Friday, 2016-03-18

« Thursday, 2016-03-17 Index Saturday, 2016-03-19 »
*** aaggarw3 has quit IRC00:06
*** thorst has quit IRC00:11
*** thorst has joined #openstack-telemetry00:12
*** thorst has quit IRC00:16
*** nicodemus_ has joined #openstack-telemetry00:24
*** thorst has joined #openstack-telemetry00:39
*** thorst has quit IRC00:41
*** thorst has joined #openstack-telemetry00:41
*** vishwanathj is now known as vishwanathj_away01:00
*** cheneydc has joined #openstack-telemetry01:01
*** boris-42 has quit IRC01:04
*** safchain has quit IRC01:13
*** gzhai1 has joined #openstack-telemetry01:23
*** gzhai has quit IRC01:27
*** marcin12345 has quit IRC01:27
*** zqfan has joined #openstack-telemetry01:27
*** caisan has joined #openstack-telemetry01:42
*** gzhai2 has joined #openstack-telemetry02:12
*** gzhai1 has quit IRC02:14
*** links has joined #openstack-telemetry02:40
*** caisan_ has joined #openstack-telemetry02:44
*** caisan has quit IRC02:44
openstackgerritliusheng proposed openstack/aodh: Fix Aodh-alarm-evaluator recreates deleted alarms in some cases  https://review.openstack.org/29096402:44
*** thorst has quit IRC02:48
*** thorst has joined #openstack-telemetry02:48
*** nicodemus_ has quit IRC02:56
*** thorst has quit IRC02:57
*** liamji has joined #openstack-telemetry02:59
*** mdih has joined #openstack-telemetry03:04
* liusheng sigh, gate is still broken03:12
liamjiliusheng: it seems to have some problems with Gnocchi?03:43
liamji:(03:44
liushengliamji: because miss uuid-runtime installation03:44
openstackgerritliusheng proposed openstack/aodh: Fix Aodh-alarm-evaluator recreates deleted alarms in some cases  https://review.openstack.org/29096403:50
*** thorst has joined #openstack-telemetry03:54
*** rbak has joined #openstack-telemetry03:58
*** thorst has quit IRC04:02
*** nicodemus_ has joined #openstack-telemetry04:10
*** dave-mccowan has quit IRC04:21
*** Gautam has joined #openstack-telemetry04:30
*** aaggarw3 has joined #openstack-telemetry04:36
*** rbak has quit IRC04:36
*** jwcroppe_ has joined #openstack-telemetry04:53
*** jwcroppe has quit IRC04:57
*** jwcroppe_ has quit IRC04:58
*** thorst has joined #openstack-telemetry04:59
*** jwcroppe has joined #openstack-telemetry05:02
*** thorst has quit IRC05:07
*** aaggarw3 has quit IRC05:32
*** aaggarw3 has joined #openstack-telemetry05:32
*** jwcroppe has quit IRC05:43
*** jwcroppe has joined #openstack-telemetry05:43
*** aaggarw3 has quit IRC05:46
*** jwcroppe has quit IRC05:47
*** cheneydc has quit IRC05:49
*** cheneydc has joined #openstack-telemetry05:52
*** yprokule has joined #openstack-telemetry05:55
*** mdih has quit IRC06:00
*** nicodemus_ has quit IRC06:01
openstackgerritLianhao Lu proposed openstack/gnocchi: devstack: allow gnocchi-api to run on different host from keystone  https://review.openstack.org/29440406:06
*** thorst has joined #openstack-telemetry06:06
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Imported Translations from Zanata  https://review.openstack.org/29271506:10
*** thorst has quit IRC06:12
*** nadya has joined #openstack-telemetry06:12
*** nadya has quit IRC06:13
*** pcaruana has quit IRC06:18
openstackgerritliusheng proposed openstack/aodh: Fix Aodh-alarm-evaluator recreates deleted alarms in some cases  https://review.openstack.org/29096406:27
openstackgerritLianhao Lu proposed openstack/ceilometer: devstack: allow ceilometer-api and keystone to run on different hosts  https://review.openstack.org/29442006:36
GautamIs there a way to get aggregated number of vcpu metric per user from ceilometer(API) using collected vcpu metrics. The idea is to get total vcpu metric consumed by instances deployed by the user. I know this is easier to get from Nova, just wanted to check if this possible from ceilometer.06:36
