Tuesday, 2015-08-25

*** _cjones_ has quit IRC00:01
*** ddieterly has joined #openstack-ceilometer00:25
*** thorst has quit IRC00:36
*** srsakhamuri has quit IRC00:37
*** ddieterl_ has joined #openstack-ceilometer00:42
*** ddieterly has quit IRC00:42
*** liusheng has quit IRC00:54
*** ankita_w_ has quit IRC01:02
*** ankita_wagh has joined #openstack-ceilometer01:02
*** julim has quit IRC01:06
*** ankita_wagh has quit IRC01:07
*** liusheng has joined #openstack-ceilometer01:20
openstackgerritliusheng proposed openstack/python-ceilometerclient: Don't try to get aodh endpoint if auth_url didn't provided  https://review.openstack.org/21619901:32
openstackgerritMerged openstack/ceilometer: Allow to run debug tox job for functional tests  https://review.openstack.org/21633701:42
*** ddieterl_ has quit IRC01:44
*** ddieterly has joined #openstack-ceilometer01:45
*** ddieterly has quit IRC01:50
*** sinisak has quit IRC02:04
*** ddieterly has joined #openstack-ceilometer02:11
*** julim has joined #openstack-ceilometer02:11
*** sinisak has joined #openstack-ceilometer02:12
*** mstyne has quit IRC02:37
*** mstyne has joined #openstack-ceilometer02:41
*** Ephur has joined #openstack-ceilometer02:41
*** liusheng has quit IRC02:44
*** liusheng has joined #openstack-ceilometer02:45
*** dikonoor has joined #openstack-ceilometer02:52
*** ankita_wagh has joined #openstack-ceilometer03:02
*** Ephur has quit IRC03:02
*** jruano has quit IRC03:13
*** rjaiswal has quit IRC03:20
*** links has joined #openstack-ceilometer03:24
*** sinisak has quit IRC03:37
*** deepthi has joined #openstack-ceilometer03:42
*** ankita_w_ has joined #openstack-ceilometer03:52
*** ankita_wagh has quit IRC03:52
*** ddieterly has quit IRC04:14
*** fawadkhaliq has joined #openstack-ceilometer04:57
*** ildikov has quit IRC04:59
*** liusheng has quit IRC05:09
*** liusheng has joined #openstack-ceilometer05:09
*** ddieterly has joined #openstack-ceilometer05:15
*** ddieterly has quit IRC05:20
*** Ephur has joined #openstack-ceilometer05:22
*** ankita_w_ has quit IRC05:28
*** nadya has joined #openstack-ceilometer05:29
*** ildikov has joined #openstack-ceilometer05:40
*** ankita_wagh has joined #openstack-ceilometer05:42
*** nadya has quit IRC05:46
*** links has quit IRC05:57
*** exploreshaifali has joined #openstack-ceilometer05:58
*** links has joined #openstack-ceilometer06:00
*** jaosorior has joined #openstack-ceilometer06:01
*** yprokule has joined #openstack-ceilometer06:03
*** r-mibu has joined #openstack-ceilometer06:04
*** boris-42 has quit IRC06:12
*** serverascode has quit IRC06:12
*** jaosorior has quit IRC06:12
*** serverascode has joined #openstack-ceilometer06:14
*** ddieterly has joined #openstack-ceilometer06:16
*** jaosorior has joined #openstack-ceilometer06:16
*** boris-42 has joined #openstack-ceilometer06:17
*** ankita_wagh has quit IRC06:20
*** ddieterly has quit IRC06:21
*** fawadk has joined #openstack-ceilometer06:34
*** fawadkhaliq has quit IRC06:35
*** exploreshaifali has quit IRC06:40
*** summer has joined #openstack-ceilometer07:01
openstackgerritRoman Dobosz proposed openstack/ceilometer: Versioned object for Meter and Sample model  https://review.openstack.org/19052107:04
openstackgerritRoman Dobosz proposed openstack/ceilometer: Versioned object initial patch  https://review.openstack.org/17867507:04
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements  https://review.openstack.org/21585507:06
*** ddieterly has joined #openstack-ceilometer07:17
*** shardy has joined #openstack-ceilometer07:21
openstackgerritRoman Dobosz proposed openstack/aodh: Versioned object initial patch  https://review.openstack.org/21624307:21
*** ddieterly has quit IRC07:22
*** cdent has joined #openstack-ceilometer07:31
* sileht wonder how many patchset about versioned object will be sent before I question is aswered...07:38
cdentIt’s a good question sileht (both of them)07:40
cdentThe versioned object thing has the feel of “fine, whatever, if you insist” so isn’t really getting the scrutiny it deserves, it is good that at least you are paying real attention.07:41
cdentI hope/suspect that now that you’ve identified a pretty specific flaw people will notice (I hadn’t, or at least hadn’t though about it). Seems like a pretty big deal.07:41
*** fawadk has quit IRC07:43
*** summerWendy has joined #openstack-ceilometer07:54
*** summer has quit IRC07:54
silehtcool, my keystonemiddleware fixes have finally landed, next release will fix the aodh/keystone configuration issue07:55
cdent\o/07:57
cdentdid you see this sileht : https://review.openstack.org/#/c/216390/07:59
silehtcdent, not yet :)07:59
silehtcdent, I really like this solution, I just wonder about if the percent of collision when we truncate the sha256 is reasonable08:01
cdentthere’s some comments about that between jd and I in the review08:01
cdentif the first 32 chars of a sha256 hexdigest are a valid UUID08:02
cdentthen they are a valid UUID08:02
