Wednesday, 2015-08-26

« Tuesday, 2015-08-25 Index Thursday, 2015-08-27 »
*** thorst has joined #openstack-ceilometer00:11
*** ankita_wagh has quit IRC00:14
*** openstack has joined #openstack-ceilometer00:33
*** thorst has quit IRC00:33
*** _cjones_ has quit IRC00:34
*** exploreshaifali has quit IRC00:37
*** changbl has quit IRC00:42
*** llu has joined #openstack-ceilometer00:45
*** jwcroppe has quit IRC00:47
*** jwcroppe has joined #openstack-ceilometer00:47
*** jwcroppe has quit IRC00:52
*** liusheng has quit IRC00:52
*** ankita_wagh has joined #openstack-ceilometer00:57
*** rjaiswal has quit IRC01:00
*** summer has joined #openstack-ceilometer01:12
*** summer has quit IRC01:17
*** jwcroppe has joined #openstack-ceilometer01:19
*** ankita_wagh has quit IRC01:25
*** ankita_wagh has joined #openstack-ceilometer01:27
*** ankita_wagh has quit IRC01:28
*** edmondsw has joined #openstack-ceilometer01:39
*** edmondsw has quit IRC01:39
*** jwcroppe_ has joined #openstack-ceilometer01:40
*** jwcroppe has quit IRC01:42
*** nadya has joined #openstack-ceilometer01:54
*** julim has quit IRC01:57
openstackgerritRohit Jaiswal proposed openstack/ceilometermiddleware: Ignore request with swift.source in env  https://review.openstack.org/21690302:02
*** nadya has quit IRC02:09
*** boris-42 has quit IRC02:20
*** ddieterly has quit IRC02:46
*** r-mibu has joined #openstack-ceilometer03:19
*** jwcroppe_ is now known as jwcroppe03:35
openstackgerritRyota MIBU proposed openstack/aodh: Fix arg for messaging.get_transport  https://review.openstack.org/21694503:40
*** deepthi has joined #openstack-ceilometer03:45
*** links has joined #openstack-ceilometer03:45
*** boris-42 has joined #openstack-ceilometer03:46
*** changbl has joined #openstack-ceilometer03:46
openstackgerritRyota MIBU proposed openstack/aodh: Fix args for get_notification_listener()  https://review.openstack.org/21694503:46
*** ddieterly has joined #openstack-ceilometer03:46
*** ddieterly has quit IRC03:52
openstackgerritRyota MIBU proposed openstack/aodh: event-alarm: devstack plugin support  https://review.openstack.org/21694703:53
r-mibuhi03:56
r-mibuceilomter gate seems to be broken03:57
r-mibucaused by WSME 0.8?03:57
r-mibuhttps://review.openstack.org/#/c/216795/03:57
*** fawadkhaliq has joined #openstack-ceilometer03:58
openstackgerritliusheng proposed openstack/ceilometer: Avoid posting extra unknown attributes to API request  https://review.openstack.org/21695304:07
openstackgerritliusheng proposed openstack/ceilometer: Avoid posting extra unknown attributes to API request  https://review.openstack.org/21695304:09
*** liusheng has joined #openstack-ceilometer04:09
*** ildikov has quit IRC04:30
*** ddieterly has joined #openstack-ceilometer04:48
*** ddieterly has quit IRC04:52
*** Ephur has quit IRC04:57
openstackgerritMerged openstack/ceilometer: Update tests to reflect WSME 0.8 fixes  https://review.openstack.org/20846705:10
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements  https://review.openstack.org/21696405:14
*** yprokule has joined #openstack-ceilometer05:19
*** RaySun has joined #openstack-ceilometer05:21
*** summer has joined #openstack-ceilometer05:44
openstackgerritRohit Jaiswal proposed openstack/ceilometer: Fixed identity trust event types  https://review.openstack.org/21697905:48
*** ddieterly has joined #openstack-ceilometer05:49
openstackgerritRohit Jaiswal proposed openstack/ceilometer: Fixed identity trust event types  https://review.openstack.org/21697905:52
*** ddieterly has quit IRC05:54
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: rest: implements /resource/generic/UUID/history  https://review.openstack.org/21665305:55
*** ankita_wagh has joined #openstack-ceilometer05:56
*** ankita_wagh has quit IRC05:57
*** ankita_wagh has joined #openstack-ceilometer05:58
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: integration: add some new tests  https://review.openstack.org/21560405:58
*** ankita_wagh has quit IRC06:00
*** ankita_wagh has joined #openstack-ceilometer06:00
openstackgerritJuPing proposed openstack/ceilometer: Fix description for "Incorrect spelling of a word"  https://review.openstack.org/21698606:05
*** afazekas has joined #openstack-ceilometer06:08
*** RaySun has quit IRC06:20
*** ddieterly has joined #openstack-ceilometer06:49
*** ddieterly has quit IRC06:54
openstackgerritMerged openstack/gnocchi: rest: Pass the project name to middleware config  https://review.openstack.org/21661606:54
*** ankita_wagh has quit IRC07:08
*** fawadkhaliq has quit IRC07:15
*** fawadkhaliq has joined #openstack-ceilometer07:34
*** Ala has joined #openstack-ceilometer07:35
*** fawadkhaliq has quit IRC07:38
*** ddieterly has joined #openstack-ceilometer07:50
*** shardy has joined #openstack-ceilometer07:52
*** cdent has joined #openstack-ceilometer07:53
*** fawadkhaliq has joined #openstack-ceilometer07:54
*** ddieterly has quit IRC07:55
openstackgerritChris Dent proposed openstack/ceilometer: Add cpu meters to yaml  https://review.openstack.org/21536707:55
openstackgerritChris Dent proposed openstack/ceilometer: Move profiler meters to yaml  https://review.openstack.org/21383207:56
openstackgerritChris Dent proposed openstack/ceilometer: support custom timestamp  https://review.openstack.org/21536107:59
openstackgerritChris Dent proposed openstack/ceilometer: Sync devstack plugin with devstack:lib/ceilometer  https://review.openstack.org/21672108:02
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: WIP: gnocchi: use events to crud resources  https://review.openstack.org/21701708:02
*** spradeepv has quit IRC08:03
*** fawadkhaliq has quit IRC08:04
*** jwcroppe has quit IRC08:16
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: WIP: gnocchi: use events to crud resources  https://review.openstack.org/21701708:16
