Wednesday, 2015-07-29

*** htruta_ has joined #openstack-ceilometer00:06
*** _cjones_ has quit IRC00:14
*** david-lyle has quit IRC00:24
*** changbl has joined #openstack-ceilometer00:34
*** pauldix has joined #openstack-ceilometer00:40
*** thorst has joined #openstack-ceilometer01:03
*** rbak has joined #openstack-ceilometer01:09
*** rbak has quit IRC01:14
*** thorst has quit IRC01:14
*** rbak has joined #openstack-ceilometer01:15
*** r-daneel has quit IRC01:19
*** pauldix has quit IRC01:22
*** pauldix has joined #openstack-ceilometer01:23
*** ankita_w_ has quit IRC01:23
*** ankita_wagh has joined #openstack-ceilometer01:24
*** ankita_wagh has quit IRC01:28
*** lexholden has quit IRC01:43
*** ankita_wagh has joined #openstack-ceilometer01:53
*** rbak has quit IRC01:54
*** htruta_ has quit IRC01:57
*** ankita_wagh has quit IRC01:58
*** ankita_wagh has joined #openstack-ceilometer01:59
*** david-lyle has joined #openstack-ceilometer02:19
*** jiaxi_ has joined #openstack-ceilometer02:20
jiaxi_Hello,everyone ?02:21
jiaxi_Is this ceilometer channel ?02:21
*** david-lyle has quit IRC02:23
*** jwcroppe has quit IRC02:27
*** jwcroppe has joined #openstack-ceilometer02:28
jasonamyersit is02:30
jasonamyersfor ceilometer development02:30
*** jwcroppe has quit IRC02:32
*** Yanyanhu has joined #openstack-ceilometer02:38
*** boris-42 has quit IRC02:40
jiaxi_jasonamyers:  I run tox .  Always error.   AttributeError: 'module' object has no attribute 'write_to_tempfile'02:50
jiaxi_jasonamyers:   File "ceilometer/tests/agent/", line 309, in test_start_with_reloadable_pipeline     pipeline_cfg_file = setup_pipeline_file(pipeline)   File "ceilometer/tests/agent/", line 287, in setup_pipeline_file     pipeline_cfg_file = fileutils.write_to_tempfile(content=pipeline, AttributeError: 'module' object has no attribute 'write_to_tempfile'02:51
jiaxi_jasonamyers: It maybe caused by different version of oslo.utils02:51
jiaxi_jasonamyers: But I installed 2.1.0. And I have read the code of 2.1.0.  fileutils  has  write_to_tempfile02:52
jiaxi_jasonamyers: So I'm really confused.02:52
*** ankita_w_ has joined #openstack-ceilometer02:55
*** ankita_wagh has quit IRC02:58
*** ankita_w_ has quit IRC03:00
*** pauldix has quit IRC03:04
liushengjiaxi_: what the result of run this command in your environment?03:09
liushengjiaxi_: echo "from oslo_utils import fileutils;fileutils.write_to_tempfile" |python03:09
jiaxi_>>> from oslo_utils import fileutils >>> fileutils.write_to_tempfile <function write_to_tempfile at 0x7f4e9d44b140>03:11
*** deepthi has joined #openstack-ceilometer03:11
liushengjiaxi_: or something wrong when tox installing packages for virtual evn ?03:12
jiaxi_liusheng: Hello, Liusheng,   the result is    from oslo_utils import fileutils >>> fileutils.write_to_tempfile <function write_to_tempfile at 0x7f4e9d44b140>03:12
liushengjiaxi_: that is ok03:12
liushengjiaxi_: is there any error when tox install packages for tox env ?03:13
jiaxi_liusheng. I 'm confused.  How to check that03:13
jiaxi_liusheng: I will look03:14
openstackgerritgordon chung proposed openstack/ceilometer: fix gnocchi resources yaml
jiaxi_liusheng:  echo 'Tue Jul 28 23:15:12.990 [FileAllocator] done allocating datafile /tmp/CEILO-MONGODB-a1ww4/local.0, size: 16MB,  took 0.008 secs' + read line + echo 'Tue Jul 28 23:15:12.993 [initandlisten] waiting for connections on port 29000' + grep -q 'waiting for connections on port 29000' + exit_code=0 + break + '[' 0 -eq 1 ']' + export CEILOMETER_TEST_MONGODB_URL=mongodb://localhost:29000/ceilometer + CEILOMETER_T03:17
jiaxi_liusheng: Seems no error in the tox install ?03:17
jiaxi_liusheng: Empty attachments:   stderr   stdout  Traceback (most recent call last):   File "ceilometer/tests/agent/", line 309, in test_start_with_reloadable_pipeline     pipeline_cfg_file = setup_pipeline_file(pipeline)   File "ceilometer/tests/agent/", line 287, in setup_pipeline_file     pipeline_cfg_file = fileutils.write_to_tempfile(content=pipeline, AttributeError: 'module' object has03:18
jiaxi_liusheng: I know you in HuaWei. Help me. Thank you.03:18
liushengjiaxi_: ;-) looking03:19
liushengjiaxi_: maybe you can check the oslo.utils package by  find /opt/stack/ceilometer/.tox/ -name oslo.utils*03:24
liushengjiaxi_: replace the path if needed03:24
*** fabian1 has joined #openstack-ceilometer03:25
jiaxi_liusheng: good suggestion. I will have a try. Thank you03:26
fabian1Hi, I have a "requests.exceptions.ConnectionError('Connection aborted.', BadStatusLine("''",))" in ceilometer execut "ceilometer alarm-list"    Please give me a help ?03:29
liushengjiaxi_: you can just run nose command to check whether the package in your *host* is ok or not03:33
liushengjiaxi_: nosetests -s -v -x ceilometer.tests.agent.test_manager:TestRunTasks03:33
*** david-lyle has joined #openstack-ceilometer03:34
liushengjiaxi_:  or this03:34
liushengjiaxi_: python testr --slowest --testr-args=ceilometer.tests.agent.test_manager.TestRunTasks03:34
jiaxi_liusheng: 厉害。03:34
fabian1liusheng: Hi, I have a "requests.exceptions.ConnectionError('Connection aborted.', BadStatusLine("''",))" in ceilometer execut "ceilometer alarm-list"    Please give me a help ?03:35
liushengfabian1: with aodh installed ?03:37
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements
fabian1liusheng: I have not install aodh03:40
liushengfabian1: other ceilometer commands are OK ?03:41
fabian1liusheng: other ceilometer commands have too error03:41
jiaxi_liusheng:     from oslotest import base ImportError: No module named oslotest03:41
jiaxi_liusheng: python testr --slowest --testr-args=ceilometer.tests.agent.test_manager.TestRunTasks can not run successfully03:42
liushengjiaxi_:  your need install requirements and test-requirements03:42
*** fawadkhaliq has joined #openstack-ceilometer03:42
liushengjiaxi_: pip instlal -r requirements.txt  and pip instlal -r test-requirements.txt03:42
liushengjiaxi_: s/instal/install/03:44
liushengfabian1: seems you need to check ceilometer services03:44
fabian1liusheng: I have try restart openstack-ceilometer-api ,restart finish ,execut command ok ,but after some times ,the error repeat03:44
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometermiddleware: Updated from global requirements
liushengfabian1: you didn't use ceilometer of master ? which version you installed ?03:45
jiaxi_liusheng: it's not automatic installed by the script ?03:45
liushengjiaxi_: no03:46
jiaxi_liusheng: use 'nosetests -s -v -x ceilometer.tests.agent.test_manager:TestRunTasks'   all is ok ...03:49
jiaxi_liusheng: tox will install a new oslo.utils in virtualenv ???03:52
jiaxi_liusheng: according to requirment and test-requirement03:52
fabian1liusheng: my openstack is juno release ,the ceilometer-api release is 1.el7.centos03:54
fabian1liusheng: The ceilometer.conf do you have configure about this ?03:55
liushengjiaxi_: yes, if you run tox without specifying '--sitepackages' parameter03:56
liushengjiaxi_: but, tox is often run failed for me ;(03:56
jiaxi_liusheng: The error come out again03:57