cdenthowever, yeah, is that actually true? I dunno.08:03
*** nadya has joined #openstack-ceilometer08:05
*** jaosorior has quit IRC08:05
silehtcdent, I just read that uuid version 5 is truncated version of sha1 (160bit truncated to 128bit)08:07
cdentdo you know which way the truncation goes (low end or big end)?08:07
*** nadya has quit IRC08:07
cdent(and would that even matter)08:07
silehtcdent, it's the beginning of the hash08:09
*** exploreshaifali has joined #openstack-ceilometer08:10
silehtcdent, I have added my proposal to use sha1 instead of sha256 to match the rfc4122 recommandation (because I have almost no cryptography knownledge)08:16
*** ddieterly has joined #openstack-ceilometer08:18
cdentoh, intresting, thanks08:18
silehtcdent, the pseugo algo is:08:19
*** deepthi has quit IRC08:19
*** ddieterly has quit IRC08:22
silehtcdent, according the rfc that should be generated like this: http://paste.openstack.org/show/426827/08:23
* cdent waits for paste08:23
cdentah, so the namespace provides a sort of seed?08:24
silehtcdent, I guess08:24
silehtyou have the code in C page 23 of the rfc08:25
cdentapparenly you can just call: uuid.uuid5(namespace, name)08:26
cdentdoes this mean the namespace needs to be static and global across all instance of gnocchi everywhere ever?08:27
silehtcdent, yes08:27
_gryfsileht, hi, i saw your comments regarding multiple database connections in versioned objects patch08:30
_gryfsileht, I agree that that problem should be resolved first.08:30
_gryfsileht, so if in ceilometer db connection is created once - is there a way to use it in object classes?08:31
sileht_gryf, I don't get the question08:33
_gryfsileht, in case of versioned object class there is a need somehow to get access to the db08:37
*** Ala has joined #openstack-ceilometer08:37
_gryfsileht, one of the obvious way was to get the connection via storage.get_connection_from_config method08:38
*** IBerezovskiy has joined #openstack-ceilometer08:38
_gryfsileht, which create new connection every time it is called, right?08:39
sileht_gryf, that my question does the lib allows to pass the connection object without storing it into a global variable or a class variable08:39
*** deepthi has joined #openstack-ceilometer08:39
_gryfsileht, hm. let me check08:39
silehtThe application already call storage.get_connection_from_config() ealier to create only one object08:39
silehtI don't see why a lib enforce the application to create more that one object to handle database connection08:40
_gryfsileht, I agree, it shouldn;t encforce to do such things08:42
*** shardy_ has joined #openstack-ceilometer08:43
*** nadya has joined #openstack-ceilometer08:44
*** shardy has quit IRC08:44
cdentnadya: I’m trying to remember, were you tracking the influxdb+gnocchi  stuff?08:45
nadyacdent: yep08:46
cdentdid you see the bug I created yesterday?08:46
* cdent finds link08:46
cdenthttps://bugs.launchpad.net/gnocchi/+bug/148802708:46
openstackLaunchpad bug 1488027 in Gnocchi "The gnocchi wsgi app experiences timeout errors when using influxdb" [Undecided,New]08:46
nadyacdent: yep, I saw it08:46
nadyalet me describe our plans08:46
cdentI haven’t had a chance to look any closer, but it seems a…problem08:47
*** spradeepv has joined #openstack-ceilometer08:47
nadyacdent: first of all we will work on it to find out what is the main problem08:48
*** shardy_ has quit IRC08:48
*** shardy has joined #openstack-ceilometer08:49
nadyacdent: but actually we were going to rewrite a data-scheme to support tags in influx, it should make it much faster08:49
cdenttags will be cool, but as far as I can tell the issue above is something more fundamentally wrong with the way that gnocchi is talking to influxdb08:50
cdentthe rate of requests isn’t super high, but still there is blocking happening, in some fashion, not something I would expect08:50
nadyacdent: we will definitely take care of it. I will assign it to ityaptin, ok?08:51
cdentyeah, that would be awesome if he has the time. I wasn’t directly looking for someone to deal with it, just wanted to make sure people were aware of it. But ityaptin can dig around and see what’s up that would be great.08:52
*** exploreshaifali has quit IRC08:52
nadyacdent: yep, it's out first priority now08:52
cdentcool08:52
*** shardy has quit IRC08:52
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: rest: Pass the project name to middleware config  https://review.openstack.org/21661608:53
silehtcdent, do we have a fix for the sql migration issue ?08:53
sileht(on gnocchi)08:53
cdentnot yet, no08:53
cdentwhat I was abl to find out was that it is our bug08:53
cdentalembic 0.8.1 is being more strict08:53
cdentI added some notes about it to the bug last night08:54
*** shardy has joined #openstack-ceilometer08:54
cdentbut I didn’t get any further than that because I have no alembic experience08:54
cdent(also it was late)08:54
*** jaosorior has joined #openstack-ceilometer08:59
silehtcdent, I wonder if the mapping shim should be on gnocchi or ceilometer dispatcher side09:02