*** rcernin has joined #openstack-telemetry06:37
lluanyone know how long does it take for  https://review.openstack.org/#/c/294233/ make effect? gnocchi gate is blocking itself and ceilometer06:45
*** ljxiash has joined #openstack-telemetry07:01
*** thorst has joined #openstack-telemetry07:09
*** Gautam has quit IRC07:12
*** Gautam has joined #openstack-telemetry07:13
*** thorst has quit IRC07:16
*** Gautam has quit IRC07:18
openstackgerritliusheng proposed openstack/aodh: Fix Aodh-alarm-evaluator recreates deleted alarms in some cases  https://review.openstack.org/29096407:28
*** pcaruana has joined #openstack-telemetry07:57
*** cheneydc has quit IRC07:57
*** safchain has joined #openstack-telemetry07:58
*** cheneydc has joined #openstack-telemetry08:11
openstackgerritliusheng proposed openstack/aodh: Fix Aodh-alarm-evaluator recreates deleted alarms in some cases  https://review.openstack.org/29096408:12
*** Gautam has joined #openstack-telemetry08:14
*** thorst has joined #openstack-telemetry08:14
*** Gautam has quit IRC08:19
*** tesseract has joined #openstack-telemetry08:20
*** tesseract is now known as Guest5995408:21
*** thorst has quit IRC08:21
*** Guest59954 is now known as new08:25
*** new has quit IRC08:26
*** tesseract- has joined #openstack-telemetry08:26
openstackgerritLianhao Lu proposed openstack/ceilometer: devstack: allow ceilometer-api and keystone to run on different hosts  https://review.openstack.org/29442008:29
openstackgerritLianhao Lu proposed openstack/gnocchi: devstack: allow gnocchi-api to run on different host from keystone  https://review.openstack.org/29440408:30
*** Gautam has joined #openstack-telemetry08:42
*** vishwanathj_away is now known as vishwanathj_zzz08:51
*** iberezovskiy|afk is now known as iberezovskiy08:54
*** ljxiash has quit IRC08:59
*** ljxiash has joined #openstack-telemetry08:59
*** ljxiash has quit IRC09:04
openstackgerritZhenyu Zheng proposed openstack/python-aodhclient: Add name support for show, update and delete  https://review.openstack.org/29230509:09
openstackgerritZhiQiang Fan proposed openstack/aodh: add missing hbase requirement  https://review.openstack.org/29447009:19
*** thorst has joined #openstack-telemetry09:19
zqfanhi, is there any reason that we skip hbase test in aodh, I know it is deprecated, and once jd__ has found hbase introduces some issue blocks the ci09:20
zqfanbut now the ci enables hbase test, but skip most of them, which seems only some kind of resource waste09:21
zqfanand hide bug in patch https://review.openstack.org/29096409:23
*** thorst has quit IRC09:27
*** nadya has joined #openstack-telemetry09:33
*** nadya has quit IRC09:34
*** yassine__ has joined #openstack-telemetry09:35
*** nadya has joined #openstack-telemetry09:36
*** links has quit IRC09:37
*** achatterjee has joined #openstack-telemetry09:42
*** cheneydc has quit IRC10:01
*** agireud has quit IRC10:03
*** agireud has joined #openstack-telemetry10:05
*** cdent has joined #openstack-telemetry10:15
*** caisan_ has quit IRC10:22
*** thorst has joined #openstack-telemetry10:24
*** caisan has joined #openstack-telemetry10:25
*** thorst has quit IRC10:32
*** links has joined #openstack-telemetry10:35
cdentjd__: I had the exact same thought about ptl non-canidacy messages10:44
*** liamji has quit IRC10:57
*** caisan_ has joined #openstack-telemetry10:59
*** caisan has quit IRC11:03
*** caisan_ has quit IRC11:03
jd__cdent: you know what they say about great minds11:22
*** achatterjee has quit IRC11:34
*** thorst has joined #openstack-telemetry11:41
*** rbak has joined #openstack-telemetry11:42
*** jwcroppe has joined #openstack-telemetry11:44
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: ceph: Try to reduce lock contention  https://review.openstack.org/29399011:48
