| *** gordc has quit IRC | 00:05 | |
| *** jaypipes has quit IRC | 00:42 | |
| *** shengjiemin has joined #openstack-ceilometer | 00:45 | |
| *** shengjiemin has quit IRC | 01:25 | |
| *** shengjiemin has joined #openstack-ceilometer | 01:47 | |
| *** prad_ has joined #openstack-ceilometer | 01:52 | |
| openstackgerrit | Pradeep Kilambi proposed a change to openstack/ceilometer: Fix docs on what an instance meter represents https://review.openstack.org/66746 | 01:54 |
|---|---|---|
| openstackgerrit | liusheng proposed a change to openstack/python-ceilometerclient: return sample info when creating sample with CLI https://review.openstack.org/66125 | 01:55 |
| *** herndon has joined #openstack-ceilometer | 02:29 | |
| *** herndon has quit IRC | 02:32 | |
| *** coolsvap has quit IRC | 02:48 | |
| *** flwang has joined #openstack-ceilometer | 02:53 | |
| *** asalkeld has joined #openstack-ceilometer | 03:04 | |
| *** asalkeld has left #openstack-ceilometer | 03:10 | |
| *** coolsvap has joined #openstack-ceilometer | 03:42 | |
| openstackgerrit | Pradeep Kilambi proposed a change to openstack/ceilometer: Fix to handle null threshold_rule values https://review.openstack.org/66757 | 03:48 |
| *** coolsvap has quit IRC | 03:49 | |
| *** coolsvap has joined #openstack-ceilometer | 04:02 | |
| *** coolsvap has quit IRC | 04:03 | |
| *** coolsvap has joined #openstack-ceilometer | 04:03 | |
| *** shengjiemin has quit IRC | 04:09 | |
| *** prad_ has quit IRC | 04:12 | |
| *** prad_ has joined #openstack-ceilometer | 04:13 | |
| *** prad_ has quit IRC | 04:34 | |
| *** prad_ has joined #openstack-ceilometer | 04:34 | |
| *** coolsvap has quit IRC | 04:50 | |
| *** prad_ has quit IRC | 04:51 | |
| *** SergeyLukjanov has joined #openstack-ceilometer | 04:56 | |
| *** prad_ has joined #openstack-ceilometer | 04:57 | |
| openstackgerrit | Pradeep Kilambi proposed a change to openstack/ceilometer: Fix to handle null threshold_rule values https://review.openstack.org/66757 | 05:01 |
| *** prad_ has quit IRC | 05:05 | |
| *** coolsvap has joined #openstack-ceilometer | 05:08 | |
| *** Alexei_987 has joined #openstack-ceilometer | 05:25 | |
| *** SergeyLukjanov has quit IRC | 05:36 | |
| *** SergeyLukjanov has joined #openstack-ceilometer | 05:44 | |
| Alexei_987 | jd__: Hi | 05:47 |
| *** SergeyLukjanov has quit IRC | 05:49 | |
| openstackgerrit | Jenkins proposed a change to openstack/ceilometer: Imported Translations from Transifex https://review.openstack.org/62808 | 06:03 |
| *** coolsvap has quit IRC | 06:19 | |
| *** AMike has quit IRC | 06:21 | |
| *** nadya_ has joined #openstack-ceilometer | 06:27 | |
| *** nadya_ has quit IRC | 06:37 | |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/ceilometer: Fixed concurrency error in session handling https://review.openstack.org/66768 | 06:43 |
| openstackgerrit | Alexei Kornienko proposed a change to openstack/ceilometer: Run unit tests against MySQL https://review.openstack.org/59489 | 06:43 |
| *** boris-42 has quit IRC | 07:02 | |
| *** coolsvap has joined #openstack-ceilometer | 07:13 | |
| *** nadya_ has joined #openstack-ceilometer | 07:17 | |
| *** ildikov_ has joined #openstack-ceilometer | 07:22 | |
| *** nadya_ has quit IRC | 07:23 | |
| *** sayali has joined #openstack-ceilometer | 07:27 | |
| *** sayali_ has joined #openstack-ceilometer | 07:27 | |
| *** xuhanp_ has joined #openstack-ceilometer | 07:54 | |
| *** AMike has joined #openstack-ceilometer | 07:57 | |
| *** AMike has joined #openstack-ceilometer | 07:57 | |
| *** boris-42 has joined #openstack-ceilometer | 08:04 | |