cdentIf it’s on the dispatcher side then the gnocchi api breaks09:03
cdentthe client (the human) needs to never know that id translation is being done if they want to be able to get their data again later09:03
cdent(at least in a reasonably friendly and non-convoluted way)09:04
silehtOk I see this is the client pov that matter (I guess yes too), that the only reasonable solution09:04
silehtSo I have to fix a bit the dispatcher too, to send the bad resource_id correclty to gnocchi :p09:05
*** fawadkhaliq has joined #openstack-ceilometer09:09
cdentbrb09:09
*** cdent has quit IRC09:09
*** cdent has joined #openstack-ceilometer09:12
*** ddieterly has joined #openstack-ceilometer09:18
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: quote the resource_id in url  https://review.openstack.org/21662809:21
silehtcdent, can I steal https://bugs.launchpad.net/gnocchi/+bug/1479819 ?09:21
openstackLaunchpad bug 1479819 in Gnocchi "From the api it is possible to GET resource history on all the resources but not one. It should be." [Undecided,New] - Assigned to Chris Dent (chdent)09:21
cdentyes, please09:22
* cdent had forgotten he had it09:22
silehtthx09:22
*** ddieterly has quit IRC09:23
openstackgerritZhai, Edwin proposed openstack/ceilometer: Do not use system config file for test  https://review.openstack.org/21663209:24
*** shardy_ has joined #openstack-ceilometer09:33
*** shardy has quit IRC09:35
*** shardy_ has quit IRC09:39
*** shardy has joined #openstack-ceilometer09:39
*** exploreshaifali has joined #openstack-ceilometer10:03
cdentsileht: we can't use uuid510:12
cdentbecause it's not repeatable10:12
silehtoO10:12
*** nadya has quit IRC10:13
*** summerWendy has quit IRC10:13
cdentso the question becomes whether sha1 is sufficient10:14
cdentor perhaps, is sha256 sufficiently inexpensive10:14
silehtcdent, why is not repetable ?10:15
cdenttry it:10:15
cdentoh wait10:15
cdenti'm making a mistake10:15
cdentjust a sec10:15
*** exploreshaifali has quit IRC10:15
silehtI have try:10:15
sileht>>> for i in xrange(0,  100000000):10:15
sileht...     if ref != uuid.uuid5(ns, "azertyuiopqsdfghjklmwxcvbn,;:!ù*=)àç_è_è-(é&"):10:15
sileht...             print "NOP"10:16
sileht...10:16
sileht>>>10:16
silehtand I have no NOP10:16
cdentno, you're right10:16
cdentmy little test case was wrong10:16
silehtouf :p10:16
*** exploreshaifali has joined #openstack-ceilometer10:16
cdent:)10:16
cdentI was distracted by a different problem (python2/3 differences on uuid5)10:16
*** ddieterly has joined #openstack-ceilometer10:19
openstackgerritZhai, Edwin proposed openstack/ceilometer: Fix disable_non_metric_meters referencing  https://review.openstack.org/21665010:21
*** ddieterly has quit IRC10:23
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: rest: implements /resource/generic/UUID/history  https://review.openstack.org/21665310:30
*** shardy_ has joined #openstack-ceilometer10:37
*** shardy has quit IRC10:38
openstackgerritRyota MIBU proposed openstack/aodh: Make event-alarm evaluator caching alarms  https://review.openstack.org/21592710:40
*** shardy_ has quit IRC10:43
*** shardy has joined #openstack-ceilometer10:43
*** nadya has joined #openstack-ceilometer10:44
*** jruano has joined #openstack-ceilometer10:44
openstackgerritMerged openstack/ceilometer: Updated from global requirements  https://review.openstack.org/21585511:02
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: move to jsonpath_rw_ext  https://review.openstack.org/21601211:18
*** ddieterly has joined #openstack-ceilometer11:20
*** fawadkhaliq has quit IRC11:20
*** ddieterly has quit IRC11:25
openstackgerritChris Dent proposed openstack/gnocchi: Introduce a resource id translation shim  https://review.openstack.org/21639011:31
openstackgerritDivya K Konoor proposed openstack/ceilometer: WIP Events RBAC needs scoped token  https://review.openstack.org/21255711:34
*** thorst has joined #openstack-ceilometer11:35
*** exploreshaifali has quit IRC11:36
*** exploreshaifali has joined #openstack-ceilometer11:41
*** gordc has joined #openstack-ceilometer11:42
gordcone last check, but this is the proposed question i will submit: https://etherpad.openstack.org/p/ceilometer-insight11:54
gordccdent: ^ your english major might prove valuable :)11:55
* cdent looks11:56
openstackgerritDivya K Konoor proposed openstack/ceilometer: Add user_id,project_id traits to audit events  https://review.openstack.org/21150911:56
*** shardy_ has joined #openstack-ceilometer11:58
*** shardy has quit IRC12:00
*** shardy_ has quit IRC12:04
gordci should add, i'm assuming there are capitals required for each bullet item... and that is it?12:04
*** shardy has joined #openstack-ceilometer12:05
cdentgordc I assume the second chunk of list is not relevant?12:05
gordccdent: correct. that was initial list.12:05
gordci think it's covered in 1st chunk but feel free to edit if not.12:06