*** fawadkhaliq has joined #openstack-ceilometer08:17
openstackgerritJulien Danjou proposed openstack/gnocchi: file: store measures atomically  https://review.openstack.org/21412308:17
openstackgerritJulien Danjou proposed openstack/gnocchi: file: use _get_tempfile() in metric storage  https://review.openstack.org/21412408:17
openstackgerritJulien Danjou proposed openstack/gnocchi: rest: remove convert_metric()  https://review.openstack.org/21523508:17
openstackgerritJulien Danjou proposed openstack/gnocchi: archive policy rule: make them available on all metric creations  https://review.openstack.org/21523608:17
*** nadya has joined #openstack-ceilometer08:18
*** nadya has quit IRC08:19
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: WIP: gnocchi: use events to crud resources  https://review.openstack.org/21701708:20
* cdent opens a wsme can of worms08:27
openstackgerritJuPing proposed openstack/ceilometer: Fix description for "Inapt spelling of 'MongoDB'"  https://review.openstack.org/21702908:29
*** ddieterly has joined #openstack-ceilometer08:51
*** ddieterly has quit IRC08:55
*** ildikov has joined #openstack-ceilometer08:56
openstackgerritMerged openstack/aodh: event-alarm: devstack plugin support  https://review.openstack.org/21694709:11
*** h00327910__ has joined #openstack-ceilometer09:14
*** eglynn has joined #openstack-ceilometer09:25
*** ildikov has quit IRC09:28
openstackgerritMerged openstack/gnocchi: file: store measures atomically  https://review.openstack.org/21412309:28
*** ildikov has joined #openstack-ceilometer09:28
openstackgerritMerged openstack/gnocchi: file: use _get_tempfile() in metric storage  https://review.openstack.org/21412409:29
silehtI received compute.instance.update before compute.instance.create.start awesome ...09:30
openstackgerritMerged openstack/gnocchi: rest: remove convert_metric()  https://review.openstack.org/21523509:30
openstackgerritMerged openstack/gnocchi: archive policy rule: make them available on all metric creations  https://review.openstack.org/21523609:30
*** goodygum has joined #openstack-ceilometer09:38
*** ddieterly has joined #openstack-ceilometer09:52
*** ddieterly has quit IRC09:57
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: WIP: gnocchi: use events to crud resources  https://review.openstack.org/21701709:57
*** ildikov has quit IRC10:05
*** Kennan2 has joined #openstack-ceilometer10:16
*** Kennan has quit IRC10:16
openstackgerritMerged openstack/gnocchi: indexer: raise an error if deleting a non-existent metric  https://review.openstack.org/21421710:17
openstackgerritMerged openstack/gnocchi: rest: implements /resource/generic/UUID/history  https://review.openstack.org/21665310:17
*** fawadkhaliq has quit IRC10:40
openstackgerritChris Dent proposed openstack/ceilometer: Create conf directory during devstack install phase  https://review.openstack.org/21708110:40
openstackgerritChris Dent proposed openstack/aodh: Create conf directory during install phase  https://review.openstack.org/21708210:40
openstackgerritChris Dent proposed openstack/gnocchi: Create conf directory during install phase  https://review.openstack.org/21708310:41
*** fawadkhaliq has joined #openstack-ceilometer10:50
openstackgerritChris Dent proposed openstack/aodh: Create conf directory during devstack install phase  https://review.openstack.org/21708210:51
openstackgerritRoman Dobosz proposed openstack/ceilometer: Versioned object for Meter and Sample model  https://review.openstack.org/19052110:52
openstackgerritRoman Dobosz proposed openstack/ceilometer: Versioned object initial patch  https://review.openstack.org/17867510:52
openstackgerritRoman Dobosz proposed openstack/aodh: Versioned object initial patch  https://review.openstack.org/21624310:52
*** ddieterly has joined #openstack-ceilometer10:52
*** ddieterly has quit IRC10:57
openstackgerritJuPing proposed openstack/ceilometer: Fix description for "Inapt spelling of 'MongoDB'"  https://review.openstack.org/21702911:01
jd__sileht: can you check if that bug statuses are up to date https://bugs.launchpad.net/aodh/+bug/1479429 ?11:03
openstackLaunchpad bug 1479429 in Gnocchi "aggregation: issue when one of the resource is no more updated" [Medium,Triaged] - Assigned to Mehdi Abaakouk (sileht)11:03
*** ildikov has joined #openstack-ceilometer11:13
*** summer has quit IRC11:15
*** summer has joined #openstack-ceilometer11:15
*** deepthi has quit IRC11:17
*** summer has quit IRC11:20
*** exploreshaifali has joined #openstack-ceilometer11:20
*** thorst has joined #openstack-ceilometer11:23
silehtjd__, yes I still work on it, the crud of gnocchi resource via events dispatcher should fix the issue11:29
*** fawadkhaliq has quit IRC11:30
*** fawadkhaliq has joined #openstack-ceilometer11:33
*** liusheng has quit IRC11:38
*** shardy_ has joined #openstack-ceilometer11:38
*** shardy has quit IRC11:39
*** shardy_ has quit IRC11:43
*** shardy has joined #openstack-ceilometer11:44
*** gordc has joined #openstack-ceilometer11:46
*** h00327910__ has quit IRC11:53
*** ddieterly has joined #openstack-ceilometer11:53
*** ddieterly has quit IRC11:58
*** dikonoor has joined #openstack-ceilometer12:02
*** MVenesio has joined #openstack-ceilometer12:02
cdentgordc: as far as I can tell the wsme stuff cleared okay this morning, and the grenade switchover is okay too12:08
cdentI sent off a fair few rechecks12:08
gordccdent: awesome. thanks for that.12:08
gordci was waiting last night but it hit 1am and.... i passed out.12:08
cdentyeah, big pile of slow last night12:09
*** Qlawy has quit IRC12:09
gordcwe should probably prioritise reviews to bps first now. we can continue with bugs after liberty-312:10