| Alexei_987 | jd__: ping | 08:06 |
| *** sayali has quit IRC | 08:09 | |
| *** sayali has joined #openstack-ceilometer | 08:10 | |
| openstackgerrit | Andrey Kurilin proposed a change to openstack/python-ceilometerclient: Replace inheritance hierarchy with composition https://review.openstack.org/60857 | 08:14 |
| sileht | Alexei_987, hi | 08:22 |
| Alexei_987 | Hi | 08:22 |
| sileht | Alexei_987, it seems we are working on the same stuff:) | 08:22 |
| Alexei_987 | what do you mean exactly? | 08:22 |
| sileht | Alexei_987, I have start to write this https://review.openstack.org/#/c/65677/3/ceilometer/openstack/common/db/sqlalchemy/session.py | 08:23 |
| Alexei_987 | ah.. sorry :) | 08:23 |
| sileht | Alexei_987, it seems https://review.openstack.org/66768 resolved the same issue | 08:23 |
| Alexei_987 | i've find this bug yesterday evening | 08:23 |
| sileht | Alexei_987, no problem :) | 08:24 |
| Alexei_987 | sileht: hm it seems that your's solution is not completely working | 08:24 |
| Alexei_987 | why do you have out of memory errors? | 08:25 |
| Alexei_987 | sileht: or is it just a gate issue? | 08:26 |
| sileht | Alexei_987, I'm working on mysql unit tests and we cannot run mysql and sqlite unit tests at the same times with the singleton implementation of the session/engine stuff of sqlalchemy | 08:26 |
| gibi | boris-42: Are you around? | 08:27 |
| Alexei_987 | sileht: yeah I have the same problem :) | 08:27 |
| boris-42 | gibi yep | 08:29 |
| gibi | boris-42: hi! We made the change in the complex filter based on your comments. Could you check what we did, please? :) | 08:30 |
| sileht | Alexei_987, I have a meeting now, we talk on that to solve this issue | 08:31 |
| ildikov_ | boris-42: hi | 08:31 |
| boris-42 | ildikov_ hi | 08:31 |
| boris-42 | gibi ildikov_ yep I will check today | 08:31 |
| sileht | s/we/we can talk later/ | 08:31 |
| Alexei_987 | sileht: ok ping me when you'll have some time to chat | 08:31 |
| ildikov_ | boris-42: regarding to the review comments, keep in mind that Icehouse-2 is really close ;) | 08:31 |
| gibi | boris-42: thanks for your effort! | 08:32 |
| boris-42 | ildikov_ yep yep | 08:32 |
| boris-42 | ildikov_ gibi I need this patch as well=) | 08:32 |
| ildikov_ | boris-42: thanks for reviewing | 08:32 |
| ildikov_ | boris-42: then we are in the same boat :) | 08:33 |
| boris-42 | ildikov_ yep | 08:33 |
| boris-42 | ildikov_ gibi thanks for work btw=) | 08:33 |
| gibi | boris-42: I enjoyed learning jsonschema during the process :) | 08:34 |
| boris-42 | gibi yep useful thing | 08:34 |
| boris-42 | gibi ildikov_ could you guys explain me one thing | 08:35 |
| boris-42 | gibi ildikov_ probably I don't know soemthing | 08:35 |
| openstackgerrit | liusheng proposed a change to openstack/python-ceilometerclient: reduce redundant parameter of some commands in CLI https://review.openstack.org/66776 | 08:35 |
| boris-42 | about this ComplexQuery | 08:35 |
| boris-42 | http://pastebin.com/N7RwMFcf | 08:36 |
| ildikov_ | boris-42: as for me, I like this task now, it's interesting and also full of learning new things, but well, it would be good to keep the schedule as we have one more BP related to the complex query | 08:36 |
| boris-42 | line 982 why we use here just plain string | 08:36 |
| boris-42 | i mean lines 979, 982 985 | 08:36 |
| boris-42 | not comments like # | 08:36 |
| boris-42 | how this work?) | 08:37 |
| gibi | boris-42: as far as I know wsme does not support json object directly | 08:37 |
| boris-42 | gibi take a look at 972, 982, 985 lines | 08:37 |