gordcthe current options can be edited as well... i just picked 6 from nadya's list12:06
cdentis it supposed to be metrics for billing _and_ tenant management?12:07
gordc...maybe tenant billing management?12:09
*** fawadkhaliq has joined #openstack-ceilometer12:10
gordcit's not tenant management... "billing tenants" is the use case.12:10
nadyacdent: what is tenant management? some quotes in tenant? I believe that all billing-metrics are tenant-specific by default12:10
_gryfsileht, cdent, I've commented the approach on gerrit, could you please take a look? https://review.openstack.org/#/c/216243/12:11
gordcnadya: yeah, i worded it wrong.12:11
openstackgerritDivya K Konoor proposed openstack/ceilometer: Add user_id,project_id traits to audit events  https://review.openstack.org/21150912:12
sileht_gryf, at first glance that looks reasonable12:12
_gryfsileht, cool12:13
*** fawadkhaliq has quit IRC12:14
cdentgordc: I think I'm done. Presumarly the survey makers will do an edit too?12:16
gordc*shrugs*12:16
gordci can ask... but i'm going to assume they're mainly just going to verify that the structure is legit.12:18
gordcif anyone else wants to vet the question, i'll give a few more minutes. https://etherpad.openstack.org/p/ceilometer-insight12:18
gordcildikov: courtesy ping ^12:19
*** fawadkhaliq has joined #openstack-ceilometer12:20
*** ddieterly has joined #openstack-ceilometer12:21
*** prad_ has joined #openstack-ceilometer12:22
*** prad_ is now known as pradk12:24
*** ddieterly has quit IRC12:25
* gordc sending question.12:33
*** ddieterly has joined #openstack-ceilometer12:35
*** edmondsw has joined #openstack-ceilometer12:36
*** deepthi has quit IRC12:37
ildikovgordc: I think it's fine :)12:39
openstackgerritMerged openstack/ceilometer: preload jsonpath_rw parsers  https://review.openstack.org/21564012:42
gordcildikov: too late anyways :P12:43
gordcpradk: shall we close your bp after the '$.' patch? any other items you want to work on?12:46
ildikovgordc: I know :)12:47
pradkgordc, yea i think that should cover majority of what we wanted in12:47
gordcpradk: cool cool.12:47
gordci guess we need to document it now12:48
*** nadya has quit IRC12:54
pradkgordc, yea i'll get a doc patch out next12:54
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: move to jsonpath_rw_ext  https://review.openstack.org/21601212:56
dikonoorityaptin : Hello . I still haven't got the ceilometer functional tests running :( . Are there any instructions anywhere on the steps13:04
*** ddieterly has quit IRC13:04
dikonoorityaptin : Are there any dependencies that need to be manually installed?13:05
*** nadya has joined #openstack-ceilometer13:08
ildikovgordc: I guess I can also start with documenting Aodh13:08
gordcdikonoor: you can run 'tox -epy-mysql'13:08
gordcildikov: that'd be great. i think sileht or liusheng might be good resources to understand exact differences.13:08
*** nadya has quit IRC13:09
ildikovgordc: ok, cool13:09
ildikovgordc: do you mean it from deployment perspective too?13:09
openstackgerritMerged openstack/ceilometer: Use jsonpath_rw_ext for meter/event definitions  https://review.openstack.org/21552913:09
*** nadya has joined #openstack-ceilometer13:10
gordcildikov: right, sileht is working on integration tests so he should have best idea how to get it all working.13:10
gordcildikov: what type of docco are you writing?13:10
*** nadya has quit IRC13:11
ildikovgordc: Iplanned to update the user facing ones which means the Admin/Install/Configuration guides13:11
*** nadya has joined #openstack-ceilometer13:11
silehtgordc, integration tests will become a voting job very soon13:12
gordcsileht: \o/13:13
gordcsileht: thanks very much!13:13
ildikovsileht: cool :)13:14
silehtYou can review some additional tests: https://review.openstack.org/#/c/215604/13:14
ildikovsileht: and thanks of course!13:14
silehtwe cannot break anymore the basic workflow with ceilometer+aodh+gnocchi13:15
gordcsileht: we have ceilometer+aodh as well?13:16
silehtgordc, this is only with aodh13:16
silehtuntil we releases a new python-ceilometer, heat talks to ceilometer-api and it's redirected to aodh13:17
sileht(python-ceilometerclient)13:17
*** changbl has quit IRC13:21
jd__cdent: do you actually understand how to fix/what is the issue on https://bugs.launchpad.net/gnocchi/+bug/1488190 ?13:22
openstackLaunchpad bug 1488190 in Gnocchi "gnocchi alembic migrations not compatible with alembic 0.8.1" [High,Triaged] - Assigned to Julien Danjou (jdanjou)13:22
jd__cdent: 'cause I cannot find where the RESTRICT is missing13:22
cdentjd__: no13:22
cdentI don't know what zzzeek meant by "both sides"13:22
dikonoorgordc: finally my functional tests work..it wasn't working earlier because i was using forward slash as separator..IT should be .13:23
cdentbut isn't it the case that gnocchi is using the upgrade process in an interesting way? I have a vague memory of that13:23
*** links has quit IRC13:23
cdentjd__: maybe best to just find zzzeek and ask him?13:23