gordcreally just rbac stuff dikonoor is working on... and the remaining mandatory limit patch12:11
gordcllu: i guess we should bump the declarative snmp work?12:11
openstackgerritPradeep Kilambi proposed openstack/ceilometer: Add validation tests for arithmetic, string and prefix expressions  https://review.openstack.org/21061912:23
openstackgerritPradeep Kilambi proposed openstack/ceilometer: Change json path's to start with $. for consistency  https://review.openstack.org/21633012:23
*** jasonamyers has joined #openstack-ceilometer12:27
*** ddieterly has joined #openstack-ceilometer12:30
*** jaosorior has joined #openstack-ceilometer12:31
*** shardy_ has joined #openstack-ceilometer12:34
*** shardy has quit IRC12:35
gordcdikonoor: any help needed on rbac work?12:36
dikonoorgordc: hi..yes..12:36
*** pradk has joined #openstack-ceilometer12:37
gordcdikonoor: cool. whats up? just so you're aware. we're freezing code next week.12:37
dikonoorgordc: https://review.openstack.org/#/c/211494/ . I got my functional tests with and without debug running .So thats good..I figured out that this breaks build because of use of temp file12:37
dikonoorgordc: you had suggested that I use a temp file ..12:38
dikonoorgordc: The temp file never gets created..I tried with creating a file under ceilometer/tests/functional/policy_fixture.json and that works12:39
dikonoorgordc: Is there a real need to use a temp file?12:39
*** shardy_ has quit IRC12:39
dikonoorgordc: Can I just create one somewhere under ceilometer tests and get it running12:39
openstackgerritRyota MIBU proposed openstack/aodh: Fix args for get_notification_listener()  https://review.openstack.org/21694512:39
*** shardy has joined #openstack-ceilometer12:40
*** edmondsw has joined #openstack-ceilometer12:40
gordcdikonoor: taking a look now12:41
dikonoorgordc: ok..12:41
gordcwe shouldn't need to create a real file for tests. (ideally)12:41
dikonoorgordc: I understand . But using the temp thing just doesn't work..It's better to have some tests than not have (There are many rules in the policy.json today but no tests for any of those)12:43
*** shardy has quit IRC12:44
dikonoorgordc: also the current set of functional tests anyway use the policy.json (non-temp file) under etc/ceilometer12:45
dikonoorgordc: so why can't we create one for test purposes under ceilometer/tests/*12:45
*** shardy has joined #openstack-ceilometer12:45
gordcdikonoor: i'll take a look. which patches do you require to be merged?12:46
gordcdikonoor: they all seem to touch same files but have different changes.12:47
dikonoorgordc: yes.. https://review.openstack.org/#/q/status:open+project:openstack/ceilometer+branch:master+topic:bp/events-rbac,n,z12:47
gordcedmondsw: i missed your question during a meeting yesterday12:49
gordcedmondsw: https://review.openstack.org/#/c/216889/ should fix your issue... i think it's a real bug12:49
edmondswgordc, tx, I'll take a look12:50
*** annasort has joined #openstack-ceilometer12:50
openstackgerritDivya K Konoor proposed openstack/ceilometer: Control Events RBAC from policy.json  https://review.openstack.org/21149412:51
*** ddieterly has quit IRC12:57
openstackgerritDivya K Konoor proposed openstack/ceilometer: Control Events RBAC from policy.json  https://review.openstack.org/21149412:59
*** jwcroppe has joined #openstack-ceilometer13:00
dikonoorgordc:I added a new patch to https://review.openstack.org/#/c/211494/ . I replaced the tmp file with a permanent file13:01
*** eandersson has joined #openstack-ceilometer13:04
eanderssonHey13:05
eanderssonI am seeing two errors in Ceilometer and I was wondering if these two are caused by something in the configuration.13:05
eandersson> ceilometer.hardware.discovery [-] Couldn't obtain IP address of instance13:05
eanderssonand I am also seeing a lot of these > OperationFailure: cursor id 'xxxxxxxxxxxxxxxxxx' not valid at server13:06
gordceandersson: can you paste (edited for privacy) errors to paste.openstack.org? one of us might be able to help...13:07
gordcalternatively you can open a bug.13:07
* gordc is going to drive to work now (be back in a bit)13:07
gordcdikonoor: will get back to you in a bit.13:07
*** gordc has quit IRC13:07
dikonoorgordc:ok13:08
eanderssonsure -> http://paste.openstack.org/show/4VZFkmctZz100VkAaySH/13:08
cdentdikonoor: can you rebase your stuff to master? I'm trying to run it against the existing tox -efunctional target and without some fixes related to WSME changes it shows a fair few failures which are unrelated to your changes13:13
dikonoorcdent: ok..let me do that13:14
*** links has quit IRC13:15
cdenteandersson: the hardware discovery message is sort of normal, more of a warning than an error13:16
cdenteandersson: on the cursor id message: which log file is that from and are you able to correspond it with an api request or something of that nature? It looks like the mongodb server may have timed out13:17
eanderssonI know that this sometime happens when I try to view the resource page in Horizon13:19
eanderssonWhich by the way takes over 10 minutes to load when it doesn't error out instantly with this error. =]13:20
cdenteandersson: ah okay, if I recall correctly not-very-recent versions of horizon make open-ended queries (without timestamps) to make their data presentation13:21
eanderssonI am actually using trunk for Horizon13:21
eanderssonCeilometer on the otherhand is running Kilo13:21
cdentah13:21
cdenta) maybe didn't do the change I thought it did, b) trunk ceilo now implicitly forces a limited query13:21
cdentso if you have the option to try turnk ceilo you might see if that makes a difference13:22
eanderssonthat is nice13:22
eanderssonthe log file is app.log btw13:22