| boris-42 | gibi why we don't use comments instead of plain string? | 08:37 |
| gibi | aah so the comments | 08:37 |
| boris-42 | gibi yep | 08:37 |
| boris-42 | gibi or there is some dirty hacky functionality around this? | 08:38 |
| boris-42 | ildikov_ ^ | 08:39 |
| gibi | boris-42: don't know the background of it but it is used this way in other classes in the api used for wsme | 08:39 |
| boris-42 | probably jd__ knows something about it ^ | 08:40 |
| boris-42 | jd__ hi | 08:40 |
| gibi | boris-42: maybe it is used for doc generation. We are cheching it ... | 08:40 |
| boris-42 | gibi i don't like when I don't know what code makes=) | 08:40 |
| boris-42 | gibi probably some kittens will die somewehre because of this code=) | 08:41 |
| ildikov_ | boris-42: it looks like the API doc is generated from these strings | 08:43 |
| *** Alexei_987 has quit IRC | 08:43 | |
| boris-42 | ildikov_ are you sure?) | 08:44 |
| boris-42 | ildikov_ actually it's terrible praxis .. | 08:44 |
| ildikov_ | boris-42: e.g. in case of Meter in case of type there is also a reference included: "The meter type (see :ref:`measurements`)" | 08:44 |
| boris-42 | practice* | 08:44 |
| ildikov_ | boris-42: you can check the V2 doc here: http://docs.openstack.org/developer/ceilometer/webapi/v2.html#Meter | 08:46 |
| boris-42 | ildikov_ ok | 08:47 |
| ildikov_ | boris-42: for us it is out of the box now, but maybe later it can be considered to find a less terrible practice for generating the API doc | 08:47 |
| boris-42 | ildikov_ yep | 08:47 |
| boris-42 | ildikov_ fully agree | 08:47 |
| gibi | boris-42: nobody wants to see dead kittens around if it is not necessary :) | 08:50 |
| xuhanp_ | scroiset, ping | 08:53 |
| xuhanp_ | sileht, ping | 08:56 |
| scroiset | xuhanp_: pong | 08:57 |
| xuhanp_ | Have a question f you have a minute. | 08:57 |
| xuhanp_ | if | 08:57 |
| scroiset | np .. if can answer | 08:57 |
| scroiset | I | 08:57 |
| xuhanp_ | I am new to Ceilometer so I was checking the architecture developer doc to understand it's service structure. (http://docs.openstack.org/developer/ceilometer/architecture.html#detailed-description). | 08:57 |
| xuhanp_ | My original understanding of Ceilometer collector from that doc page is that it is responsible for notification collection from other openstack components. But after checking the code and my devstack env, I found that the notification collection code has been separated into notification agent and the service left in Ceilometer collector is to collect UDP data. Could you help me understand what kind of UDP data is collected and where is this | 08:58 |
| xuhanp_ | UDP data from? | 08:58 |
| xuhanp_ | is my question clear? | 08:58 |
| scroiset | it's clear. I never used the UDP collector, but I guess it accept any data from anywhere (not authÃ), You just have to send 'meter' to it .. | 08:59 |
| scroiset | I don't know any usecase ... It's keep secret :) | 09:00 |
| openstackgerrit | Nadya Privalova proposed a change to openstack/ceilometer: Insertion in HBase should be fixed https://review.openstack.org/52670 | 09:00 |
| xuhanp_ | scroiset, so just like the notification agent receive notification as the source? | 09:00 |
| scroiset | xuhanp_: just like collector receving meters from 'notification handler' and 'pollsters (compute, central) | 09:01 |
| xuhanp_ | then just resend that UDP data into RPC to record it in Database? | 09:02 |
| scroiset | yes, IIRC | 09:04 |
| nprivalova | As I know UDP is used only in publishers. There are RPC and UDP publishers | 09:04 |
| *** coolsvap has quit IRC | 09:04 | |