openstackgerritMerged openstack/aodh: Make event-alarm evaluator caching alarms  https://review.openstack.org/21592713:23
jd__cdent: I thought he meant migation + base model, but it's there13:26
jd__cdent: I'm gonna track this man down!13:26
cdent13:26
jd__gosh he seems away13:26
jd__he might have heard us!13:26
cdentfoo13:26
*** jwcroppe has quit IRC13:27
gordcsileht: right. i meant testing against ceilometer v2 db rather than gnocchi? in case user wants ceilometer+aodh but haven't adopted gnocchi yet.13:28
gordcsileht: also, the latest python-ceilometercilent should have redirect support as well.13:29
silehtgordc, I can't write another job for that, it should be easy just the alarm definition change13:29
silehtgordc, I will add that to my todo13:29
gordcsileht: cool cool. no worries.13:29
*** jwcroppe has joined #openstack-ceilometer13:30
*** jwcroppe has quit IRC13:34
jd__obviously that bug is only on MySQL, haha :(13:38
*** rbak has joined #openstack-ceilometer13:38
jd__sileht: you meant "I can write" ?13:38
silehtjd__, of course :)13:39
gordc:) i just assumed you didn't want to do it.13:39
*** jwcroppe has joined #openstack-ceilometer13:47
*** shardy_ has joined #openstack-ceilometer13:50
*** r-mibu has quit IRC13:50
*** r-mibu has joined #openstack-ceilometer13:50
*** shardy has quit IRC13:51
*** bapalm_ is now known as bapalm13:51
openstackgerritChris Dent proposed openstack/ceilometer: Sync devstack plugin with devstack:lib/ceilometer  https://review.openstack.org/21672113:52
*** thumpba has joined #openstack-ceilometer13:53
*** shardy_ has quit IRC13:55
*** shardy has joined #openstack-ceilometer13:56
*** jruano has quit IRC13:57
openstackgerritJulien Danjou proposed openstack/gnocchi: Exclude Alembic>=0.8.1  https://review.openstack.org/21672213:57
*** jaypipes has joined #openstack-ceilometer13:58
cdentjd__: I +A that13:58
jd__thanks cdent13:58
cdentlots of stuff sitting around13:58
*** yprokule has quit IRC13:58
jd__I just sent a mail to Mike about that bug with you in Cc13:59
cdentyeah, saw that, will be curious to read the response13:59
jd__it seems fair he spends 5 minutes figuring out what's wrong rather than me spending 1 day13:59
cdentyeah13:59
*** yprokule has joined #openstack-ceilometer13:59
*** shardy has quit IRC14:00
*** shardy has joined #openstack-ceilometer14:01
*** nadya has quit IRC14:13
*** annasort_ has joined #openstack-ceilometer14:21
*** annasort has quit IRC14:24
*** annasort_ is now known as annasort14:24
cdentgordc: when you went to the monasca meeting last week where/when was it?14:28
gordcit was a webex meeting14:29
gordcsidenote: i'm pretty sure i got a virus from it.14:29
gordcbut yeah, i think you need windows14:30
*** nadya has joined #openstack-ceilometer14:30
cdentoh, then I'll have to send my apologies14:30
cdentwhat's roland's nick?14:30
*** annasort has quit IRC14:30
gordcno idea. he might be in monasca channel14:30
gordchttps://wiki.openstack.org/wiki/Meetings/Monasca14:31
gordcit might work with fedora... it complained about java when i tried.14:31
gordci had to use webex from windows and then dial in from phone... i think ildikov had to do the same.14:32
cdentwow14:32
ildikovthe meeting is at 1500UTC14:32
ildikovI'm not sure at the moment whether I was able to make webex partially work in my browser on Linux, but for audio I called in form phone14:33
ildikovcdent: ^^14:33
ildikovs/form/from/14:34
jd__the gate is overwhelmed14:35
jd__gordc: so how things are turning out wrt Monasca?14:36
gordcjd__: ildikov and i joined last week's meeting and gave them overview of what we were doing14:37
gordcjd__: they were interested in the event-alarming work as they are currently porting their threshold engine to python14:37
ildikovjd__: we were also discussing about having at least one joint session on the next Summit14:38
jd__ok14:38
jd__good to hear14:38
ildikovjd__: and prolly they are interested in the further componentization work too, but we/they will see14:38
gordcjd__: there is redundancy in monasca and gnocchi as their entire model is pretty much centre around storage(vertica) so i don't know if there's integration there. maybe outside of that domain i guess.14:41
jd__gordc: I thought they were also interested in InfluxDB?14:42
*** links has joined #openstack-ceilometer14:44
gordcyeah, there's a influxdb driver as well14:44
*** yprokule has quit IRC14:49
*** yprokule has joined #openstack-ceilometer14:50
*** r-mibu has left #openstack-ceilometer14:51
*** nadya has quit IRC14:53
*** exploreshaifali has quit IRC14:59
*** annasort has joined #openstack-ceilometer15:09
*** jruano has joined #openstack-ceilometer15:11
*** annasort_ has joined #openstack-ceilometer15:12
*** ahilan has quit IRC15:13
*** annasort has quit IRC15:14
*** annasort_ is now known as annasort15:14
*** fawadkhaliq has quit IRC15:15
cdentsileht: dhellmann just released a new wsme15:18
silehtcool15:18
silehttomorrow, I will update my change to use it15:18