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: tools: Add tools to quickly add measurements  https://review.openstack.org/29453711:48
*** rbak has quit IRC11:48
*** dave-mccowan has joined #openstack-telemetry12:05
*** ljxiash has joined #openstack-telemetry12:09
*** gordc has joined #openstack-telemetry12:09
*** Gautam_ has joined #openstack-telemetry12:11
*** Gautam has quit IRC12:15
*** Gautam_ has quit IRC12:17
openstackgerritZhiQiang Fan proposed openstack/ceilometer: remove SERVICE_TENANT_NAME from devstack plugin  https://review.openstack.org/29455712:24
jd__sileht: did you see my comment about the dead lock retry to merged in the previous patchset btw?12:27
silehtjd__, Yes I keep separatly until I sure it works :)12:28
jd__ok!12:28
silehtjd__, I have added a comment on the previous patch about the reason of the deadlock12:28
silehtjd__, did you see it ?12:28
jd__not yet12:28
jd__I'm gonna check it right now12:29
jd__I spend my yesterday rewriting overtest in Python -_-12:29
jd__poor choice of life12:29
silehtahah12:30
silehtjd__, I have passed one day to use rados_lock instead of xattr in ceph, this morning I have bench it, it was worse as before, I check why in ceph code, ceph uses xattr to store locks info...12:31
*** pradk has joined #openstack-telemetry12:32
*** peristeri has joined #openstack-telemetry12:32
jd__I saw that indeed12:33
silehtjd__, so I will start to replace xattr by omap, omap allow to configure the lock in our case, we don't about attributes values, so I will put SKIPRWLOCKS everywhere12:33
jd__and gordc has posted good report on the threading system we use12:33
jd__sileht: that sounds like a good lead12:33
jd__sileht: we can have that in 2.1?12:33
silehtjd__, I dunno, I will write a PoC first to ensure it's better, I always got surprise with ceph12:34
silehtjd__, and also this will add cradox as hard requirements to got omap support with old Ceph cluster12:34
jd__btw I asked -infra for uuidgen, and the images are still not being refreshed for whatever reason12:34
jd__sileht: fair enough, though we need to check with pradk how cradox is packaed12:34
jd__packaged12:34
silehtjd__, currently is not :)12:34
jd__so we need him/someone to do it ASAP :)12:35
pradkjd__, i'll take a look at it, how soon do you need?12:36
*** diogogmt has quit IRC12:37
silehtjd__, about the deadlock, are you ok to merge the code with RetryOnDeadlock now, and then try to improve the 'drop table' to not trigger the deadlock12:37
jd__"The issue seems only occurs with postgres :("12:39
jd__dude, that's because MySQL probably does not enforce anything lol12:39
silehtjd__, I think this is because Mysql lock the whole DB during drop table operation :p12:39
jd__haha12:39
jd__sileht: actually I had a comment in https://review.openstack.org/#/c/269844/30/gnocchi/indexer/sqlalchemy.py and you did not reply12:41
jd__I was not sure the RLock was needed12:41
jd__at least for _cache12:41
jd__but I guess you need to make Base.metadata thread safe since SQLAlchemy does not do it for you12:41
*** diogogmt has joined #openstack-telemetry12:42
silehtexactly12:43
gordcis that what the lock is for? to make thread safe? i don't really have problem merging. i just have a feeling multiple processes will make it possible to have multiple actions at same time.12:44
gordcincluding drop.12:44
*** openstackgerrit has quit IRC12:48
*** openstackgerrit has joined #openstack-telemetry12:49
*** Gautam has joined #openstack-telemetry12:55
*** alejandrito has joined #openstack-telemetry12:56
jd__gordc: yes but the Base.metadata object is local to a process12:56
jd__so we need to protect it because our work on it is not atomic12:57
jd__we iterates over the table and THEN create_all()12:57
jd__which means 2 thread could to the same thing12:57
jd__if 2 processes do the same it's not a big deal since it's not the same Base object :)12:57