| xuhanp_ | so can I ask why collector need to collect data from UDP publishers? sorry for the silly questions! | 09:05 |
| scroiset | nprivalova: There is a udp collector too | 09:05 |
| scroiset | xuhanp_: Somebody need it! | 09:05 |
| scroiset | "git blame" is telling me that jd__ wrote the udp collector .. may be ask to him? | 09:07 |
| nprivalova | scroiset, could you please give a link on the code? | 09:07 |
| xuhanp_ | jd__, ping | 09:08 |
| nprivalova | scroiset, I've found | 09:08 |
| *** coolsvap has joined #openstack-ceilometer | 09:09 | |
| *** coolsvap has quit IRC | 09:10 | |
| xuhanp_ | scroiset, nprivalova https://github.com/openstack/ceilometer/blob/master/ceilometer/collector.py#L63 | 09:13 |
| xuhanp_ | the udp collector code | 09:13 |
| *** SergeyLukjanov has joined #openstack-ceilometer | 09:13 | |
| scroiset | xuhanp_: yes | 09:13 |
| scroiset | started here https://github.com/openstack/ceilometer/blob/master/ceilometer/collector.py#L56 | 09:13 |
| *** yassine has joined #openstack-ceilometer | 09:15 | |
| xuhanp_ | nprivalova mentioned that there are UDP publishers too, so is there a case that this collector wants to receive data published by other ceilometer agents? | 09:15 |
| xuhanp_ | maybe for data recording purpose? | 09:15 |
| xuhanp_ | just think loud here. | 09:16 |
| *** Alexei_987 has joined #openstack-ceilometer | 09:16 | |
| scroiset | xuhanp_: I guess you're right .. the loop is closed. | 09:23 |
| ildikov_ | xuhanp_: if I know right, then the data from compute-agent comes by using rpc publisher | 09:23 |
| ildikov_ | xuhanp_: see the figure of how samples are published, here: http://docs.openstack.org/developer/ceilometer/architecture.html#multi-publisher | 09:24 |
| xuhanp_ | ildikov_, yeah. that picture confused me by only pointing udp publisher to external system, not collector. | 09:25 |
| xuhanp_ | you are saying RPC and UDP are both used for collector to receive data from compute agent, right? | 09:25 |
| nprivalova | as I understand collector main goal is to understand how to deal with collected data. It may call directly "record_metering_data" if we want to store data into backends (Mongo, SQL...). it may also republish smth | 09:26 |
| xuhanp_ | nprivalova, you mean like this? https://github.com/openstack/ceilometer/blob/master/ceilometer/collector.py#L81 | 09:27 |
| nprivalova | so there is no udp-collector, collector just should know how to work with data recieved by udp-publisher | 09:28 |
| scroiset | nprivalova: I don't think collector can republish smth | 09:28 |
| nprivalova | scroiset, oh, sorry, I meant 'transform' | 09:31 |
| nprivalova | I don't know the concrete service, but in theory we may have a service that uses udp for sending notifications (instead of rpc). And collector will listen the port and process messages | 09:35 |
| nprivalova | as I know nova, glance and so on use only rpc now. because udp publisher was broken for a long time and nobody noticed that | 09:38 |
| xuhanp_ | nprivalova, what do you mean by that? should nova and glance be the receiver of ceilometer udp publisher? | 09:40 |
| nprivalova | xuhanp_, no | 09:40 |
| nprivalova | xuhanp_, I'm just saying that nove just sends notifications using rpc and ceilometer process it in collector. but instead of rpc udp may be used and colloctor is able to process notification in the sama way | 09:42 |
| nprivalova | nova and glance are just examples | 09:43 |
| xuhanp_ | nprivalova, Ok. got it | 09:43 |
| *** _ruhe is now known as ruhe | 09:46 | |