*** julim has joined #openstack-ceilometer13:22
openstackgerritIlya Tyaptin proposed openstack/ceilometer: [ES] Implement event time to live  https://review.openstack.org/20057313:23
eanderssonIs there a way to reduce the risk of the mongodb timing out?13:26
eanderssonNot really an mongodb expert unfortunately :p13:26
cdenteandersson: are you just testing or is this for a real deployment?13:28
cdentthe usual tricks are: make sure mongo has a _ton_ or ram, make sure it has its own machine/disk, make sure that you are only recording the meters that you actually need (by editing pipeline.yaml), make sure that you are only recording meters at the interval that is required (also by editing pipeline.yaml)13:29
cdenttrunk ceilo has some improvements to how mongodb does big queries13:29
openstackgerritRyota MIBU proposed openstack/aodh: Fix args for get_notification_listener()  https://review.openstack.org/21694513:30
ityaptincdent, eandersson: I'm affraid that Horizon issue is not caused only by mongodb.13:30
ityaptincdent, eandersson: Code that uses Ceilo in Horizon is not so good. It makes N (N=resource_count) statistics requests in 100 threads which block ceilometer api for a long time.13:33
eanderssonok perfect thnaks13:35
eanderssonIt's kinda what I figured, but was hoping for a magical fix :p13:35
cdenteandersson: oh we have that, but you're not allowed to have it13:36
cdentcuz it's magic13:36
eanderssonhaha13:36
eanderssonbtw are there any recommended dashboards or reporting tools that is based on Ceilometer ? or in general any documentation on how to make good use of Ceilometer?13:36
* cdent looks at everyone else13:36
cdentI've been asking that question for about a year13:36
eanderssonheh13:36
eanderssonlooking at some commercial solutions, but ideally I would prefer something open =]13:37
*** ddieterly has joined #openstack-ceilometer13:38
openstackgerritDivya K Konoor proposed openstack/ceilometer: Add user_id,project_id traits to audit events  https://review.openstack.org/21150913:38
*** Ephur has joined #openstack-ceilometer13:41
openstackgerritDivya K Konoor proposed openstack/ceilometer: WIP Events RBAC needs scoped token  https://review.openstack.org/21255713:41
*** annasort_ has joined #openstack-ceilometer13:43
*** annasort has quit IRC13:43
*** annasort_ is now known as annasort13:43
*** Ephur has quit IRC13:46
*** gordc has joined #openstack-ceilometer13:49
*** annasort has quit IRC13:52
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements  https://review.openstack.org/21696413:53
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometermiddleware: Updated from global requirements  https://review.openstack.org/21224513:53
*** fawadkhaliq has quit IRC13:56
openstackgerritRyota MIBU proposed openstack/python-ceilometerclient: refactor: move 'repeat_actions' to common args  https://review.openstack.org/21715913:57
*** Kennan2 has quit IRC14:01
*** Kennan has joined #openstack-ceilometer14:02
*** fawadkhaliq has joined #openstack-ceilometer14:02
*** links has joined #openstack-ceilometer14:05
*** nadya has joined #openstack-ceilometer14:07
*** rbak has joined #openstack-ceilometer14:11
*** ddieterly has quit IRC14:11
*** nadya has quit IRC14:11
*** annasort has joined #openstack-ceilometer14:15
*** annasort_ has joined #openstack-ceilometer14:18
*** thumpba has joined #openstack-ceilometer14:19
*** annasort has quit IRC14:20
*** annasort_ is now known as annasort14:20
*** thumpba has quit IRC14:20
*** thumpba has joined #openstack-ceilometer14:20
*** fawadkhaliq has quit IRC14:35
edmondswgordc, looking at https://review.openstack.org/#/c/216889/ and having doubts. I was thinking the request_id should be aligned with the request_id you'd find in the API logs for the service that is being audited, but seems that is not the case.14:53
edmondswI found an "id" trait does match up between audit request/responses already, but I can't tell where the value comes from. Maybe that is what I was looking for?14:53
gordcedmondsw: well the request_id currently in traits is not that, it defaults to the context of message14:54
edmondswwhat is "context of message"?14:54
gordcbut yes, there is an id and also a correlation_id that would match req/response messages14:54
edmondswoh nm14:54
gordcwhen notifications are published, they are published under a defined context.14:55
edmondswwhat is the difference between id and correlation_id?14:55
gordcid is unique identifier for a single event (in audit middleware there's a single event shared between req and response)14:55
edmondswhow is that possible? ceilometer splits a single event into 2 events?14:56
gordccorrelation_id is designed to match across different events.. similar to request_id. so if you create an instance all the events from nova, glance, neutron, would all theorectically have same correlation_id but different event id14:56
*** Ephur has joined #openstack-ceilometer14:58
gordcedmondsw: no. ceilometer doesn't do anything. in audit middleware, we track a single request event in two parts: incoming and response. it's a single event. but audited at different stages14:58
gordcthat's the purpose of reporter attribute... so a single event can be audited at different/multiple stages.14:59
*** links has quit IRC14:59
edmondswI'm not following14:59
ildikovcdent: hey15:00
ildikovcdent: did you finally join the Monasca meeting yesterday?15:00
gordcit's how CADF is designed. so say we have a request event.15:00
edmondswwhen I ask ceilometer for events, I get two... one for the request and another for the response. So there are 2 events. But you're saying there is one. So we must be using the word "event" to mean 2 different things?15:00