jd__we just want to avoid to have a corrupted Python object12:58
*** yprokule has quit IRC12:58
jd__s/to have/having/12:58
gordcjd__: ah i see. makes more sense now.12:58
jd__yeah it was not obvious to me neither at first read12:59
*** tesseract- has quit IRC12:59
gordcsomeone can approve. not sure if uuidgen issue is fixed12:59
jd__it's not fixed12:59
gordcsuper12:59
jd__and I wanted to have the @retrydeadlock patch merged with that one12:59
*** yprokule has joined #openstack-telemetry12:59
jd__-infra said the images are not refreshed yet12:59
jd__viva el cloud12:59
silehtjd__, I will update it soon12:59
*** diogogmt has quit IRC12:59
jd__sileht: yeah take your time… THE GATE IS BROKEN12:59
jd__:]13:00
jd__subtle hint13:00
jd__gonna grab a coffee and make some more python13:00
*** Gautam has quit IRC13:00
gordccoffee is good idea.13:01
*** MVenesio has joined #openstack-telemetry13:04
*** tesseract has joined #openstack-telemetry13:08
*** tesseract is now known as Guest8468813:08
openstackgerritZhiQiang Fan proposed openstack/ceilometer: remove unused field 'triggers' defined in sample event_pipeline.yaml  https://review.openstack.org/29458213:10
*** ljxiash has quit IRC13:15
*** ljxiash has joined #openstack-telemetry13:15
*** ljxiash has quit IRC13:20
*** links has quit IRC13:22
*** ametts has joined #openstack-telemetry13:37
jmloweAm I being led astray by the docs, I'm trying to get compute node stats but after setting compute_monitors I see warnings "Excluding nova.compute.monitors.cpu monitor virt_driver. Not in the list of enabled monitors (CONF.compute_monitors)."13:37
jmlowe?13:38
*** nicodemus_ has joined #openstack-telemetry13:43
gordcjmlowe: no idea. you'll need to ask that in nova.13:43
jd__gordc: is the uuidgen stuff fixed or you are randomly trying?13:44
gordcjd__: randomly trying.lol13:44
gordcAJaeger is looking at it for us13:44
jd__ack13:45
gordc09:39 < AJaeger> new images will be build and uploaded automatically startedn at 15:00 UTC but it takes a few hours  to upload all...13:45
gordcso i guess we need to wait for a few hours... and then a few hours more.13:45
gordci'm going to sleep13:45
jd__sounds like a plan13:46
nicodemus_morning13:48
gordcmorning13:50
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Use omap insteads of xattr  https://review.openstack.org/29459813:50
*** KrishR has joined #openstack-telemetry13:53
openstackgerritlitong01 proposed openstack/ceilometer: Add http publisher  https://review.openstack.org/29062413:55
*** cdent has quit IRC13:56
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Use omap insteads of xattr  https://review.openstack.org/29459813:57
*** rcernin has quit IRC14:01
*** cdent has joined #openstack-telemetry14:03
*** liusheng has quit IRC14:05
*** rbak has joined #openstack-telemetry14:06
nicodemus_gordc, if you have a moment... I'd like to ask you about a presentation we found, "Stabilizing the Jenga tower: Scaling out Ceilometer"14:06
nicodemus_more specifically, the lambda design14:07
*** links has joined #openstack-telemetry14:10
gordcnicodemus_: sure.14:14
nicodemus_the one thing I don't quite understand about such design, is how to make the ceilometer agent to push measurements to two different queues14:15
* gordc needs to google to refresh14:16
gordcoh. it's all done by pipeline.14:16
nicodemus_gordc, http://www.slideshare.net/praddevel/stabilizing-the-jenga-tower-scaling-out-ceilometer14:16
gordcin your pipeline, there is notifier:// publisher in sink14:16
gordcby default it publishes to 'metering' topic14:17
*** diogogmt has joined #openstack-telemetry14:17
gordcbut you can also add multiple publishers and configure it to publish to another topic.14:17
gordcie notifier://?topic=<new_topic>14:18
nicodemus_sweet14:18