| scroiset | nprivalova: AFAIK, Oslo don't provide udp publisher for notifications just AMQP, only ceilo's agents (central, compute) are able to send by udp | 09:48 |
| jd__ | scroiset: ? | 09:51 |
| nprivalova | scroiset, I'm just trying to explain why we may need to listen udp in collector. We may imagine that there is a service that doesn't use oslo and just want to send smth to ceilometer. So Ceilometer is able to process that message | 09:51 |
| nprivalova | jd__, we're just trying to understand udp flow | 09:52 |
| jd__ | I don't have time to read the backlog right now but ask if you need | 09:52 |
| xuhanp_ | just wondering why UDP collector is needed and which service will send data to the UDP collector. | 09:53 |
| xuhanp_ | for example, nova glance can send UDP data or ceilometer compute agent will | 09:54 |
| scroiset | jd__: hi, just trying to explain/understand udp use case for xuhanp_ ^^ | 09:54 |
| jd__ | it's faster and lighter than using RPC | 09:56 |
| jd__ | it's good enough to transport metrics for alarming for example | 09:56 |
| jd__ | you can transport some metrics over UDP to go faster | 09:56 |
| nprivalova | jd__, do we have a service in OpenStack that sends smth to Ceilometer over UDP? | 09:57 |
| scroiset | nprivalova: I would say no, only ceilo's agents send by UDP, AFAIK | 09:59 |
| nprivalova | scroiset, I used to think in the same way, just wanted to be sure | 10:01 |
| jd__ | nprivalova: Ceilometer itself :) | 10:02 |
| jd__ | it's a sample transport only | 10:02 |
| xuhanp_ | jd__, scroiset, nprivalova, that helps a lot. Thanks! just want to confirm that collector record data by directly calling "record_metering_data" or by RPC? | 10:04 |
| nprivalova | xuhanp_, collector has a dispatcher_manager who knows where the data should be written. it doesn't matter how a message was received (by UDP or RPC) | 10:06 |
| xuhanp_ | nprivalova, I will check that. Thanks | 10:08 |
| *** xuhanp_ has quit IRC | 10:12 | |
| *** flwang has quit IRC | 10:16 | |
| *** ruhe is now known as ruhe_away | 10:37 | |
| *** ruhe_away is now known as _ruhe | 10:46 | |
| *** eglynn has joined #openstack-ceilometer | 11:05 | |
| *** boris-42_ has joined #openstack-ceilometer | 11:18 | |
| *** boris-42 has quit IRC | 11:19 | |
| *** boris-42_ is now known as boris-42 | 11:19 | |
| *** Alexei_9871 has joined #openstack-ceilometer | 11:22 | |
| Alexei_9871 | sileht: ping | 11:40 |
| *** sayali has quit IRC | 11:46 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Replace mongo aggregation with plain ol' map-reduce https://review.openstack.org/65962 | 11:47 |
| openstackgerrit | A change was merged to openstack/python-ceilometerclient: Python 3: fix test_sample_list https://review.openstack.org/65836 | 11:48 |
| openstackgerrit | A change was merged to openstack/python-ceilometerclient: client looking at wrong cacert argument name https://review.openstack.org/65082 | 11:48 |
| openstackgerrit | A change was merged to openstack/ceilometer: cacert is not picked up correctly by alarm services https://review.openstack.org/65087 | 11:48 |
| openstackgerrit | A change was merged to openstack/python-ceilometerclient: Supports bash_completion for ceilometerclient https://review.openstack.org/63718 | 11:48 |
| openstackgerrit | A change was merged to openstack/ceilometer: tests: allow to skip if no database URL https://review.openstack.org/65081 | 11:48 |
| *** prad has joined #openstack-ceilometer | 12:52 | |
| *** sayali has joined #openstack-ceilometer | 12:54 | |
| *** shengjiemin has joined #openstack-ceilometer | 13:13 | |
| *** flwang has joined #openstack-ceilometer | 13:18 | |