cdentildikov: no, webex was a killer for me, but at the tc meeting last night they were told they need to use irc for meetings, so starting next week they are15:01
gordcedmondsw: right. so it's one "api request" event.15:01
gordcedmondsw: but there's two discrete, auditable steps to it.15:01
*** fawadkhaliq has joined #openstack-ceilometer15:02
edmondswI understand it's tied to a single api request... you just mean "event" in the real world sense, rather than as a code construct?15:02
ildikovcdent: yeap, webex can be tough...15:03
edmondswgordc, I think I start to follow, in that case15:03
gordcright. if you look at the req and corresponding response in ceilometer event, it's actually the same data except the response entry has been modified to include response code, outcome, (a few more items)15:03
ildikovcdent: will that then be IRC only?15:03
cdentI think so15:03
cdentthe reason is that tc requires meetings to be logged15:03
cdent(to be in big tent)15:04
edmondswgordc, right15:04
ildikovok, cool, tnx for info!15:04
edmondswgordc, so I think we should abandon your change15:04
cdentildikov: there was posting on os-dev earlier today about changing the meeting time to wednesday15:04
ildikovyeap right, I read it somewhere15:04
edmondswgordc, correlation_id: can tie together multiple API requests, id: can tie together request and response from a single API request, request_id: uniquely to the audit notification (either a request or response... not shared between the 2)15:05
gordcedmondsw: well it's sort of valid still, in context of keystonemiddleware messages, both req and response should be tied together, they aren't separate messages from different requests.15:05
edmondswand id does that, right?15:06
edmondswthey are separate notificates, though, aren't they15:06
gordcfor CADF message yes.15:06
ildikovcdent: found it, thanks15:06
gordcthe request_id here or context we use for notifications is a completely separate item from auditing and CADF messages15:07
*** r-daneel has joined #openstack-ceilometer15:07
edmondswgordc, sort of15:07
edmondswgordc, they're tied together in the sense that keystonemiddleware.audit only ever does auditing/CADF messages, so this code is specific to that15:08
edmondswwould you want other places sending multiple notifications with the same request_id between them?15:09
gordcright. but the context part is an oslo.messaging concept.15:09
edmondswok, I leave it to your judgement15:09
gordcso for instance, our notifications we use for IPC in ceilometer services, it's not a new context for every single notification.15:10
*** boris-42 has quit IRC15:10
edmondswsure15:10
gordcedmondsw: cool cool. but yeah, id/correlation_id's would be the correct way to match up req/response cadf messages :)15:11
edmondswgordc tx :)15:11
edmondswoh, that brings me to something else I was going to say... id don't see correlation_id in the events15:12
edmondswit's not in the event def yaml, and I don't see it hardcoded... which should it be?15:12
edmondswlooks like it's specific to keystonemiddleware.audit, so event def yaml then?15:12
gordci probbaly missed it. it should be added in event def yaml.15:12
edmondswcool... I'll open a bug and throw up a commit15:13
*** jaypipes has joined #openstack-ceilometer15:13
gordcedmondsw: cool cool. thanks15:13
cdentgordc: just a heads up: I've made the ceilometer devstack plugin related reviews non-wip:15:14
cdenthttps://review.openstack.org/196383 and https://review.openstack.org/19644615:14
gordccdent: we ready to merge and do our own thing?15:14
cdentI've been using it for weeks now15:14
cdentthe tricky part is making sure the infra things do what's expected15:15
gordc...good luck with that.15:15
cdentand we also need to make sure that people using ceilometer in their own gate tests make some adjustments15:15
cdentI guess I need to send some mail15:15
gordccdent: shall we wait until monday (or until we're sure no other bps need to be merged)?15:16
gordcin case it all goes to hell?15:16
cdentif you prefer, I don't really mind15:16
cdentI don't expect the devstack part to be hellish15:17
cdentbut there may be issues with other projects15:17
*** vishwanathj has joined #openstack-ceilometer15:19
gordccdent: can you send mail before merging or is it dependent?15:19
cdentI can send mail before15:20
cdentbecause it is a warning15:20
*** changbl has quit IRC15:21
gordck, maybe send mail first and if not one yells we can proceed? i'd probably still aim for monday just so we less hurdles for specs/bugs in this last week.15:21
cdentnoted15:21
*** MVenesio has quit IRC15:31
openstackgerritgordon chung proposed openstack/ceilometer: WIP Events RBAC needs scoped token  https://review.openstack.org/21255715:32
gordcdikonoor: ^^ i fixed your tests. i told you to inherit wrong test class. apologies.15:32
openstackgerritgordon chung proposed openstack/ceilometer: Events RBAC needs scoped token  https://review.openstack.org/21255715:33
*** afazekas has quit IRC15:40
*** fawadkhaliq has quit IRC15:43
-openstackstatus- NOTICE: restarting gerrit due to a slow memory leak15:43
gordc... 9 hrs of queuing.. so close.lol15:45
cdentgordc: i wipped the changes and have sent some mail15:47
gordccdent: cool cool15:52
*** exploreshaifali has quit IRC15:52
gordcthe ordering for decorators go from outer to inner right?15:53
EmilienMgordc: I go back to bed15:53
gordcEmilienM: ... i think everyone realises not we have have an #idiots channel15:54
gordcs/not/now/15:54
EmilienMgordc: ahah15:54
EmilienMgordc: no actually, I go back to bed because gate is slow15:54
EmilienMbut it's a joke :)15:54
gordcyou should go to bed. it' more productive15:54
*** adurbin_ has joined #openstack-ceilometer15:55
EmilienMgordc: ok15:58