nicodemus_so I could poll at a long interval for billing purposes, and poll also every five minutes and publish to a second queue/separate infrastructure for live metrics14:19
*** yprokule has quit IRC14:19
gordcin theory yes.14:19
gordcyou need to play around with source and sink configurations to do that but yeah, it's possible14:20
gordcyou'll need multiple collectors or consumers though. a collector can only listen to one topic currently.14:20
nicodemus_do you think that could be a good strategy for scaling?14:21
nicodemus_because we changed out pipeline to poll every 20 mins, since for billing we don't need the measures to be that frequent14:23
nicodemus_but would be nice to also have points every 5 minutes or so for live metrics14:24
*** pradk has quit IRC14:24
*** pradk has joined #openstack-telemetry14:24
gordcdepends what you're trying to scale? it is a common architecture pattern so it might be useful14:25
nicodemus_in our case scenario, testing showed that having a 5-minute polling with a single archive policy generated too much data for processing14:27
jd__is this a single archive policy with 10 lines definition like you used to have? :)14:28
nicodemus_yes, only it has eight14:28
nicodemus_http://paste.openstack.org/show/491112/14:29
gordcyeah, that's 64 calculations each datapoint14:30
nicodemus_perhaps if we have one gnocchi api/metricd with a long-term, long polling period and a second collector/gnocchi api/metricd/ceph pool with another short term archive policy... we could better scale14:30
nicodemus_plus the short-term archive policy needs only to have median for aggregation method14:30
nicodemus_(yesterday we changed the first definition from 5-minute granularity to 20 min)14:32
*** Gautam has joined #openstack-telemetry14:33
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Use omap insteads of xattr  https://review.openstack.org/29459814:35
gordcnicodemus_: i don't think it's the actual granularity that is the issue. i think it's partly because it needs to do 8 aggregate calculations across 8 archive policies for each datapoint of a metric14:35
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Use omap insteads of xattr  https://review.openstack.org/29459814:37
*** Gautam has quit IRC14:37
*** pcaruana has quit IRC14:38
nicodemus_gordc, yes, come to think about it... it is too much to process14:39
nicodemus_another thing I'd like to ask, is it possible to enable polling on a per-instance basis? Meaning: collect measures from instance A but not B ?14:41
gordcbut yeah, i'm not sure what will happen if you split short term and long term data. it might work.14:41
gordcnicodemus_: nope. it's been something that was requested a while (years) ago. we just couldn't figure out a good way to control it.14:42
gordcnicodemus_: similar: https://blueprints.launchpad.net/ceilometer/+spec/project-based-data-collection-configuration14:42
*** Guest84688 has quit IRC14:48
*** cheneydc has joined #openstack-telemetry14:48
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Add some resource types tests  https://review.openstack.org/27041914:51
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute number  https://review.openstack.org/27009114:51
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute bool  https://review.openstack.org/27041814:51
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute uuid  https://review.openstack.org/27009014:51
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Don't create Ceilometer resource types by default.  https://review.openstack.org/27032214:51
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Move legacy Ceilometer resource into indexer.  https://review.openstack.org/27026614:51
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource type CRUD.  https://review.openstack.org/26984414:51