jiaxi_liusheng: AttributeError: 'module' object has no attribute 'write_to_tempfile'03:57
jiaxi_That's annoying...03:57
fabian1liusheng: I have a three ceilometer node,you tell me master,do you configure master ?03:58
jiaxi_liusheng: and added a new error .                        pymongo.errors.OperationFailure: database error: Can't take a write lock while out of disk space03:58
liushengfabian1: sorry I'm not familiar with the centos package of ceilometer juno, it seems something wrong with ceilometer-api service or something related of your deployment solution04:01
liushengjiaxi_: out of disk space.. is the reason04:01
jiaxi_liusheng: is the reason for 'AttributeError: 'module' object has no attribute 'write_to_tempfile''  too ?????????????????????????04:02
fabian1liusheng: thanks,maybe this error Associated with the cluster04:03
liushengjiaxi_: obviously not, that is just package issue04:06
jiaxi_liusheng: 我刚刚给vm加了40G的云硬盘。 仍然报错04:08
jiaxi_liusheng. When will you get out for lunch ?04:08
*** vishwanathj has joined #openstack-ceilometer04:10
jiaxi_liusheng: .tox/py27/local/lib/python2.7/site-packages/oslo_utils$ 里面utils 和 外面的不一样 是1.7.0  requirement  didn't specify oslo.utils to 1.7.004:11
*** jwcroppe has joined #openstack-ceilometer04:14
*** yprokule has joined #openstack-ceilometer04:27
*** ankita_wagh has joined #openstack-ceilometer04:39
*** pauldix has joined #openstack-ceilometer05:05
*** pauldix has quit IRC05:10
openstackgerritliusheng proposed openstack/aodh: Add alembic support for aodh
*** jwcroppe has quit IRC05:25
*** jwcroppe has joined #openstack-ceilometer05:25
*** jwcroppe has quit IRC05:30
*** liusheng has quit IRC05:38
*** liusheng has joined #openstack-ceilometer05:39
*** tshtilma has joined #openstack-ceilometer05:41
*** jwcroppe has joined #openstack-ceilometer05:45
*** _nadya_ has joined #openstack-ceilometer05:59
openstackgerritLan Qi Song proposed openstack/ceilometer: Add keystone V3 support for service credentials
jiaxi_Is there anyone here ?06:00
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: retry with a new token on 401
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Adds batch dispatcher
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: concurrently post resource/measures
*** afazekas has joined #openstack-ceilometer06:09
*** liusheng has quit IRC06:12
*** liusheng has joined #openstack-ceilometer06:15
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: DONT MERGE: check live tests run in gate
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: gate: fixup gabbi live tests
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Add tests to not break ceilometer dispatcher
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Imported Translations from Transifex
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: gate: fixup gabbi live tests
openstackgerritMehdi Abaakouk (sileht) proposed openstack/gnocchi: Add tests to not break ceilometer dispatcher
openstackgerritOpenStack Proposal Bot proposed openstack/aodh: Imported Translations from Transifex
*** kfox1111_ has quit IRC06:47
openstackgerritMehdi Abaakouk (sileht) proposed openstack/aodh: functionnal: fix gating
*** _nadya_ has quit IRC06:55
*** _nadya_ has joined #openstack-ceilometer06:56
*** ankita_wagh has quit IRC06:57
*** ankita_wagh has joined #openstack-ceilometer06:57
*** tongli_ has quit IRC07:00
*** _nadya_ has quit IRC07:00
*** ankita_wagh has quit IRC07:02
openstackgerritliusheng proposed openstack/aodh: Add alembic support for aodh
*** _nadya_ has joined #openstack-ceilometer07:07
*** _nadya_ has quit IRC07:08
*** _nadya_ has joined #openstack-ceilometer07:09
*** _nadya_ has quit IRC07:13
*** ankita_wagh has joined #openstack-ceilometer07:21
*** jwcroppe has quit IRC07:22
*** yprokule has quit IRC07:25
*** yprokule has joined #openstack-ceilometer07:27
*** Yanyan has joined #openstack-ceilometer07:29
*** Yanyanhu has quit IRC07:32
*** ankita_wagh has quit IRC07:33
*** ankita_wagh has joined #openstack-ceilometer07:33
*** ankita_wagh has quit IRC07:37
*** jwcroppe has joined #openstack-ceilometer07:51
*** safchain has joined #openstack-ceilometer08:08
*** jwcroppe has quit IRC08:35
*** exploreshaifali has joined #openstack-ceilometer08:42
*** yassine has joined #openstack-ceilometer08:43
*** _nadya_ has joined #openstack-ceilometer08:47
*** Ala has joined #openstack-ceilometer08:49
*** jiaxi_ has quit IRC08:52
*** nijaba has quit IRC09:02
*** jwcroppe has joined #openstack-ceilometer09:02
*** nijaba has joined #openstack-ceilometer09:07
*** nijaba has quit IRC09:07
*** nijaba has joined #openstack-ceilometer09:07
*** liusheng has quit IRC09:10
*** liusheng has joined #openstack-ceilometer09:11
*** lexholden has joined #openstack-ceilometer09:17
*** arunman has joined #openstack-ceilometer09:21
*** alexpilotti has joined #openstack-ceilometer09:24
-openstackstatus- NOTICE: Currently our CI system is broken, jobs are not getting processed at all.09:27
*** ChanServ changes topic to "Currently our CI system is broken, jobs are not getting processed at all."09:27
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: retry with a new token on 401
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: reduce the number of patch to gnocchi API
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Adds batch dispatcher
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: concurrently post resource/measures
*** yassine has quit IRC09:59
openstackgerritIlya Tyaptin proposed openstack/ceilometer: Initial separating unit and functional tests
*** jwcroppe has quit IRC10:00
*** tshtilma has quit IRC10:00
*** Kennan has quit IRC10:00
*** mitz has quit IRC10:00
*** vgnbkr has quit IRC10:00
*** mitz has joined #openstack-ceilometer10:00
*** Kennan has joined #openstack-ceilometer10:01
*** tshtilma has joined #openstack-ceilometer10:01
*** vgnbkr has joined #openstack-ceilometer10:02
*** clduser_ has joined #openstack-ceilometer10:05
*** Ala has quit IRC10:05
*** clduser has quit IRC10:05
*** ildikov has quit IRC10:05
*** ildikov has joined #openstack-ceilometer10:05
*** fabian1 has left #openstack-ceilometer10:07
*** Yanyan has quit IRC10:12
*** Ala has joined #openstack-ceilometer10:18
*** _nadya_ has quit IRC10:23
* sileht was wondering why its stack doesn't grow up: he got negative values for cpu_util ....10:23
*** exploreshaifali_ has joined #openstack-ceilometer10:39
*** exploreshaifali has quit IRC10:43
*** arunman has left #openstack-ceilometer10:44
*** lexholden has quit IRC10:46
*** cdent has joined #openstack-ceilometer10:55
*** yassine has joined #openstack-ceilometer11:04
*** yassine has quit IRC11:04
*** yassine has joined #openstack-ceilometer11:04
*** pauldix has joined #openstack-ceilometer11:07
*** rbak has joined #openstack-ceilometer11:22
*** deepthi has quit IRC11:24
openstackgerritJason Myers proposed openstack/ceilometer: Instance Cache in Node Discovery Pollster
*** rbak has quit IRC11:29
*** belmoreira has joined #openstack-ceilometer11:31
ityaptincdent: Hi! Could you review a new patch with aggregation pipeline?
