptop | eventlet.greenthread.sleep(seconds=0) | 16:32 |
| sileht | "certain time" | 16:32 |
| eglynn | llu-laptop: now() - time_start < certain time | 16:32 |
| *** safchain has quit IRC | 16:32 | |
| llu-laptop | that would make the test case at assertion, not killed by CI | 16:33 |
| llu-laptop | make the test case fail at assertion, not failed with atexit | 16:33 |
| sileht | sound good | 16:33 |
| eglynn | llu-laptop: could you code up that idea, so we can run it thru' the gate a few times? | 16:34 |
| llu-laptop | ok, i'll try to cook a pathc tomorrow, and do some recheck to see if this is real devil | 16:34 |
| eglynn | llu-laptop: excellent :) | 16:34 |
| eglynn | sileht, llu-laptop: thank you sirs! | 16:34 |
| *** gordc has joined #openstack-ceilometer | 16:36 | |
| *** llu-laptop has quit IRC | 16:36 | |
| *** openstackgerrit has quit IRC | 16:38 | |
| *** isviridov is now known as isviridov|away | 16:47 | |
| *** zqfan has quit IRC | 16:50 | |
| *** _nadya_ has quit IRC | 16:50 | |
| *** zqfan has joined #openstack-ceilometer | 16:53 | |
| *** harlowja_away is now known as harlowja | 16:54 | |
| *** zqfan has quit IRC | 16:55 | |
| *** zqfan has joined #openstack-ceilometer | 16:56 | |
| *** erecio has quit IRC | 16:58 | |
| *** erecio has joined #openstack-ceilometer | 16:59 | |
| *** rwsu has joined #openstack-ceilometer | 17:00 | |
| *** erecio has quit IRC | 17:06 | |
| *** nati_ueno has joined #openstack-ceilometer | 17:06 | |
| zqfan | eglynn: is havana only accept security backport now? I have some backport patch not related to security, if that is true, seems better to abandon them | 17:06 |
| eglynn | zqfan: yes, stable/havana is only for security patches | 17:08 |
| *** erecio has joined #openstack-ceilometer | 17:09 | |
| *** blinky_ghost has joined #openstack-ceilometer | 17:17 | |
| blinky_ghost | hi all, I'm looking for a solution for monitoring my openstack deployment/services. My question is if ceilometer does this job or only monitors resources that run on the cloud, like instances, networks, storage? Thanks | 17:19 |
| *** rwsu has quit IRC | 17:30 | |
| *** rwsu has joined #openstack-ceilometer | 17:32 | |
| *** fnaval has quit IRC | 17:36 | |
| *** fnaval has joined #openstack-ceilometer | 17:37 | |
| *** fnaval has quit IRC | 17:41 | |
| *** admin0 has joined #openstack-ceilometer | 17:51 | |
| *** admin0 has quit IRC | 17:55 | |
| *** blinky_ghost has quit IRC | 18:00 | |
| *** zul has quit IRC | 18:03 | |
| *** zul has joined #openstack-ceilometer | 18:03 | |
| *** eglynn has quit IRC | 18:06 | |
| *** admin0 has joined #openstack-ceilometer | 18:11 | |
| *** ildikov has quit IRC | 18:12 | |
| *** jdob has quit IRC | 18:12 | |
| *** openstackgerrit has joined #openstack-ceilometer | 18:12 | |
| *** jdob has joined #openstack-ceilometer | 18:13 | |
| *** admin0 has quit IRC | 18:17 | |
| *** admin0 has joined #openstack-ceilometer | 18:19 | |
| *** greghaynes has joined #openstack-ceilometer | 18:22 | |
| *** jdob has quit IRC | 18:25 | |
| *** jdob has joined #openstack-ceilometer | 18:25 | |
| *** admin0 has quit IRC | 18:28 | |
| *** admin0 has joined #openstack-ceilometer | 18:30 | |
| *** erecio has quit IRC | 18:34 | |
| *** erecio has joined #openstack-ceilometer | 18:34 | |
| *** fnaval has joined #openstack-ceilometer | 18:38 | |
| greghaynes | Hello ceilometerers, I am working on leader election code in Tripleo and someone linked me to https://etherpad.openstack.org/p/icehouse-summit-ceilometer-central-agent - Does anyone know if any of the information there has been acted upon? | 18:44 |
| greghaynes | specifically, it would be less than ideal for tripleo and ceilometer to use two different systems for this, so hoping to prevent that | 18:45 |
| *** admin0 has quit IRC | 18:46 | |
| *** zqfan has quit IRC | 18:47 | |
| *** admin0 has joined #openstack-ceilometer | 18:47 | |
| *** erecio has quit IRC | 18:53 | |