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute string  https://review.openstack.org/26988814:51
*** nadya has quit IRC14:52
*** links has quit IRC14:54
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: tools: Add tools to quickly add measurements  https://review.openstack.org/29453714:54
*** ametts has quit IRC14:57
*** tesseract has joined #openstack-telemetry15:00
*** tesseract is now known as Guest7688315:00
*** thorst has quit IRC15:03
*** nadya has joined #openstack-telemetry15:03
*** thorst has joined #openstack-telemetry15:07
jmlowefor the life of me I can't seem to figure out how to get CPU related meters from the compute host machines into gnocchi, is there a howto somewhere, has somebody else successfully done this?15:07
*** thorst has quit IRC15:10
*** thorst has joined #openstack-telemetry15:16
*** cdent has quit IRC15:28
*** cdent has joined #openstack-telemetry15:30
*** thorst has quit IRC15:30
*** Gautam has joined #openstack-telemetry15:31
*** david-lyle has quit IRC15:32
*** cdent has quit IRC15:33
*** david-lyle has joined #openstack-telemetry15:33
*** Gautam has quit IRC15:35
*** thorst has joined #openstack-telemetry15:36
*** thorst has quit IRC15:41
*** agireud has quit IRC15:42
*** thorst has joined #openstack-telemetry15:42
*** cdent has joined #openstack-telemetry15:43
*** agireud has joined #openstack-telemetry15:44
*** thorst has quit IRC15:46
*** thorst has joined #openstack-telemetry15:52
*** cdent_ has joined #openstack-telemetry15:56
*** thorst has quit IRC15:56
*** thorst has joined #openstack-telemetry15:58
*** cdent has quit IRC15:58
*** cdent_ is now known as cdent15:58
*** cheneydc has quit IRC15:58
*** david-lyle has quit IRC16:00
*** david-lyle has joined #openstack-telemetry16:01
*** thorst has quit IRC16:02
*** thorst has joined #openstack-telemetry16:04
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Use omap insteads of xattr  https://review.openstack.org/29459816:10
*** thorst has quit IRC16:13
*** thorst has joined #openstack-telemetry16:14
openstackgerritlitong01 proposed openstack/ceilometer: Add http publisher  https://review.openstack.org/29062416:18
*** thorst has quit IRC16:19
*** thorst has joined #openstack-telemetry16:20
*** cdent has quit IRC16:24
*** cdent has joined #openstack-telemetry16:24
*** thorst has quit IRC16:25
*** thorst has joined #openstack-telemetry16:26
*** thorst has quit IRC16:31
*** thorst has joined #openstack-telemetry16:32
*** thorst has quit IRC16:37
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Use omap insteads of xattr  https://review.openstack.org/29459816:38
*** thorst has joined #openstack-telemetry16:38
*** nadya has quit IRC16:41
*** vishwanathj_zzz is now known as vishwanathj16:44
*** jwcroppe has quit IRC16:44
*** jwcroppe has joined #openstack-telemetry16:45
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Fix an IN-predicate SAWarning  https://review.openstack.org/29471116:47
*** pcaruana has joined #openstack-telemetry16:47
*** thorst has quit IRC16:48
*** thorst has joined #openstack-telemetry16:49
*** jwcroppe has quit IRC16:49
*** Gautam has joined #openstack-telemetry16:49
*** thorst has quit IRC16:53
*** ljxiash has joined #openstack-telemetry16:54
*** Gautam has quit IRC16:54
*** thorst has joined #openstack-telemetry16:55
*** thorst has quit IRC16:58
*** thorst has joined #openstack-telemetry16:59
*** ljxiash has quit IRC16:59
*** cdent has quit IRC17:02
*** cdent has joined #openstack-telemetry17:03
*** cdent has quit IRC17:03
*** cdent has joined #openstack-telemetry17:08
*** peristeri has quit IRC17:15
*** Gautam has joined #openstack-telemetry17:21
*** gordc has quit IRC17:42
*** david-lyle_ has joined #openstack-telemetry17:44
*** david-lyle has quit IRC17:45
*** david-lyle_ is now known as david-lyle17:46
*** ljxiash has joined #openstack-telemetry18:07
*** gordc has joined #openstack-telemetry18:08
*** ljxiash has quit IRC18:12
*** Guest76883 has quit IRC18:15
*** zqfan has quit IRC19:02
*** chlong has quit IRC19:12
jmloweok, can somebody explain exactly where my measurements are going?  from he central log "Polling pollster hardware.network.incoming.bytes in the context of meter_snmp", this is the last trace I see before it disappears?19:14