cdentyeah, will look again this afternoon11:33
*** jwcroppe has joined #openstack-ceilometer11:47
*** thorst has joined #openstack-ceilometer11:47
*** fabian4 has joined #openstack-ceilometer11:47
*** jwcroppe has quit IRC11:48
*** fawadkhaliq has quit IRC11:48
silehtcdent, jd__  I have started written to write the autoscaling use-case (ceilo+heat+aodh+gnocchi) with gabbi, it works well, I just wonder in which repository I will put the yaml file11:53
silehts/written/to write11:53
cdenta) \o/11:54
cdentb) I don't have a strong opinion, but based on his reaction to the other similar test I guess jd__ does?11:54
*** fawadkhaliq has joined #openstack-ceilometer11:55
silehtcdent, I have used the 'poll' and check the stack status to reach "CREATE_COMPLETE", to wait the heat stack to be created11:58
silehtcdent, I wonder if we can add something to bail out when the status is CREATE_FAILED11:58
*** fabian4 has quit IRC11:59
silehtlike: failure_response_json_paths:11:59
silehtcdent, I can write something in gabbi, if that good for you12:00
cdentsorry, sileht, was away, catching up12:04
*** rbak has joined #openstack-ceilometer12:04
cdentsileht: what's wrong with bailing out when the poll never succeeds?12:05
* cdent is not keen on adding syntax on the left hand side12:05
silehtcdent, the stack creation often fails quickly, when it take some time to sucess12:06
* cdent thinks12:06
cdentare you wanting the entire test to be complete if CREATE_FAILED?12:07
*** exploreshaifali_ has quit IRC12:07
jasonamyersgood morning12:07
cdent(and not go on and try the other tests)12:07
cdentmorning jasonamyers12:07
silehtcdent, I don't think so, because I have other poll in next test to wait that the stack grow up12:09
silehtcdent, if the stack is not created that doesn't make sense to wait for more nova vm12:09
cdentthat's what I meant: if CREATE_FAILED happens, should the entire test run be over?12:09
*** rbak has quit IRC12:10
silehtcdent, yes , with gabbi-run I use the -x for that12:10
*** rbak has joined #openstack-ceilometer12:10
* cdent continues thinking12:10
silehtcdent, side question: '-x' is available throught the python api ?12:10
cdentsileht: it depends on the test runner, but if the test runner supports some kind of --fail-fast then yes12:11
sileht(I guess not, gabbit seem just generate tests for unittest framework))12:11
cdentand unittest can failfast12:11
silehtcdent, oh yes, I can put that in testr :p12:11
cdenttestr does a marvelous job of making it hard and complex though... :)12:12
*** _nadya_ has joined #openstack-ceilometer12:12
*** fabian4 has joined #openstack-ceilometer12:14
*** rbak has quit IRC12:16
silehtsniff, testr doesn't have failfast...12:17
cdentyou can pass it to subunit but only if you are concurrency=112:17
fabian4Hi , I have a question. my ceilometer execute "ceilometer --debug event-list"  error droped connection12:18
cdentfabian4: have you tried limiting the result set12:20
cdentsounds like it is timing out trying to get too much information12:20
*** rbak has joined #openstack-ceilometer12:20
*** gordc has joined #openstack-ceilometer12:21
fabian4cdent: limiting the result set ? I don't known ,where is configure12:21
cdentfabian4: I mean by setting a query: ceilometer event-list -q=<some query here>12:21
fabian4cdent: no,I have select "ceilometer --debug event-type-list or ceilometer --debug alarm-list"12:23
fabian4cdent: when i execut command ceilometer ,error abort connect12:23
cdentcan you use to paste the full command and error message you are getting?12:24
cdentsileht: I'm still thinking about it but I can't think of any clean ways of getting it. Can you show me the tests you've built so far and the three different responses you get from heat (success, fail, not done yet but keep trying)?12:25
*** edmondsw has joined #openstack-ceilometer12:27
*** pauldix has quit IRC12:30
*** fabian4 has quit IRC12:30
cdentsileht: if CREATE_FAILED happens is status still 200?12:40
silehtcdent, yes12:40
cdentthat must be a bug?12:40
silehtcdent, because we just get the description of the resource12:40
silehtcdent, and check for the state of the resource12:40
cdentso, not only does OpenStack hate python it also hates HTTP. ossum12:41
silehtcdent, the creation of the stack can fail in two ways:12:42
sileht* heat detect something wrong when you do the POST and return 40X12:42
sileht* heat fail to create the compute (for example) -> change the stack_status to CREATE_FAIL with the nova error in a description field12:43
*** rbak has quit IRC12:43
cdentyeah, I get what's going on, I just don't agree with the implementation12:44
gordcanything we want in for liberty-2 folks? if not, i'm going to defer everything unfinished to liberty-312:44
*** rbak has joined #openstack-ceilometer12:44
cdentgordc: you going to accept my batch_ thing, or is it still up in the air because we don't know what the code is supposed to look like? ;)12:44
gordccdent: it's failing gate.12:45
cdentgate is failing gate12:45
pradgordc, it would be good to get this in
gordcfair enough.lol12:45
*** fabian4 has joined #openstack-ceilometer12:45
gordci deferred from yesterday to wait for batch patch12:45
*** rbak has quit IRC12:46
fabian4cdent: are you here?12:46
cdentfabian4: yes, you apparently got disconnected12:46
gordcprad: cdent: i thought we said there was a way to manage a single file?12:46
cdentthat _is_ the single file gordc12:47
fabian4cdent: maybe program where is it ?12:47
gordccdent: so we don't have one in etc/ceilometer?12:47
cdentgordc: correct12:47
cdentfabian4: I'm sorry, I don't understand your question.12:48
cdentprad: no point to recheck, ci dead12:48
fabian4cdent: do you see  this url?12:48
pradcdent, wait, we do.. the one in data/_meters is the hidden file with just the needed meters for tempest/grenade12:48
*** ChanServ changes topic to "OpenStack Telemetry (Ceilometer) |"12:48
-openstackstatus- NOTICE: zuul's disks were at capacity. Space has been freed up and jobs are being re-queued.12:48
cdentprad, gordc: that, then, wasn't quite my idea12:49
cdentas there's no point having duplication12:49
gordcyeah, that's what i thought12:49
cdentput all the data in data/_meters.yaml12:49
cdent_and_ use that file as the installable config file12:49
pradcdent, you were suggesting to have a disable approach, which i thought we dint want12:49
cdentyeah, I gave up on the override concept12:49
cdentbut still think that there should be an in-site-packages full file12:50
cdent(gave up because you guys didn't like it)12:50
pradso in setup.cfg we map the hidden file to etc/ceilometer ?12:50
cdentwhy does meters.yaml need to be in setup.cfg at all?12:51
gordccdent: just to confirm but there's a way to see that file in-site-package file without telling users to look at source right?12:51
cdentprad: events.yaml (or whatever it is called) is not in setup.cfg is it?12:51