openstackgerritgordon chung proposed openstack/ceilometer: Events RBAC needs scoped token  https://review.openstack.org/21255715:59
*** changbl has joined #openstack-ceilometer16:00
*** Ala has quit IRC16:13
*** fawadkhaliq has joined #openstack-ceilometer16:23
*** _cjones_ has joined #openstack-ceilometer16:26
dikonoorgordc:thanks for fixing my tests..16:26
gordcdikonoor: np, i sent you down wrong path.16:29
dikonoorgordc: could you take a look at this and let me know your comments >> https://review.openstack.org/#/c/211509/16:30
gordcsure16:36
*** exploreshaifali has joined #openstack-ceilometer16:50
*** cdent has quit IRC16:54
*** jasonamyers has quit IRC16:54
*** yprokule has quit IRC16:57
*** boris-42 has joined #openstack-ceilometer17:07
*** ankita_wagh has joined #openstack-ceilometer17:07
openstackgerritRohit Jaiswal proposed openstack/ceilometermiddleware: Fixes parsing req path for account HEAD/GET  https://review.openstack.org/21730917:08
openstackgerritMerged openstack/ceilometer: gnocchi: move to jsonpath_rw_ext  https://review.openstack.org/21601217:13
pradkgood to see things getting merged.. gate's been so slow17:33
gordcpradk: i appreciate your optimism.17:42
gordcit's foolish... but it's appreciated.lol17:42
*** eandersson has quit IRC17:49
*** exploreshaifali has quit IRC17:58
openstackgerritDivya K Konoor proposed openstack/ceilometer: Add user_id,project_id traits to audit events  https://review.openstack.org/21150917:58
dikonoorgordc: gordc: implemented your comment and uploaded a new patch https://review.openstack.org/#/c/211509/18:00
dikonoorgordc: on a separate note, we had talked earlier about this scenario - admins should be able to see events for their project + all events with empty/null/no project ..The current trait filter query doesn't have the support to run a OR condition .18:01
gordcdikonoor: right. that would be support you'd need to add in...18:02
dikonoorgordc: If the code freeze is next week , I don't think I will be able to add that support (the OR filter support).18:02
gordcdikonoor: we can request an extension... assuming you're comfortable working on it (it would involve all drivers.18:03
gordcdid you have another idea?18:03
*** Samolo has joined #openstack-ceilometer18:04
SamoloHi all18:04
dikonoorgordc: not really..although wondering if anyone else in the community would be able to take that up.. The other option was to run multiple queries and combine them but that has two problems -1) performance 2) limit18:05
SamoloRunning Kilo with ceilometer, My pipeline is set with an intervall of 600sec (10min), when I run a sample list I have some sample every 10min, thats good, but I also get some sample every hour (14h00, 15h00, 16h00 ......)18:06
SamoloBut I didnt set any pulling at 1h interval18:07
SamoloI've read my pipeline 10000 time and I dont see where this 1hour sample come from18:07
gordcdikonoor: i can try looking tomorrow. i think we need it to move forward with your patches18:07
gordcSamolo: in addition to ceilometer, the other services emit their own periodic meters.18:08
gordcSamolo: by default, nova will emit information about system every hour.18:09
dikonoordikonoor: if you can, that'd be great..18:09
SamoloWhen I use sample list with the deprecated instance:<flavor> meter, I get sample every 10min and every hour, but when I use the new ceilometer sample-list -m instance -q metadata.instance_type=<value>, the returned sample are just the one every hour18:10
SamoloI'm lost18:10
*** annasort has quit IRC18:10
openstackgerritDivya K Konoor proposed openstack/ceilometer: Add user_id,project_id traits to audit events  https://review.openstack.org/21150918:10
Samologordc: thanks I thing you get something interesting18:10
Samologordc: do you know where can I find the flag that use nova to set at 1h please?18:11
gordcSamolo: the second one is coming from Nova's periodic notifications (trying to find it in docs)18:11
dikonoorgordc: updated events.rst https://review.openstack.org/#/c/211509/18:11
*** jasonamyers has joined #openstack-ceilometer18:12
Samologordc: thanks for your time18:12
gordcSamolo: there's an instance_usage_audit_period config option18:12
gordchttp://docs.openstack.org/kilo/config-reference/content/list-of-compute-config-options.html18:12
gordcalthough it seems like it defaults to monthly.18:12
Samologordc: thanks, I dont see this flag on my controllers18:13
dikonoorgordc: https://review.openstack.org/#/c/211494/ works fine when i run with debug with pdb in my local but fails wit sudo tox -e functional.. I am sure it has something to do with that policy file not getting loaded18:14
gordcdikonoor: looking at it now. i'm pretty sure it's broken (just that it's hiding it in debug)18:15
*** openstackgerrit has quit IRC18:17
*** openstackgerrit has joined #openstack-ceilometer18:17
openstackgerritvenkatamahesh proposed openstack/ceilometer: Path of the links are updated  https://review.openstack.org/21733518:17
Samologordc: If you want this two command explain my question better http://pastebin.com/es4xEYSb18:18
dikonoorgordc:most likely it's broken and i am unable to debug it without my debugger helping .. in debug mode, the test is able to find the policy_fixture in the right path (when I used debugger with temp file, the file couldn;t be found) and the response has the right status and message18:18
SamoloThe old and new command doesnt return the same sample18:18
Samologordc: note that the first command is the new, the second is the deprecated18:19
SamoloMy goal is to set the polling returned by the first command to 10min and not 1hour18:20
gordcSamolo: let me check how we build metadata when we poll18:22
Samologordc: thanks a lot18:22