*** cdent has quit IRC19:16
jmlowewhy doesn't it just go into the meter_sink and off to the collector for storage by the dispatcher?19:17
*** nadya has joined #openstack-telemetry19:17
*** Gautam has quit IRC19:29
jmloweha, I knew something was off "rabbitmqctl list_queues |grep metering19:47
jmlowemetering.sample942564"19:47
jmlowenow I have no idea what to do about it19:47
*** yassine__ has quit IRC19:51
*** hongbin has joined #openstack-telemetry20:00
gordcjmlowe: that means your collector isn't reading from that queue OR it's reading very very slowly for some reason20:20
*** jdandrea has quit IRC20:21
*** jdandrea has joined #openstack-telemetry20:22
jmlowelooks like not enough api workers for gnocchi20:26
gordcjmlowe: yeah, you need to customise number of processes/threads20:34
jmloweI should probably figure out how to run it under apache20:35
*** alejandrito has quit IRC20:37
nicodemus_jmlowe, maybe this could help: http://paste.openstack.org/show/491157/20:40
jmloweyep that looks pretty good, thanks20:55
*** MVenesio has quit IRC20:57
*** safchain has quit IRC20:58
nicodemus_you need apache-mod-wsgi for it to work20:58
*** hongbin has left #openstack-telemetry21:01
*** nicodemus_ has quit IRC21:05
*** david-lyle has quit IRC21:06
*** david-lyle has joined #openstack-telemetry21:07
*** nadya has quit IRC21:07
*** Nepoc has quit IRC21:09
*** Nepoc has joined #openstack-telemetry21:15
*** diogogmt has quit IRC21:21
*** thorst has quit IRC21:21
*** diogogmt has joined #openstack-telemetry21:24
*** thorst has joined #openstack-telemetry21:24
*** yarkot1 has joined #openstack-telemetry21:24
*** yarkot has quit IRC21:25
*** j_king has quit IRC21:25
*** adreznec has quit IRC21:26
*** dhellmann has quit IRC21:26
*** adreznec has joined #openstack-telemetry21:26
*** j_king has joined #openstack-telemetry21:26
*** dhellmann has joined #openstack-telemetry21:26
*** ityaptin has quit IRC21:26
*** ityaptin has joined #openstack-telemetry21:27
*** thorst has quit IRC21:28
*** Nepoc has quit IRC21:39
*** Nepoc has joined #openstack-telemetry21:39
*** thorst has joined #openstack-telemetry21:42
*** Nepoc has quit IRC21:44
*** Nepoc has joined #openstack-telemetry21:45
*** thorst has quit IRC21:46
*** pradk has quit IRC21:58
*** cdent has joined #openstack-telemetry22:15
*** cdent has quit IRC22:17
*** KrishR has quit IRC22:23
*** r-mibu has quit IRC22:30
*** r-mibu has joined #openstack-telemetry22:30
openstackgerritAtul Aggarwal proposed openstack/ceilometer: Enhancing Retry logic to Coordination when joining partitioning grp  https://review.openstack.org/29424422:31
*** thorst has joined #openstack-telemetry22:36
*** thorst_ has joined #openstack-telemetry22:36
*** thorst has quit IRC22:40
*** gordc has quit IRC22:42
*** thorst_ has quit IRC22:46
openstackgerritMerged openstack/gnocchi: Added docs about new snmp related resource types  https://review.openstack.org/29391722:57
openstackgerritMerged openstack/aodh: Use assertIn and assertNotIn for test  https://review.openstack.org/29343923:03
openstackgerritAtul Aggarwal proposed openstack/ceilometer: Enhancing Retry logic to Coordination when joining partitioning grp  https://review.openstack.org/29424423:30
*** thorst has joined #openstack-telemetry23:49
*** thorst has quit IRC23:57
« Thursday, 2016-03-17 Index Saturday, 2016-03-19 »

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