| *** Alexei_987 has quit IRC | 19:00 | |
| *** Shaan7 has quit IRC | 19:02 | |
| *** changbl has quit IRC | 19:02 | |
| *** hoangdo has quit IRC | 19:02 | |
| *** erecio has joined #openstack-ceilometer | 19:03 | |
| *** changbl has joined #openstack-ceilometer | 19:03 | |
| *** hoangdo has joined #openstack-ceilometer | 19:03 | |
| *** r3pl4y has quit IRC | 19:07 | |
| *** skelpter has joined #openstack-ceilometer | 19:08 | |
| *** Shaan7 has joined #openstack-ceilometer | 19:10 | |
| *** fnaval has quit IRC | 19:17 | |
| *** _nadya_ has joined #openstack-ceilometer | 19:17 | |
| *** ildikov has joined #openstack-ceilometer | 19:24 | |
| *** skelpter has quit IRC | 19:25 | |
| *** admin0 has quit IRC | 19:46 | |
| *** nati_ueno has quit IRC | 19:47 | |
| *** admin0 has joined #openstack-ceilometer | 19:47 | |
| *** admin0 has quit IRC | 19:48 | |
| *** _nadya_ has quit IRC | 19:49 | |
| openstackgerrit | Fabio Giannetti proposed a change to openstack/python-ceilometerclient: Update python-ceilometerclient to support Keystone V3 API https://review.openstack.org/96323 | 19:49 |
| *** admin0 has joined #openstack-ceilometer | 19:51 | |
| *** admin0 has quit IRC | 19:52 | |
| *** chmartinez has left #openstack-ceilometer | 19:53 | |
| *** nati_ueno has joined #openstack-ceilometer | 19:56 | |
| *** admin0 has joined #openstack-ceilometer | 19:59 | |
| *** fnaval has joined #openstack-ceilometer | 20:06 | |
| *** Alexei_987 has joined #openstack-ceilometer | 20:07 | |
| *** admin0 has quit IRC | 20:12 | |
| *** admin0 has joined #openstack-ceilometer | 20:13 | |
| *** admin0 has quit IRC | 20:14 | |
| *** boris-42 has quit IRC | 20:18 | |
| *** boris-42 has joined #openstack-ceilometer | 20:18 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Adds caches for image and flavor in compute agent https://review.openstack.org/98301 | 20:19 |
| openstackgerrit | A change was merged to openstack/ceilometer: Use TYPE_GAUGE rather than TYPE_CUMULATIVE https://review.openstack.org/98075 | 20:23 |
| *** admin0 has joined #openstack-ceilometer | 20:23 | |
| *** nati_ueno has quit IRC | 20:28 | |
| *** julim has quit IRC | 20:35 | |
| *** openstackgerrit has quit IRC | 20:35 | |
| *** openstackgerrit has joined #openstack-ceilometer | 20:35 | |
| *** admin0 has quit IRC | 20:39 | |
| *** nati_ueno has joined #openstack-ceilometer | 20:44 | |
| *** erecio has quit IRC | 20:53 | |
| *** admin0 has joined #openstack-ceilometer | 20:56 | |
| *** jdob has quit IRC | 20:59 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ceilometer: Updated from global requirements https://review.openstack.org/96819 | 21:01 |
| *** DrBacchus has quit IRC | 21:03 | |
| *** nati_ueno has quit IRC | 21:07 | |
| *** admin0 has quit IRC | 21:12 | |
| *** thomasem has quit IRC | 21:35 | |
| *** thomasem_ has joined #openstack-ceilometer | 21:35 | |
| *** thomasem_ has quit IRC | 21:39 | |
| openstackgerrit | A change was merged to openstack/ceilometer: remove default=None for config options https://review.openstack.org/99368 | 21:42 |
| *** ccrouch has quit IRC | 21:51 | |
| *** kun_huang has quit IRC | 21:55 | |
| *** admin0 has joined #openstack-ceilometer | 22:01 | |
| *** nati_ueno has joined #openstack-ceilometer | 22:03 | |
| *** ccrouch has joined #openstack-ceilometer | 22:04 | |
| *** ccrouch has quit IRC | 22:04 | |
| *** dmsimard is now known as dmsimard_away | 22:05 | |
| *** gordc has quit IRC | 22:07 | |
| *** ccrouch has joined #openstack-ceilometer | 22:12 | |
| *** jergerber has quit IRC | 22:12 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Adding alarm list filtering by state and meter https://review.openstack.org/96582 | 22:28 |
| *** admin0 has quit IRC | 22:33 | |
| *** robbybb111 has quit IRC | 23:03 | |
| *** fnaval has quit IRC | 23:11 | |
| *** ccrouch has quit IRC | 23:44 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!