gordcSamolo: are you using ceilometer for resource health monitoring? i should mention, that will probably cause your db to implode.18:23
openstackgerritDivya K Konoor proposed openstack/ceilometer: Add user_id,project_id traits to audit events  https://review.openstack.org/21150918:23
Samologordc: actually its for billing purpose18:24
*** gaud has joined #openstack-ceilometer18:24
gordcSamolo: just curious but what value in instance meter are you using? being that meter is deprecated as well.18:25
gordcSamolo: is it possible for you to view the metadata for the 10min polled samples? it seems like there should be an instance_type for those values18:27
Samologordc: I'm not sure that I understand your question but its the instance flavor that I want18:27
Samoloto bill specificly each flavor18:27
Samoloaccording to the time run with this flavor18:27
Samologordc: How can I view the metadata for a sample please ?18:29
SamoloI dont see any sample id or anything18:29
Samolo(my bad must be a noob question, just staring with ceilo)18:29
Samolostarting*18:29
*** jaypipes has quit IRC18:31
Samologordc: Ok I do find an ID18:32
Samolodidnt see them at first :s18:32
gordcSamolo: :) no worries.18:33
Samologordc: sorry it was a mixed up,  I do not find sample ID when doing sample-list, I mixed up with the meter-list, that map an ID to each meter18:35
SamoloSo I have find each meter ID, but no sample ID18:35
*** rbak_ has joined #openstack-ceilometer18:35
SamoloSo I dont know how to see the metadata for a 10min poll sample18:36
gordcSamolo: odd. i'll need to start up an environment18:38
*** rbak has quit IRC18:38
*** cdent has joined #openstack-ceilometer18:40
cdentgordc: dood: rejected!18:40
Samologordc: oh ok, sorry18:40
gordcyeah :(18:40
*** jaypipes has joined #openstack-ceilometer18:41
gordclist18:41
gordcSamolo: try dropping the -m part of query18:42
ildikovcdent: gordc: pradk: hey. any sessions accepted on your side?18:44
Samologordc: I thinks I have found the pb18:45
gordcildikov: no dice.18:45
ildikovI found a Ceilosca one and one more including Ceilo18:45
Samologordc: you where rigth I have an instance_usage_audit_period = hour18:45
ildikovit's not my lucky summit either :(18:45
Samolobut It was not on the controller but on the compute18:45
pradkildikov, yea bummer18:46
gordcildikov: list is posted?18:46
pradkyea18:46
pradkhttps://www.openstack.org/summit/tokyo-2015/schedule/main-conference18:46
pradkgordc, its sad to see barely any ceilo sessions accepted this time18:46
ildikovgordc: yeap, as pradk linked above18:46
ildikovpradk: :(18:46
Samologordc: I removed the -m part http://pastebin.com/uQRr0Q2X18:48
SamoloThe first command is with -m and the second without18:49
gordccan you run ceilometer sample-show f406731e-4c13-11e5-9937-002590d3517318:49
gordcyou might want to make sure whatever you show me doesn't contain ip/pw18:50
gordci don't know what's in the metadata :)18:50
*** jwcroppe has quit IRC18:50
Samologordc: Yes thanks to notice :)18:51
cdentreading the schedule looks like they are being very “here’s openstack actually doing stuff” oriented18:53
cdentwhich is a fair play18:53
cdentgiven that attendance is swinging away from contributors18:53
cdentand packagers and the like18:53
Samologordc: sample show as requested : http://pastebin.com/eQ2pTCcj18:53
gordcSamolo: whoops. i asked for the sample that already matches.18:55
Samologordc: no pb :p wich one do you wanted ?18:56
gordcSamolo: in you sample-list query, can you also drop metadata.instance_type=Popular_118:58
*** dikonoor has quit IRC18:59
Samologordc: there is the query withoud metadata.instance_type http://pastebin.com/rUmbjUCj19:00
gordcSamolo: ok, try sample-show 5ae2db72-4c23-11e5-9bb1-0025904bfa7c19:01
gordcSamolo: apologies on the jumping around. i'm too lazy to start up an environment and you haven't complained yet :)19:02
Samologordc: I can't complained about somehone that is helping me19:02
Samologordc: its not fair :p19:02
Samologordc: there is the sample-show http://pastebin.com/R1XKYaWw19:04
gordcSamolo: hmm... yeah that's what i figured. the instance_type is an id from polling samples and a name in periodic nova notifications19:06
gordcSamolo: that is a bug. they should be consistent19:06
Samologordc: oh good here we go :)19:07
*** exploreshaifali has joined #openstack-ceilometer19:07
Samologordc: is it a know bug that I can follow the fix or ...?19:08
gordcSamolo: opening one now.19:08
*** jwcroppe has joined #openstack-ceilometer19:08
gordcSamolo: i think you could probably use complex queries to track both instance_type value OR flavour.name... but it should be fixed regardless.19:09
Samologordc: thanks a LOT, can you please send me the opened bug once its done so I can follow the fix ?19:10
gordcSamolo: for tracknig purposes, what version of Ceilometer are you using?19:10
Samolo1.0.13 in Kilo19:10
gordcSamolo: ok.19:11
gordci shoudl also mention there are a few authentication issues with the 1.0.13 client so you may want to ugprade19:12
Samologordc: yhanks for the information19:12
gordcSamolo: https://bugs.launchpad.net/ceilometer/+bug/148913219:14
openstackLaunchpad bug 1489132 in Ceilometer "instance_type metadata is different between notifications and polling" [High,Triaged]19:14
gordcfeel free to add information19:14
* cdent goes back to sleep19:16
* cdent waves19:16
*** cdent has quit IRC19:16
Samologordc: Thanks a lot for your time19:17
Samologordc: I've to go to a meeting, thanks again ! very appreciated :)19:17
gordcSamolo: no problem.19:17
gordchttps://blog.berendt.io/querying-ceilometer/ this might help to find an existing workaround19:17
Samologorc: thanks !19:18
gordccheers19:19