pradright my point is will the file be in etc/ceilometer after the install12:51
pradit is12:51
pradoh not in setuip.cfg no12:52
gordci don't thin it isi12:52
pradbut thats in etc/ceilometer12:52
cdentthere is no reason for meters.yaml to be in etc/ceilometer12:52
cdentit just needs to _end_up_  in /etc/ceilometer12:52
pradthen what will the user edit to enable/disable meters?12:52
pradcorrect agreed12:52
pradit should be visible in that location12:52
cdentfabian4: looking now12:53
pradafter the install so user can edit12:53
pradcdent, so how will the file end up in /etc/ceilometer .. through devstack plugin or whatever external packaging means?12:54
cdentyeah, whatever method gets files out of etc/ceilometer into /etc/ceilometer can also get files from
fabian4cdent: The program is requests.exceptions.ConnectionError: ('Connection aborted.', BadStatusLine("''",))  and INFO (connectionpool) Resetting dropped connection:12:55
gordclol nice disclaimer.12:55
pradgordc, when are we cutting the l2, today?12:56
gordcprad: it's either this afternoon or early tomorrow.12:56
cdentgordc: I can't speak for every packager everywhere12:56
gordcprad: we should probably wait on your patch though.12:56
*** jaosorior has joined #openstack-ceilometer12:56
jasonamyerscan I beg to get my two patches in :)12:57
* jasonamyers pretty please troll face12:57
gordcprad: tbh, it sort of bothers me we enable stuff but don't disable and delete corresponding code12:57
cdentfabian4: can you do: 'curl -v' and see what you get?12:57
pradgordc, we have that in the following dependent patch?12:57
cdentfabian4: my guess is that either you are having a timeout or there is someone on the network between you and the ceilometer server and it is behaving poorly12:57
pradgordc, obviously i want all to get in, if not we might as well wait12:58
gordcjasonamyers: i already +2'd... any core want to look at
jasonamyersyou rock extra cookies for you in Dublin12:59
* gordc switching accounts12:59
fabian4cdent: when i excute "curl -v"  return curl: (52) Empty reply from server13:00
*** gordc is now known as gordc_idle13:00
cdentfabian4: are you using haproxy or something like that on your network?13:00
*** rbak has joined #openstack-ceilometer13:00
fabian4cdent: I have three nodes , node ip:   and is a bind ip ,and bind ip too13:01
*** gordc has joined #openstack-ceilometer13:03
*** rbak has quit IRC13:04
fabian4cdent: you say timeout;  in ceilometer.conf  able to configure ?13:05
cdentfabian4: the "Empty reply from server" suggests that some proxy in your system is not working13:05
cdentdo you get the same error when you ceilometer meter-list ?13:06
*** rbak has joined #openstack-ceilometer13:07
*** eglynn has joined #openstack-ceilometer13:07
fabian4cdent: ye,when i excute "ceimeter alarm-list or ceilimeter event-type-list or ceilometer sample-list -m cpu_util" al command that error appear?13:07
cdentthen your network is broken13:08
cdentyou should look on the server where your ceilometer api is running and see if it is even getting your requests, I'm guessing that it is probably not13:08
fabian4cdent: this is able to appear,I  will check network and network proxy13:10
*** pauldix has joined #openstack-ceilometer13:10
fabian4cdent: about ceilometer.conf in the doc document have not too more configure ,do you configure have any more ?13:11
*** jiaxi_ has joined #openstack-ceilometer13:12
cdentfabian4: these are the configuration docs for ceilometer under kilo:
cdentjasonamyers: I had some comments, not really blocking, but enough to make me not +A13:16
openstackgerritJason Myers proposed openstack/ceilometer: Instance Caching
cdentsileht: after much thought and discussion with another gabbi person, still haven't come up with a reasonable idea on how to deal with the heat thing that doesn't involve changing either heat or gabbi13:19
cdentif changing gabbi turns out to be required (I'd still resist until we had a few more use cases to make sure we had the generalization correct) I'd think that an "abort:" key of some kind within the poll: dict would be the way to go, but not entirely sure13:20
openstackgerritJason Myers proposed openstack/ceilometer: Instance Cache in Node Discovery Pollster
jasonamyerscdent: changes pushed except for the isoformat one13:21
jasonamyersbut I pushed up a command13:21
*** jruano has joined #openstack-ceilometer13:21
fabian4cdent: thanks , i open this url, about test ceilometer ,do you have something in mind ?13:21
cdentI think you need to inspect the log files13:22
*** rbak has quit IRC13:23
cdentjasonamyers: not isoformat because it's a bad match or just because it doesn't matter or because you hate me?13:23
jasonamyersnah it's missing the Z terminator13:24
*** r-mibu has left #openstack-ceilometer13:24
-openstackstatus- NOTICE: zuul jobs after about 07:00 UTC may need a 'recheck' to enter the queue. Look if your change is in and recheck if not.13:25
*** lexholden has joined #openstack-ceilometer13:26
gordcjasonamyers: i don't think you submitted your comment13:26
jasonamyersdang it13:27
jasonamyershow does one interwebz13:27
*** fabian4 has quit IRC13:28
jasonamyersthere we go13:29
gordcjasonamyers: just curiuos, but why do we need the 'Z'?13:31
cdentthat was my questions too13:31
jasonamyersmatches what they expect as input13:31
cdenttime is hard13:32
gordcsays it assumes UTC if missing... is that not true?13:32
jasonamyersthey are being nice and letting you break their required format13:32
gordcah i se13:32
jasonamyersproper to have the Z13:32
jasonamyersI'm happy to switch if it bugs ya13:33
jasonamyersIf the time is in UTC, add a Z directly after the time without a space. Z is the zone designator for the zero UTC offset. "09:30 UTC" is therefore represented as "09:30Z" or "0930Z". "14:45:15 UTC" would be "14:45:15Z" or "144515Z".13:33
gordcnot the end of the world for me.13:33
cdentI'm fine with it too13:34
cdentit seems rather odd that isoformat() doesn't, you know, isoformat13:34
jasonamyersI like standards especially those like ISO 8601 that have been around for ages13:35
jasonamyerscdent: datetime.utcnow().replace(tzinfo=simple_utc()).isoformat() will give you the +0:00 that works as well13:36
jasonamyersand is valid per spec13:36
cdentoslo_utils can be friendly here but seems meh:13:37
cdent(that True is to the  with_timezone kwarg of utcnow13:38
cdentwhoops typo13:38
cdentwhat a palaver13:38
gordcyeah i just saw taht13:39
gordclet's use timeutils13:39
cdentfor want of a horse a kingdom was lost13:40
cdentor something13:40
*** jwcroppe has joined #openstack-ceilometer13:43
jasonamyerscdent: how did you import oslo_utils13:43
cdentin real code (ie. not above): from oslo_utils import timeutils13:43
openstackgerritJason Myers proposed openstack/ceilometer: Instance Cache in Node Discovery Pollster