| *** _ruhe is now known as ruhe | 13:21 | |
| sileht | Alexei_9871, sorry I'm still in a meeting, I'm not sure I have the time today finally, but for sure I'm free tomorrow (no meeting planned :) ) | 13:23 |
| Alexei_9871 | sileht: ok no problem | 13:24 |
| *** julienvey_ has joined #openstack-ceilometer | 13:28 | |
| *** julienvey_ has quit IRC | 13:30 | |
| *** jdob has joined #openstack-ceilometer | 13:30 | |
| *** julienvey_ has joined #openstack-ceilometer | 13:30 | |
| nprivalova | sileht, hi! | 13:34 |
| *** xuhanp has joined #openstack-ceilometer | 13:36 | |
| nprivalova | sileht, ah, you're on meeting, sorry :) | 13:36 |
| *** eglynn is now known as eglynn-lunch | 13:36 | |
| *** boris-42_ has joined #openstack-ceilometer | 13:38 | |
| *** boris-42 has quit IRC | 13:39 | |
| *** boris-42_ is now known as boris-42 | 13:40 | |
| *** gordc has joined #openstack-ceilometer | 13:40 | |
| nprivalova | gordc, hi! could you please take a look https://review.openstack.org/#/c/66580/ when you have time? I will work on migration, but want to be sure that idea is ok | 13:48 |
| gordc | nprivalova: sure. i'll try to get around to it. i should have time later today. | 13:49 |
| nprivalova | gordc, ok, thanks! | 13:50 |
| *** thomasem has joined #openstack-ceilometer | 13:51 | |
| thomasem | Goooood morning! | 13:52 |
| ildikov_ | thomasem: ...Vietnaaaaaaaaaam! :) | 13:53 |
| thomasem | ildikov_, good evening! :) | 13:53 |
| thomasem | what time is it there? | 13:54 |
| ildikov_ | thomasem: 2:54 pm | 13:54 |
| thomasem | ildikov_, Good afternoon. :D | 13:54 |
| ildikov_ | thomasem: it's ok now, I'll suffer on the meeting tonight at 10 pm | 13:54 |
| ildikov_ | thomasem: thanks :D | 13:55 |
| thomasem | ildikov_, Ohhh! :( | 13:55 |
| ildikov_ | thomasem: my luck is that I like coffee | 13:56 |
| thomasem | Ohhh me too. Just got my cup for the morning. :) | 13:56 |
| ildikov_ | thomasem: today I will definitely need one, as we have the important Icehouse-2 status topic on the agenda | 13:56 |
| ildikov_ | thomasem: I've just had mine for the afternoon ;) | 13:57 |
| *** dperaza1 has left #openstack-ceilometer | 13:57 | |
| thomasem | Yay! | 13:59 |
| *** ruhe is now known as ruhe_away | 14:02 | |
| *** ildikov_ has quit IRC | 14:04 | |
| *** ildikov_ has joined #openstack-ceilometer | 14:05 | |
| *** sayali has quit IRC | 14:05 | |
| *** ruhe_away is now known as ruhe | 14:10 | |
| *** prad has quit IRC | 14:16 | |
| *** SergeyLukjanov_ has joined #openstack-ceilometer | 14:36 | |
| *** SergeyLukjanov has quit IRC | 14:39 | |
| *** SergeyLukjanov_ is now known as SergeyLukjanov | 14:39 | |
| *** dmsimard has joined #openstack-ceilometer | 14:41 | |
| *** dmsimard has quit IRC | 14:42 | |
| *** dmsimard has joined #openstack-ceilometer | 14:42 | |
| *** ruhe is now known as ruhe_away | 14:44 | |
| *** SergeyLukjanov has quit IRC | 14:44 | |
| *** ruhe_away is now known as ruhe | 14:44 | |
| *** ruhe is now known as ruhe_away | 14:45 | |
| *** SergeyLukjanov_ has joined #openstack-ceilometer | 14:45 | |
| *** SergeyLukjanov_ is now known as SergeyLukjanov | 14:45 | |
| *** ruhe_away is now known as ruhe | 14:46 | |
| *** jmckind has joined #openstack-ceilometer | 14:50 | |
| *** SergeyLukjanov is now known as SergeyLukjanov_ | 14:53 | |
| *** boris-42_ has joined #openstack-ceilometer | 14:54 | |
| *** SergeyLukjanov_ is now known as SergeyLukjanov | 14:54 | |
| *** prad has joined #openstack-ceilometer | 14:54 | |
| *** kwhitney1 has quit IRC | 14:56 | |
| *** boris-42_ has quit IRC | 14:56 | |