*** fawadkhaliq has joined #openstack-ceilometer15:27
openstackgerritMerged openstack/ceilometer: Do not use system config file for test  https://review.openstack.org/21663215:29
*** shardy_ has joined #openstack-ceilometer15:31
*** shardy has quit IRC15:32
*** ildikov has quit IRC15:33
*** shardy_ has quit IRC15:36
*** shardy has joined #openstack-ceilometer15:37
*** mstyne has quit IRC15:38
sileht2215:39
openstackgerritChris Dent proposed openstack/ceilometer: Update gabbi xfail tests to reflect WSME 0.8 fixes  https://review.openstack.org/21677215:42
*** mstyne has joined #openstack-ceilometer15:43
*** lsmola has quit IRC15:44
*** links has quit IRC15:44
silehtgordc, cdent can you review https://review.openstack.org/#/c/215604/, infra want it merged before they approve the infra part15:46
cdentdone15:47
gordclooking15:51
*** nadya has joined #openstack-ceilometer15:53
gordcsileht: +A15:54
silehtthanks !15:54
openstackgerritChris Dent proposed openstack/ceilometer: Update gabbi xfail tests to reflect WSME 0.8 fixes  https://review.openstack.org/21677215:54
*** changbl has joined #openstack-ceilometer15:55
*** nadya has quit IRC15:58
cdentsileht: there's a pile of functional alarm tests that are bad under wsme 0.8, I'll get those, I've abandonded my gabbi-related one16:00
gordcpradk: weird. your patches didn't get triggered for merge... i just rechecked.16:00
*** dikonoor has quit IRC16:00
pradkyea was going to recheck before i head out for lunch, thx16:00
*** _cjones_ has joined #openstack-ceilometer16:03
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Update gabbi xfail tests to reflect WSME 0.8 fixes  https://review.openstack.org/20846716:05
cdentsileht: actually looks like a real change in wsme behavior: do a tox -r -efunctional -- test_alarm_scenarios16:05
* sileht is trying16:06
*** Ala has quit IRC16:08
gordcpradk: ^16:08
pradkyea16:08
silehtcdent, it seems something is broken16:09
* cdent nods :(16:09
gordccrud16:09
* gordc watches gate implode16:09
openstackgerritMerged openstack/gnocchi: Exclude Alembic>=0.8.1  https://review.openstack.org/21672216:10
cdentthe schema checking has gone wrong...16:11
cdentsome of the problems are just different error messages and are okay/fixable16:14
cdentbut some of the problems are more interesting16:14
silehtcdent, all of this failures seems normal I will fix them in my changes16:22
cdentthere's one that's not16:22
cdentline 1988 in test_alarm_scenarios16:22
cdentwhich tests that an unknown field is just ignored16:23
cdentin WSME 0.8 it is not ignored16:23
silehtcdent, that not really a issue16:23
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Update tests to reflect WSME 0.8 fixes  https://review.openstack.org/20846716:23
silehtcdent, ignoring incorrect input is a bug16:24
cdentbut what about enterprisey backwards compatibility?16:24
cdentwhich means even bugs need to be backwards compatible?16:24
silehtcdent, we will not really break backwards compatible, but show up bugs into other apps :p16:25
cdentif you're happy with that, I'm happy with that16:25
pradki dont know what i'm seeing is related but we're seeing kilo tests fail too in osp716:26
cdentpradk: it occurs to me that wsme 0.7 is new since kilo as well16:26
cdentso you might try less than it16:26
cdent(to see if is possible to avoid a backport)16:26
pradkhttp://fpaste.org/259081/44052001/ si what i'm seeing.. seems like sileht's above fix addresses a few of thos exception messages16:27
*** exploreshaifali has joined #openstack-ceilometer16:27
silehtOn my side ceilometer.tests.api.v2.test_alarm_scenarios.TestAlarms.test_post_invalid_alarm_input_enabled_int doesn't fail16:28
silehtbut the error remind my something16:29
pradkhmm16:29
silehtpradk, sorry what I have in mind that was not the same error16:32
pradkkk16:33
cdentaodh is all cheesed up on WSME too16:33
gordcany help needed or are both covered?16:33
cdentgordc: I think it's under control16:34
cdentsileht: I'll try to apply your changes to aodh16:34
gordccdent: cool cool. i guess i'll wait to either fast approve and recheck stuff this afternoon16:34
silehtK16:34
*** eglynn has joined #openstack-ceilometer16:35
openstackgerritgordon chung proposed openstack/ceilometer: remove jsonpath-rw requirement  https://review.openstack.org/21679416:44
openstackgerritChris Dent proposed openstack/aodh: Update tests to reflect WSME 0.8.0 changes  https://review.openstack.org/21679516:45
*** ankita_wagh has joined #openstack-ceilometer16:45
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: rest: implements /resource/generic/UUID/history  https://review.openstack.org/21665316:54
*** ankita_w_ has joined #openstack-ceilometer16:55
*** ankita_wagh has quit IRC16:58
openstackgerritMerged openstack/python-ceilometerclient: Don't try to get aodh endpoint if auth_url didn't provided  https://review.openstack.org/21619917:03
cdentsileht: your wsme change modified requirements.txt so presumably it can't merge in ceilo17:08
cdentunless we do a global-requirements change17:09