openstackgerritJason Myers proposed openstack/ceilometer: Instance Caching
jasonamyersboth updated13:46
*** fabian4 has joined #openstack-ceilometer13:47
silehtcdent, I have found something in the heat API:
*** jiaxi_ has quit IRC13:50
cdentsileht: interesting. does that redirect until it has something to say?13:52
silehtcdent, I wait for the second event of the stack resource to occurs13:52
silehtcdent, and then I check the status of the stack to see if that a success or a error13:53
*** fawadkhaliq has quit IRC13:53
cdentI asked about the redirects because you've got redirects set to true and count set to just 1 so it seems like there must be more requests than are immediately visible actually happening?13:53
silehtcdent, the redirect is because heat stacks/integration_test/ to stacks/integration_test/<stack-id>13:53
silehtheat/heat transforms13:53
cdentso does count need to go up?13:54
silehtcdent, the count is wrong it's for testing13:54
cdentah, okay13:54
cdentmakes more sense :)13:54
cdentI'm glad you were able to find something that is workable13:54
*** rbak has joined #openstack-ceilometer13:56
*** fawadkhaliq has joined #openstack-ceilometer13:56
*** rbak has quit IRC13:58
*** fawadkhaliq has quit IRC14:00
cdentjasonamyers: I'm going all unintentionally nitpicky, just because I happen to be doing multiple reads of the same block :(14:01
*** fawadkhaliq has joined #openstack-ceilometer14:01
*** rbak has joined #openstack-ceilometer14:02
*** fc__ has quit IRC14:03
*** jd__ has quit IRC14:04
*** titilambert has quit IRC14:05
openstackgerritJason Myers proposed openstack/ceilometer: Instance Cache in Node Discovery Pollster
jasonamyerscdent: I changed the whole comment now14:05
*** EmilienM has quit IRC14:05
openstackgerritIlya Tyaptin proposed openstack/ceilometer: [MongoDB] Use a aggregate pipeline in statistics
cdentjasonamyers: word14:06
*** titilambert has joined #openstack-ceilometer14:06
*** fc__ has joined #openstack-ceilometer14:07
cdentergh, my comment while approving the other one was not quite correct14:07
cdentit'll come out in the wash14:07
jasonamyerscdent: should it be using per_host instead of getting everything?14:08
* cdent is asleep at the wheel14:08
jasonamyersI don't know the hardware code as well14:08
cdentno, you're right14:08
cdenti wasn't reading well enough14:09
cdent(despite also reading too much)14:09
*** EmilienM has joined #openstack-ceilometer14:09
jasonamyersadmit it you like to nit pick my stuff because you're a hater :P14:10
*** jd__ has joined #openstack-ceilometer14:10
* cdent is totes a hater14:17
*** rbak has quit IRC14:19
*** r-daneel has joined #openstack-ceilometer14:26
*** jruano has quit IRC14:27
*** liusheng has quit IRC14:27
*** Ephur has joined #openstack-ceilometer14:38
gordccdent: were you joking that this is a gate failure and not real failure?
cdentgordc: I was joking that the gate was dead so I couldn't check14:39
gordcah i see.14:39
cdentI can't replicate the failure locally, so was going to do a recheck14:39
* cdent does one14:40
gordci feel like the gate will take 8hrs per check.14:41
gordcmaybe it's best to release tomorrow morning.14:41
cdentyeah, what happened? things used to be a bit more nippy14:43
* gordc remembers 3 years ago.14:44
* cdent kickstarts hardware for infra14:46
jasonamyersthe funny thing about that is... there are like 6 gold foundation members that are hardware ppls14:52
* cdent isn't laughing14:52
silehtjd__, I have a really wierd carbonara agregation result:
silehtjd__, the measure is always around 10014:52
cdentwhich aggregation is that sileht ?14:53
silehtcdent, the default one: blackmagic ?14:54
silehtit's mean14:54
* cdent wonders if his pandas fix broke things14:54
* cdent wouldn't have though so14:54
silehtcdent, which one ?14:55
* cdent locates14:55
silehtcdent, hum that looks better after the revert14:57
* sileht retries without the revert to ensure14:57
cdentuh oh!14:57
jd__we likely miss tests14:57
* cdent nods14:57
*** eglynn_ has joined #openstack-ceilometer14:57
*** jwcroppe has quit IRC14:58
*** eglynn has quit IRC14:58
cdentat a guess the lenght on input is being used to calculate mean rather than the length of the deuped timeserie14:58
silehtcdent, not sure finally I need to wait more measure to got the issue14:59
*** belmoreira has quit IRC14:59
* sileht waits14:59
cdentso yeah, in AggregatedTimeSerie, the call to set_values calls _resample with the original values, not the deduped (post call to super) values15:01
*** rex_lee has quit IRC15:02
*** jasonamyers has quit IRC15:02
cdentalthough since is 'min' the min won't actually change15:02
* cdent shrugs and waits15:02
*** fabian4 has left #openstack-ceilometer15:04
silehtcdent, this change seems the culprit15:06
cdentjd__: you hear that15:07
cdentsileht: can you write a failing test? If so I'll see about fixing it. Or would you prefer we revert it (it fixes a different bug)15:08
jd__having a test would be great15:08
cdentfix one thing, break another15:11
silehtcdent, oh I just got the issue without the change15:11
cdentuh oh15:11
cdentin that case I'd guess a race in metricd15:12
* cdent is pulling at straws15:12
silehtdamn it I just drop the measurements and seen that one:  "2015-07-29T15:08:51+00:00", 1.0, -2437.87811620991915:15
silehtcdent, jd__ ^15:15
cdentwhat do you mean by "drop"?15:15
*** gordc_idle has quit IRC15:16
silehtI have copy/paste all points that are not really aggregated15:16
silehtwith 1.0 as aggregate period15:17
* sileht will see if ceilometer send that15:17
cdentah, so you have bad inputs in the first place?15:17
silehtI think I will confirm15:17
cdentwhat's the metric?15:19
silehtthe cpu metric doesn't have any negative value15:19
*** jruano has joined #openstack-ceilometer15:21
*** ddaskal has joined #openstack-ceilometer15:23
*** tshtilma has quit IRC15:23
cdentprad, gordc: how did we resolve the question of "This is a temporary duplicate file to staisfy grenade requirements." on data/_meters.yaml? Asking because I'm wondering if there's another patchset coming on that review15:25
pradcdent, yea based on what we discussed above, i'm working on a patch to remove that comment as this is now the permanent location15:26
gordcprad: can you enable one set of notifications at a time... and disable/delete the corresponding old code?15:27
pradcdent, since the new file location doesnt exists, tempest will thinking i'll first submit with file left in etc/ceilo .. then update devstack plugin and remove the etc/ceilo file15:28
gordcso maybe just add glance meters first.15:28
* cdent blinks at the hoops15:28
pradcdent, hehe yea15:28
gordcwhy would tempest fail?15:28
cdentprad, won't tempest checkout the current code15:28
cdentand wont' tempest use the default15:29
cdentnot need to put anything in devstack15:29