*** jasonamyers has quit IRC19:25
*** jasonamyers has joined #openstack-ceilometer19:25
openstackgerritgordon chung proposed openstack/ceilometer: Control Events RBAC from policy.json  https://review.openstack.org/21149419:25
*** adurbin_ has quit IRC19:34
*** marcinzzz has joined #openstack-ceilometer19:35
marcinzzzHello guys19:35
marcinzzzCould anybody point me how to limit number of meters collected by Ceilometer?19:36
marcinzzzI know I have to change pipeline.yaml on the computes, but I am not sure what to change19:36
marcinzzzplease give me some pointers19:36
*** Ephur has quit IRC19:39
pradkmarcinzzz, this should help http://docs.openstack.org/developer/ceilometer/configuration.html#id119:40
*** Ephur has joined #openstack-ceilometer19:41
openstackgerritgordon chung proposed openstack/ceilometer: Control Events RBAC from policy.json  https://review.openstack.org/21149419:44
*** chinmay_g has quit IRC19:54
*** thumpba has quit IRC19:55
openstackgerritMerged openstack/ceilometer: Add cpu meters to yaml  https://review.openstack.org/21536719:59
*** jwcroppe has quit IRC20:03
marcinzzzthanks20:06
*** jwcroppe has joined #openstack-ceilometer20:10
*** jasonamyers has quit IRC20:12
*** jwcroppe has quit IRC20:14
*** jasonamyers has joined #openstack-ceilometer20:23
*** Samolo has quit IRC20:24
*** eglynn_ has joined #openstack-ceilometer20:24
*** eglynn has quit IRC20:24
*** fawadkhaliq has quit IRC20:25
*** rjaiswal has joined #openstack-ceilometer20:25
rjaiswalhi gordc: is there an alternative to submitting patches to stable/kilo20:26
gordcrjaiswal: alternative?20:28
rjaiswalhttps://bugs.launchpad.net/ceilometer/+bug/1488707 affects kilo, https://review.openstack.org/134252 was the fix done to kilo, but it doesnt change the meter name to identity.trust.created|deleted, the liberty fix https://review.openstack.org/#/c/216979/ is good since it has the name as identity.trust.created|deleted20:28
openstackLaunchpad bug 1488707 in Ceilometer "ceilometer identity.trust.* samples are not stored " [Undecided,In progress] - Assigned to Rohit Jaiswal (rohit-jaiswal-3)20:28
*** thumpba has joined #openstack-ceilometer20:29
*** zeus- has joined #openstack-ceilometer20:29
rjaiswalgordc: the plugin based notifications is deprecated in liberty20:30
rjaiswalso we cant make a fix and backport20:30
rjaiswalto kilo20:30
gordcrjaiswal: yeah, you can't do a backport easily.20:30
gordcyou'll have to 'git checkout origin/stable/kilo && git checkout -B<branch name>'20:31
gordcthen do a completely different patch which does the same thing20:31
*** zeus- is now known as zeus20:31
rjaiswalgordc: ok, so i can submit a patch to stable/kilo20:31
rjaiswali thought it was closed20:31
*** zeus is now known as Guest9246320:31
gordcusually kilo and master patches should share same Change-id but i don't think it matters since it's completely different code-wise20:32
gordcstable/icehouse is closed. you can still submit to juno/kilo20:32
rjaiswalgordc: oh ok20:32
rjaiswalgordc: probably https://review.openstack.org/#/c/216979/ should affect kilo20:32
rjaiswali mean https://bugs.launchpad.net/ceilometer/+bug/148870720:33
openstackLaunchpad bug 1488707 in Ceilometer "ceilometer identity.trust.* samples are not stored " [Undecided,In progress] - Assigned to Rohit Jaiswal (rohit-jaiswal-3)20:33
rjaiswalgordc: how do we add kilo as an affected project?20:33
gordcrjaiswal: there is a 'target to series' option (although it may just be available to maintainers()20:34
*** smekel has joined #openstack-ceilometer20:34
gordci added it for you20:34
rjaiswalgreat, thanks gordc20:35
gordcyeah, it's only for maintainers.20:35
rjaiswalok20:35
gordcstepping out, bbl.20:43
*** gordc has quit IRC20:43
*** Guest92463 has quit IRC20:45
openstackgerritMerged openstack/ceilometer: support custom timestamp  https://review.openstack.org/21536120:51
*** julim has quit IRC20:58
*** jwcroppe has joined #openstack-ceilometer20:59
*** openstackgerrit has quit IRC21:01
*** openstackgerrit has joined #openstack-ceilometer21:01
*** sudorandom has quit IRC21:02
*** sudorandom has joined #openstack-ceilometer21:02
*** jasonamyers has quit IRC21:03
*** r-daneel has quit IRC21:05
*** jaosorior has quit IRC21:15
openstackgerritFrederic Guillot proposed openstack/ceilometer-specs: Timeseries dispatcher  https://review.openstack.org/21683821:16
*** zeus has joined #openstack-ceilometer21:29
*** zeus is now known as Guest5341921:29
*** thorst has quit IRC21:32
*** jaypipes has quit IRC21:37
*** sudorandom has quit IRC21:39
*** alejandrito has joined #openstack-ceilometer21:39
*** thumpba has quit IRC21:40
*** thumpba has joined #openstack-ceilometer21:40
*** sudorandom has joined #openstack-ceilometer21:42
*** shardy has quit IRC21:46
*** thumpba has quit IRC21:47
*** thumpba has joined #openstack-ceilometer21:49
*** thumpba has quit IRC21:51
*** alejandrito has quit IRC21:51
*** jaypipes has joined #openstack-ceilometer21:52
edmondswgordc, your thoughts on my last comment in https://bugs.launchpad.net/ceilometer/+bug/1489049 ?21:56
openstackLaunchpad bug 1489049 in Ceilometer "correlation_id missing from http.audit.request/response events" [Low,Triaged]21:56
*** edmondsw has quit IRC21:59
*** thorst has joined #openstack-ceilometer22:27
*** jwcroppe has quit IRC22:30
*** gaud has quit IRC22:31
*** jaypipes has quit IRC22:33
*** nihilifer has joined #openstack-ceilometer22:40
*** exploreshaifali has quit IRC22:45
*** rbak_ has quit IRC22:56
*** ankita_wagh has quit IRC23:18
*** ankita_wagh has joined #openstack-ceilometer23:19
*** ankita_wagh has quit IRC23:48
llugordc: I'm still working on that. Hope to get patches on gerrit for review by this week23:49
llugordc: sorry for the slow progress23:49
« Tuesday, 2015-08-25 Index Thursday, 2015-08-27 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!