cdentwhich will break things like ironic17:09
*** julim has quit IRC17:09
cdent(which has fixes for wsme 0.8 in progress too)17:10
openstackgerritChris Dent proposed openstack/ceilometer: Update tests to reflect WSME 0.8 fixes  https://review.openstack.org/20846717:14
cdenti've removed the requirements.txt change for now17:14
*** rjaiswal has joined #openstack-ceilometer17:14
*** jaosorior has quit IRC17:15
*** yprokule has quit IRC17:18
silehtk17:22
*** liusheng has quit IRC17:24
cdentI assume that was the better option for now?17:25
* cdent doesn't really know17:25
*** liusheng has joined #openstack-ceilometer17:25
*** exploreshaifali has quit IRC17:40
*** nadya has joined #openstack-ceilometer17:43
*** lintonv has joined #openstack-ceilometer18:01
lintonvWhen an instance is created in openstack, can ceilometer poll via the cli or api to notify?18:01
*** ildikov has joined #openstack-ceilometer18:03
*** shardy_ has joined #openstack-ceilometer18:04
*** shardy has quit IRC18:05
lintonvanybody have info on ceilometer being notified on instance creation?18:07
gordclintonv: we get data both ways. nova will send a notification that it's been created. but eventually taht instance will be polled as well18:07
lintonv@gordc, thank you18:07
gordcyou should enable 'store_events' in ceilometer if you want to know when specific actions happen18:08
lintonvwhen you say that instance will be polled - what entity polls it?18:08
gordcnp18:08
gordcthe polling agent (configured to compute namespace) will gather metrics on the instance.18:09
lintonvyes, i do have store_events set to 'true' in the conf file18:09
gordci believe it's packaged as compute agent18:09
*** shardy_ has quit IRC18:09
lintonvcan I do something with ceilometer-polling to view the event?18:10
lintonvI have tried combinations of it but cannot see the event18:10
*** shardy has joined #openstack-ceilometer18:10
gordcceilometer-polling just gathers measurement data.18:10
lintonvok18:10
gordcif you want to view the create event itself, you need to enable store_events in ceilometer and run ceilometer event-list18:10
gordcto view the create event (or any event in OpenStack)18:11
lintonvok. that was the missing piece18:11
lintonvthank you so much gordc18:12
gordcnot a problem.18:12
gordci should mention event support is being added to horizon but i don't know if it'll merge for liberty18:13
lintonvOk, does that mean it is being pulled out of openstack? or will be merged with another release?18:14
gordcthe next release maybe.18:15
gordcthis is the patch: https://review.openstack.org/#/c/188134/18:15
cdentgordc: i'll be back later to make sure that wsme stuff is moving along properly18:16
*** cdent has quit IRC18:16
*** Qlawy has quit IRC18:17
*** Qlawy has joined #openstack-ceilometer18:18
lintonvinteresting - in the meantime should we look at other ways to poll or get notified about events?18:18
lintonvoh i misread - you said added to Horizon18:20
gordclintonv: well you can still query the data from ceilometer. it just isn't visible in Horizon currently.18:20
gordcyep :)18:20
lintonvso cli support will still be there18:20
lintonvmy bad :)18:20
gordcyep18:20
*** jfluhmann has joined #openstack-ceilometer18:22
openstackgerritMerged openstack/gnocchi: Introduce a resource id translation shim  https://review.openstack.org/21639018:41
*** changbl has quit IRC18:49
openstackgerritMerged openstack/gnocchi: storage: remove create_metric()  https://review.openstack.org/21380318:49
*** cdent has joined #openstack-ceilometer18:51
cdentgordc: the grenade stuff has been approved18:55
cdentso we’re probably going to need to do a bit of making sure everything we think is supposed to be happening is happening tomorrow18:57
gordccdent: i noticed.18:59
gordcthat mean we don't have grenade running anymore?18:59
cdentno, we do18:59
cdentbut via the plugin18:59
*** changbl has joined #openstack-ceilometer18:59
gordcok. so we just need to make sure the plugin is doing what it's suppose to be doing.18:59
cdentyeah18:59
cdentthere may be a brief gap where one or two things merge without grenade19:00
cdentbut based on what’s currently pending in the gate etc (and the fact that wsme is blocking everything at the moment) I don’t reckon it’ll be much19:00
cdentbtw, pradk, if your rechecks are about the wsme related failures, that thing is still waiting to merge19:01
cdentgate be slow19:01
pradkcdent, ah kk, i'll wait19:01
*** jruano has quit IRC19:02
gordccdent: ok. i see the experimental job. will run check experimental19:02
cdentgordc, the infra change to change that job from experimental to real is merging as wel19:02
cdentl19:02
gordcoh. read it wrong.19:03
gordci thought it was going to experimental... even better19:03
openstackgerritFrederic Guillot proposed openstack/ceilometer-specs: Timeseries dispatcher  https://review.openstack.org/21683819:04
cdent“Integrated queue delay: 5.2 hours “19:10