pradso for devstack plugin to copy the file from data dir, it needs to first exist?15:29
pradand for tempest tests to work, that file needs to exist15:29
cdentdevstack doesn't need to copy the file15:30
cdentthat's the entire point15:30
pradunless i can make both commits depend on each other15:30
pradit has to to etc/ceilo15:30
gordcprad: i'm under impression that if file doesn't exist  it will fall back to internal file15:30
cdentwhen tempest runs, it will do a checkout of your branch15:31
cdentthat branch will be develop15:31
cdentwhich means that the pkg_resources call to find data/_meters.yaml will work15:31
cdentat least that's the theory :)15:31
pradso for someone using devstack they will have to go update the meters in data dir and not etc15:31
pradforget about tempest.. just for development lets say15:32
cdentfor someone running devstack locally, the devstack code can copy the file15:32
cdentbut it doesn't _have_ to copy the file15:32
cdentso in grenade, which doesn't run the entire "new" devstack code, the file will never make it to /etc/ceilo15:33
cdentbut for other devstack runs it will, because the entire devstack code can copy it over15:33
cdenthowever for your specific change15:33
pradyep agree on grenade part, for a sec i forgot about the fallback logic i added15:33
cdentthere's no requirement that the file be in /etc/ceilo15:33
cdentso your patchset can go in no problem15:34
pradso only reason i see devstack might want to copy that file is for easy of development then15:34
cdentand a later change to devstack could add a cp to /etc/ceilo if desired15:34
cdentyah, exactly15:34
pradwhich can go in subsequent patch without dependency15:34
*** alexpilotti has quit IRC15:34
* gordc switching computers.15:35
*** gordc has quit IRC15:35
pradcool agreed15:35
*** gordc has joined #openstack-ceilometer15:39
cdentsileht: learn anything?15:41
silehtcdent, no it's not more happenning :(15:41
cdentoh dear :(15:41
*** vishwanathj has quit IRC15:41
gordcsileht: is the problem that you have negative cpu_util value?15:44
silehtgordc, it's happen sometimes15:46
silehtwith gnocchi, that's polute a lot the timeseries15:46
gordcsileht: yeah... i see it in mine as well15:46
silehtgordc, do you think libvirt report that ?15:47
silehtgordc, or the transformer is buggeed ?15:47
gordci have 342 samples which are negative for cpu_util15:49
gordci have no cpu samples which are negative15:51
cdentso there's a pollster for cpu_util15:51
cdentand there's also a transformer for generating it15:52
cdentpresumably only some virt drivers will give it15:53
cdentgordc: are you getting transformed or polled cpu_util?15:53
*** fawadkhaliq has quit IRC15:54
gordccdent: what' the easy way to tell?15:56
cdentgood question. In such a situation I'd ask the PTL15:56
*** adurbin_ has joined #openstack-ceilometer15:56
cdentbut you might be able to tell in your acompute.log15:56
cdentmy logs say "Obtaining CPU Util is not implemented for LibvirtInspector"15:58
cdentand I'm all defaulted15:58
gordccan't see...15:59
cdentso unless you have a special virtdriver, I reckon you're the same and it is the transformer15:59
gordccrap. need to go to meeting15:59
* cdent feels gordc's pain15:59
openstackgerritPradeep Kilambi proposed openstack/ceilometer: Enable entry points for new declarative meters
pradgordc, ^^ since you felt strongly about deprecating.. I added glance, magnetodb and heat meters and dropped the corresponding code in the above commit16:04
pradwill add the rest in subsequent commit, but hopefully we can get the above in for liberty 216:05
pradalso should have the meter.yaml moved to data dir entirely16:06
cdentsileht: what's your polling interval where you've been seeing these problems?16:06
silehtI use 5 for my tests16:07
cdentsileht: the pollsters are now not doing transforms16:08
cdentthey insead send a notification16:09
cdentso there's a chance that here:
cdenttime_delta could go negative16:09
cdentbut I'd be quite surprised by 5 second latency on the bus?16:09
*** tshtilma has joined #openstack-ceilometer16:10
cdenteven that's not the cause of the problem here, it's definitely a potential bug16:10
*** jaypipes has joined #openstack-ceilometer16:12
*** zhiyan has quit IRC16:13
*** serverascode has quit IRC16:13
*** yprokule has quit IRC16:14
*** _nadya_ has quit IRC16:14
*** _nadya_ has joined #openstack-ceilometer16:15
*** rbak has joined #openstack-ceilometer16:17
*** david-lyle has quit IRC16:18
*** _nadya_ has quit IRC16:19
*** lsmola has quit IRC16:20
*** belmoreira has joined #openstack-ceilometer16:23
*** afazekas has quit IRC16:27
openstackgerritPradeep Kilambi proposed openstack/ceilometer: Enable entry points for new declarative meters
*** yassine has quit IRC16:29
*** exploreshaifali has joined #openstack-ceilometer16:33
*** jasonamyers has joined #openstack-ceilometer16:39
jasonamyersgate broken again :(16:40
cdentany details?16:40
jasonamyersmy stuff no run16:41
jasonamyersthat's all I see16:41
jasonamyersalso irc cloud is down16:41
jasonamyersrough day16:41
cdentcould just be backed up. do you know about:
jasonamyersI do16:41
jasonamyersIt was runnings my gate-neutron stuff earlier16:41
cdentthat's good16:41
*** serverascode has joined #openstack-ceilometer16:41
jasonamyersbut now it's queued again16:42
*** david-lyle has joined #openstack-ceilometer16:42
cdentit took months for that to be revealed to me by some kind of magical tautological exegesis of myth16:42
cdentor something16:42
* jasonamyers goes to google like half those words16:42
cdentthey're not a good match16:42
cdentI just wanted to get the religiousness visible16:42
* prad has been there, cdent just makes up words16:43
jasonamyersahh word painting16:43
jasonamyersprad: haha16:43
cdentprad this time around they are all real16:43
cdentthey just might not be the _right_ ones16:43
cdenttautological exegesis has a nice ring to it16:43
cdent"I would argue that the meaning or definition that we attribute to reconciliation is not going to be illuminated by tautological exegesis on the term itself but by the..."16:44
cdentthat's the quote from the first (of two) results on searching google for the quoted phrase16:44
pradcdent, can i get some eyes on .. should be in good shape16:44
pradty sir16:45
*** alexpilotti has joined #openstack-ceilometer16:46
*** zhiyan has joined #openstack-ceilometer16:46
openstackgerritPradeep Kilambi proposed openstack/ceilometer: Convert identity, sahara and volume to meters yaml
*** ankita_wagh has joined #openstack-ceilometer16:53
*** _cjones_ has joined #openstack-ceilometer16:53
*** _cjones_ has quit IRC16:53
*** _cjones_ has joined #openstack-ceilometer16:53
*** Ala has quit IRC16:54
*** rjaiswal has joined #openstack-ceilometer16:55
*** dikonoor has joined #openstack-ceilometer17:01
dikonoorgordc:Hi gordc..To submit a new BP, is it enough if we fill this thing up?
jasonamyersalso need to add a spec in
edmondswright, that would be my question... don't we need the spec approved, then create the blueprint?17:04
jasonamyersthat has been the process for me so far17:04
*** rbak has quit IRC17:04
cdentthe blueprint process is a bit vague, the important part is make the spec and get it approved17:04
*** rex_lee has joined #openstack-ceilometer17:05
gordcdikonoor: edmondsw: spec is good. i can create the corresponding blueprint17:14
gordcor you can do both... just don't do a bp and stop... no one will look at it.17:14
*** alexpilotti has quit IRC17:14
gordcand i will delete it in a year17:14
edmondswI'll have the spec up shortly17:14
gordcedmondsw: cool cool17:15
*** krykowski has quit IRC17:15
cdentgordc when you get a chance go up the backlog to where I ask sileht about the potential for out of order: search for "could go negative"17:16
*** _nadya_ has joined #openstack-ceilometer17:16
*** jasonamyers_ has joined #openstack-ceilometer17:16
cdentI'm trying to write a test that confirms that and I'm getting errors from the transformer.Namespace object but the test basically confirms that negative time delta is possible17:16
silehtcdent, I think, this come up now because pollster doesn't do the transform anymore17:17
cdentyeah, that's what I said above17:17
silehtcdent, the collector can have workers17:17
silehtcdent, the transformer doesn't share the state across works ...17:17
gordchmm... maybe it's because i turned off coordinated workers then17:18
cdentsileht, gordc's notification agent changes make it so state is managed across notification agents17:18
cdentthat wouldn't result in negatives17:18
cdentthat would just result in stuff being dropped17:18
gordcyeah. now that i think about it. i was testing uncoordinated workers yesterday17:18
cdentor rather: lack of coordination wouldn't change things17:18
ityaptinsileht, jd__: hi!17:18
silehtityaptin, hi17:19
jd__hey ityaptin17:19
cdentyou still have to have a way for some value to inject a negative number17:19
cdentand from what I can tell that's only if time_delta is negative17:19
cdentvolume_delta protects against negatives17:20
gordccdent: there's clause the prohibits prev_volume < current_volume17:20
cdentbut time_delta does not17:20
gordcmaybe we should allow it if time_delta is negative too?17:20
cdentthat makes for some funky match17:20
gordcfor cumulative only17:21
ityaptindue the tests of influxdb driver, I catch one error then trying to accept metric.archive_policy in influxdb driver:
gordccumulative it's the assumption that bigger value is a more recent value17:21
ityaptinIn the normal run of gnocchi api this exception doesn't happen.17:22
cdentgordc: yeah, but the main issue here is you should not calculate a rate of change that violates the arrow of time17:23
ityaptinCould you  suggest where I should look?17:23
*** alexpilotti has joined #openstack-ceilometer17:23
cdentDoesn't that mean that if the current sample is before the previous one the only logical thing to do is to drop it?17:24
cdentIf we're not happy with that they we'll need to keep a cache of more things17:24
gordccdent: probably... how do we get a more sample first if it's all the same queue?17:25
gordcmore recent*17:25
cdentmind you: we haven't even confirmed that this is the cause of the problem you are seeing...17:25
cdentis the message bus an actual queue or is it a bus?17:25
cdentif there are multiple compute or notification agents then either can have latency17:26
ityaptinjd__, sileht: I looked onto this comment, but how i understand we load archive_policy in get_metrics17:26
cdentif the notification agent republishes to get in the right queue, it might dawdle17:26
gordci don't have multiple compute... i guess the order gets thrown out of wack when oslo.messaging does it routing17:26
cdentunfortunately I don't grok the mechanics of o.messaging to know what is or is not possible17:27
* gordc confirms getting data via transform17:27
jd__ityaptin: that comment might be outdated17:27
gordci guess we should probably add a clause to drop if sample is older than previous17:28
cdentgordc: that's what I'm looking into now. Is that sufficient or does it upset some kind of audit trail?17:28
cdentpresumably it ought to be rare, but who knows17:29
cdent(it's also possible for time_delta to be zero and that would cause no end of excitemnt...)17:29
ityaptinjd__: I think the same and now I try  to find out the reason of Metric instance behavior changing in the tests.17:31
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: tests: add an integration test
silehtcdent, :)17:33
silehttomorrow I will do the infra stuffs17:33
gordccdent: i hope we're not polling the same resource at the exact same time17:34
gordcbut yeah, i guess it's something to handle17:35
cdentdefensive programming sir17:35
cdentI know it is a concept that is pretty alien around these parts, but still ;)17:35
cdentmeh: TypeError: 'Namespace' object is not callable17:35
*** dikonoor has quit IRC17:37
*** krykowski has joined #openstack-ceilometer17:37
*** jaypipes has quit IRC17:40
openstackgerritMehdi Abaakouk (sileht) proposed openstack/aodh: api: fix alarm deletion
gordcfor any cores interested (forgot if i plugged already):
cdentthat one it is _totally_ awesome17:43
openstackgerritMerged openstack/ceilometer: Instance Caching
gordcYay caching is in17:43
cdent(btw, the previously failing test passed , but the entire job is not yet done)17:43
*** r-daneel has quit IRC17:44
openstackgerritMerged openstack/ceilometer: [MongoDB] Use a aggregate pipeline in statistics
gordccdent: i don't know why it failed.17:44
openstackgerritMerged openstack/ceilometer: Instance Cache in Node Discovery Pollster
cdentmaybe it was at that instance that ci exploded17:45
openstackgerritMehdi Abaakouk (sileht) proposed openstack/aodh: api: fix alarm deletion
openstackgerritMehdi Abaakouk (sileht) proposed openstack/aodh: functionnal: fix gating
gordcmeh. future me will look at it.17:45
openstackgerritMatthew Edmonds proposed openstack/ceilometer-specs: Events RBAC via Policy
gordci wonder what else we can mark essential now that caching is in.17:48
pradwhen using cfg.CONF.find_file .. if the file doesnt exists i soemtimes see NoSuchOptError: no such option: config_dir .. is this a bug in oslo_config? or am i not setting something here17:50
gordcin test?17:52
openstackgerritMatthew Edmonds proposed openstack/ceilometer-specs: Events RBAC via Policy
pradsee it in the tests, but can reproduce isolated as well ..