openstackgerritMerged openstack/gnocchi: Make pagination gabbi test no longer xfail  https://review.openstack.org/21521619:11
openstackgerritMerged openstack/gnocchi: file: fix potential race condition in storing measure  https://review.openstack.org/21412219:12
*** lintonv has quit IRC19:16
*** eglynn has quit IRC19:17
*** ddieterly has joined #openstack-ceilometer19:24
*** exploreshaifali has joined #openstack-ceilometer19:27
openstackgerritMerged openstack/aodh: Update tests to reflect WSME 0.8.0 changes  https://review.openstack.org/21679519:29
openstackgerritRohit Jaiswal proposed openstack/ceilometer: Adds support for dynamic event pipeline  https://review.openstack.org/20806819:33
*** ankita_w_ has quit IRC19:43
*** ankita_wagh has joined #openstack-ceilometer19:47
*** fawadkhaliq has quit IRC19:50
*** nadya has quit IRC19:51
*** jfluhmann has quit IRC19:54
*** julim has joined #openstack-ceilometer20:06
*** exploreshaifali has quit IRC20:17
*** MVenesio has joined #openstack-ceilometer20:35
*** nadya has joined #openstack-ceilometer20:52
*** changbl has quit IRC20:53
*** nadya has quit IRC20:56
*** thorst has quit IRC21:09
*** ddieterly has quit IRC21:23
*** exploreshaifali has joined #openstack-ceilometer21:26
*** annasort has quit IRC21:39
*** rjaiswal has quit IRC21:40
*** thumpba has quit IRC21:48
*** shardy has quit IRC21:52
*** pradk has quit IRC21:53
edmondswgordc, I'm seeing the request_id trait have different values for corresponding http.audit.request and http.audit.response events. Shouldn't those match, or am I misunderstanding?21:54
*** jaypipes has quit IRC21:55
*** edmondsw has quit IRC21:58
*** jaypipes has joined #openstack-ceilometer21:59
*** MVenesio has quit IRC22:00
*** rjaiswal has joined #openstack-ceilometer22:07
rjaiswalgordc: question regarding ceilometermiddleware.swift, which type of requests are expected to be metered?22:07
rjaiswalgordc: Should it meter the client HEAD requests as well?22:11
*** ddieterly has joined #openstack-ceilometer22:12
gordcrjaiswal: i don't think the middleware meters explicit requests currently22:12
gordcrjaiswal: i didn't write original implementation but i think it basically just creates a meter from any request that uploads/downloads data.22:14
rjaiswali see that HEAD requests that are made by the swiftclient as part of doing a swift upload are metered, but not when doing a swift stat22:14
rjaiswalor swift list22:15
rjaiswalgordc: if we are metering HEAD, then they should all be metered?22:16
gordcrjaiswal: going to grab the code we lifted off swift22:16
*** rbak has quit IRC22:17
gordcrjaiswal: we lifted this code from swift which calculates the data taht is uploaded/downloaded https://github.com/openstack/swift/blob/master/swift/common/utils.py#L2776-L281322:17
rjaiswalgordc: ok22:18
rjaiswalI observed that https://github.com/openstack/ceilometermiddleware/blob/master/ceilometermiddleware/swift.py#L182 raises a valueError in those cases when the HEADs get ignored22:18
gordcrjaiswal: i'd probably have to dig into the code myself. i just ported over the original logic.22:21
gordcunfortunately, i'm not 100% the logic behind how we calculate meter and for which req22:21
rjaiswalok22:21
rjaiswalnp, i updated https://bugs.launchpad.net/ceilometermiddleware/+bug/1485711, can reproduce it, but the above is another issue, i guess22:23
openstackLaunchpad bug 1485711 in ceilometermiddleware "ceilomiddleware.swift ignore the swift_source in environ" [Undecided,New] - Assigned to Rohit Jaiswal (rohit-jaiswal-3)22:23
gordcrjaiswal: cool cool. haven't really looked at many bugs in past few days. i think once we get libery-3 cut next week, i can focus on bugs22:24
rjaiswalgordc: alright, thanks22:25
gordcsorry22:25
cdentI reckon there’s gonna be tons of bugs gordc22:25
cdentbut in the good way22:25
gordccdent:  fun22:25
gordcare you waiting on gate?22:25
cdent22:25
cdentno, I was doing some other stuff22:25
cdentthe gate is taking its sweet time22:26
gordci see. yeah, that sucker isn't clearing anytime soon.22:26
gordci'm hopnig it clears before eod for me so i can recheck everything and go to sleep22:26
cdentI just haven’t managed to leave IRC22:26
cdentif it doesn’t i’ll recheck everything in my morning22:26
gordccool cool.22:26
*** RaySun has joined #openstack-ceilometer22:35
gordcedmondsw: if you read logs: https://review.openstack.org/#/c/216889/22:35
cdent‘night22:45
*** cdent has quit IRC22:45
*** jaypipes has quit IRC22:48
*** RaySun has quit IRC22:58
*** gordc has quit IRC23:17
*** bitblt has joined #openstack-ceilometer23:28
*** bitblt has quit IRC23:28
*** changbl has joined #openstack-ceilometer23:29
*** ddieterly has quit IRC23:30
openstackgerritRohit Jaiswal proposed openstack/ceilometermiddleware: Ignore request with swift.source in env  https://review.openstack.org/21690323:35
*** annasort has joined #openstack-ceilometer23:40
*** annasort has quit IRC23:41
*** ddieterly has joined #openstack-ceilometer23:47

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