openstackgerritMatthew Edmonds proposed openstack/ceilometer-specs: Events RBAC via Policy
openstackgerritMehdi Abaakouk (sileht) proposed openstack/aodh: api: fix alarm deletion and update
openstackgerritIlya Tyaptin proposed openstack/gnocchi: InfluxDB storage driver
*** prad has quit IRC18:04
*** jasonamyers has quit IRC18:08
*** jasonamyers_ is now known as jasonamyers18:08
*** jasonamyers has quit IRC18:08
cdentprad and progress on the find_file thing? I get frequently confused by oslo_config18:09
cdentseems to do magic things18:09
*** jwcroppe has joined #openstack-ceilometer18:11
*** _nadya_ has quit IRC18:13
*** _nadya_ has joined #openstack-ceilometer18:13
openstackgerritChris Dent proposed openstack/ceilometer: Drop out-of-time-sequence rate of change samples
cdentgordc ^18:14
* cdent dines18:14
*** _nadya_ has quit IRC18:18
*** jasonamyers has joined #openstack-ceilometer18:19
*** prad has joined #openstack-ceilometer18:20
*** prad has quit IRC18:24
openstackgerritMerged openstack/aodh: functionnal: fix gating
*** prad has joined #openstack-ceilometer18:27
*** openstackgerrit has quit IRC18:31
*** openstackgerrit has joined #openstack-ceilometer18:32
*** rbak has joined #openstack-ceilometer18:34
*** jaypipes has joined #openstack-ceilometer18:34
*** thorst has quit IRC18:36
*** rbak has quit IRC18:37
*** safchain has quit IRC18:37
*** rbak has joined #openstack-ceilometer18:40
*** vishwanathj has joined #openstack-ceilometer18:42
gordcprad: i've no idea when it's set18:43
*** jwcroppe has quit IRC18:43
*** jwcroppe has joined #openstack-ceilometer18:44
*** rbak has quit IRC18:45
*** fawadkhaliq has joined #openstack-ceilometer18:45
*** vishwanathj has quit IRC18:46
gordcprad: i would just say you assume you need fixture.Config()18:47
*** jwcroppe has quit IRC18:48
pradyea not sure, feels like something is not initialized, but all the code we have where we use find_file uses  it directly18:48
pradso was wondering if it was a bug in oslo perhaps.. i'll play around18:49
*** thorst has joined #openstack-ceilometer18:51
*** fawadk has joined #openstack-ceilometer18:55
*** fawadkhaliq has quit IRC18:56
*** rbak has joined #openstack-ceilometer19:11
gordci think rule is use fixture.Config or call prepare_service.. either or19:13
*** rbak has quit IRC19:14
*** fawadkhaliq has joined #openstack-ceilometer19:16
*** fawadk has quit IRC19:19
*** ankita_wagh has quit IRC19:23
*** ankita_wagh has joined #openstack-ceilometer19:23
*** fawadk has joined #openstack-ceilometer19:24
*** fawadkhaliq has quit IRC19:26
*** ankita_wagh has quit IRC19:28
pradgordc, yea i already have fixture.Config in my test, but dint take effect.. prepare_service seems to help19:28
pradrunning the full suite to make sure19:28
*** vishwanathj has joined #openstack-ceilometer19:35
openstackgerritPradeep Kilambi proposed openstack/ceilometer: Enable entry points for new declarative meters
*** belmoreira has quit IRC19:47
cdentgordc: your comments on the out-of-time stuff make no sense so I left a comment stating as much19:49
*** krykowski has quit IRC19:50
*** vishwanathj has quit IRC19:57
openstackgerritSaba proposed openstack/ceilometer: Correct unit of snmp based harware disk and memory meters
*** edmondsw has quit IRC20:01
*** fawadk has quit IRC20:02
*** rjaiswal has quit IRC20:04
*** vishwanathj has joined #openstack-ceilometer20:10
gordcyou make no sense.20:14
gordc*mic drop*20:14
cdentIt's true20:14
cdentbut it takes one to know one20:15
cdenti'm rubber and you're glue20:15
gordcmy comment was that your change is breaking the comment in code that specifically says delta can be negative20:15
cdentit's talking about a different delta, the one that follows20:15
cdentif it was talking about the time_delta the comment would have been above both deltas20:16
*** openstackgerrit has quit IRC20:16
*** openstackgerrit has joined #openstack-ceilometer20:16
cdent"deltas of samples" is pretty clear about the value of the samples20:16
* gordc is thinking whether order matters for delta/gauge20:19
gordci read it differently but meh20:19
cdentCan I claim that I'm right by virtue of a degree in English?20:19
cdentthat's about the only thing it can do for me20:19
gordcyou have a degree in english?20:19
gordci was top of class in high school english... at the midterm... afterwards not so much20:20
*** vishwanathj has quit IRC20:20
cdentyeah, once you peak, why keep on trying?20:20
cdentI have experienced this problem too20:20
cdentmind you gordc, in case you get the wrong impression, the only reason I finished that degree (after 14 years of off and on) is because I wanted to go to grad school20:23
gordcyou went ot grad school for english?20:24
cdentno, info science, but I had to have some kind of undergrad degree to get a grad oen20:25
gordcwhat the hell are you doing in arguably the most illiterate profession?20:25
cdenttrying to make it more literate?20:26
gordcdo my emails anger you?20:26
cdentnot that I've noticed20:26
gordccool cool.20:27
cdentI was an english _lit_ major and was rather focused on lit from 1920s so your flexibly stylings fit right in20:27
gordci have no more comments on your patch specifically.20:27
gordci don't know what that means. i took philosophy... and dropped it because it was too much reading... or i didn't feel like buying the books20:28
cdentyou'll note that I have no particular connection to good grammar or spelling words correctly20:28
cdentgordc: is there anything you want me to do to that patch or can I go to bed tonight knowing that I may have done something good today?20:31
gordccdent: do you feel like making it smarter?20:32
cdentyou mean keeping a longer cache? not immediately no20:33
cdentI think we might consider that eventually but would prefer to see the immediate bug band-aided and then a more considered fix put in place20:33
gordcdo you want me to fast approve?20:34
cdentNot for my sake. I know sileht and you were both seeing bugs that _might_ be related to this and it would be interesting to see if it fixed it, but there's no rush from my standpoint.20:34
gordci'm opening a bug to add longer cache20:35
*** ankita_wagh has joined #openstack-ceilometer20:37
*** tshtilma has quit IRC20:38
*** jruano has quit IRC20:40
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements
*** jasonamyers has quit IRC20:43
* cdent nights20:50
*** cdent has quit IRC20:50
*** edmondsw has joined #openstack-ceilometer20:50
*** Saba has joined #openstack-ceilometer21:04
SabaHi , What is the freeze date for Bug fixing in ceilometer for Liberty ?21:04
*** Saba is now known as Guest5721121:04
gordcGuest57211: liberty in general?
*** Guest57211 has quit IRC21:06
gordcpretty much up to release date... except it'll need to be higher and higher severity as you get closer21:06
*** ekarlso has quit IRC21:16
*** htruta_ has joined #openstack-ceilometer21:34
*** eglynn_ has quit IRC21:38
*** gordc has quit IRC21:39
*** eglynn_ has joined #openstack-ceilometer21:39
*** exploreshaifali has quit IRC21:43
*** ekarlso has joined #openstack-ceilometer21:44
*** thorst has quit IRC21:58
*** prad has quit IRC21:59
*** Kennan2 has joined #openstack-ceilometer22:02
*** Kennan has quit IRC22:03
openstackgerritPradeep Kilambi proposed openstack/ceilometer: Enable entry points for new declarative meters
*** jaypipes has quit IRC22:43
*** thorst has joined #openstack-ceilometer22:45
*** pauldix has quit IRC22:48
*** htruta has quit IRC22:50
*** Ephur has quit IRC22:51
*** vishwanathj has joined #openstack-ceilometer22:53
*** eglynn_ has quit IRC22:58
*** alexpilotti has quit IRC23:05
*** ddaskal has quit IRC23:06
*** thorst has quit IRC23:19
*** vishwanathj has quit IRC23:20
*** jaosorior has quit IRC23:21
*** ankita_wagh has quit IRC23:47
*** ankita_wagh has joined #openstack-ceilometer23:48
*** ankita_wagh has quit IRC23:52
*** Kennan2 is now known as Kennan23:53

Generated by 2.14.0 by Marius Gedminas - find it at!