| *** boris-42 has quit IRC | 14:57 | |
| *** tongli has joined #openstack-ceilometer | 15:09 | |
| *** dmsimard has left #openstack-ceilometer | 15:09 | |
| *** SergeyLukjanov has quit IRC | 15:09 | |
| *** SergeyLukjanov_ has joined #openstack-ceilometer | 15:10 | |
| *** SergeyLukjanov_ is now known as SergeyLukjanov | 15:10 | |
| *** herndon has joined #openstack-ceilometer | 15:10 | |
| *** zqfan has joined #openstack-ceilometer | 15:17 | |
| *** SergeyLukjanov has quit IRC | 15:19 | |
| *** SergeyLukjanov has joined #openstack-ceilometer | 15:20 | |
| nprivalova | folks, help me please! question about testing. testr runs in parallel so how do we test real backends? I do not see in the code any fake-names for tables | 15:20 |
| *** SergeyLukjanov has quit IRC | 15:22 | |
| *** SergeyLukjanov has joined #openstack-ceilometer | 15:23 | |
| *** xuhanp has quit IRC | 15:23 | |
| *** zqfan has quit IRC | 15:24 | |
| *** ruhe is now known as ruhe_away | 15:25 | |
| openstackgerrit | Julien Danjou proposed a change to openstack/ceilometer: Update oslo https://review.openstack.org/65661 | 15:28 |
| nprivalova | jd__, please help me to find expert in tests | 15:29 |
| jd__ | you found him | 15:29 |
| jd__ | what's your problem nprivalova? | 15:29 |
| nprivalova | jd__, testr runs in parallel so how do we test real backends? I do not see in the code any fake-names for tables | 15:29 |
| jd__ | so build a db for each test | 15:30 |
| jd__ | check ceilometer/tests/db.py | 15:30 |
| jd__ | s/so/so we/ | 15:30 |
| nprivalova | jd__, the problem that we get 'table already exists' exception all the time | 15:30 |
| jd__ | nprivalova: are you trying to test several SQL backends? | 15:31 |
| nprivalova | no, HBase :( but as I understand the same problem should be in SQL backends. Ah, do you use the same table names but different databases? | 15:32 |
| jd__ | nprivalova: exactly | 15:33 |
| jd__ | you can do that for HBase too if you can | 15:33 |
| nprivalova | jd__, in HBase we have only table names, no 'namespaces' above | 15:33 |
| *** ruhe_away is now known as _ruhe | 15:34 | |
| jd__ | ah :-/ | 15:34 |
| nprivalova | jd__, and what do you do with Mongo? | 15:34 |
| jd__ | nprivalova: we use a different db in MongoDB too | 15:35 |
| jd__ | maybe you can suffix your tables? | 15:35 |
| nprivalova | great news for us ^) | 15:35 |
| nprivalova | jd__, thank you so much! we will think about approaches. Yep, look like we have to use prefixes somehow | 15:36 |
| openstackgerrit | A change was merged to openstack/ceilometer: Fix to tackle instances without an image assigned https://review.openstack.org/65356 | 15:38 |
| *** eglynn-lunch is now known as eglynn | 15:38 | |
| jd__ | nprivalova: ok cool let me know | 15:39 |
| *** SergeyLukjanov is now known as SergeyLukjanov_a | 15:39 | |
| jd__ | nprivalova: otherwise the other approach is to use a lock to run only one test at a time, but it makes the test muuuuch slower | 15:40 |
| *** SergeyLukjanov_a is now known as SergeyLukjanov_ | 15:40 | |
| nprivalova | jd__, yes, especially for HBase. I think it would be not too difficult to generate table names for each test | 15:41 |
| nprivalova | jd__, btw, I've created a bp about unit tests on real HBase, I've added you as approver | 15:42 |
| jd__ | nprivalova: ok nice | 15:43 |
| *** SergeyLukjanov_ is now known as SergeyLukjanov | 15:43 | |
| *** SergeyLukjanov is now known as SergeyLukjanov_ | 15:44 | |
| jd__ | nprivalova: if you want it to be tracked, add a milestone to i3 | 15:44 |
| jd__ | otherwise I've marked it as approve | 15:44 |
| jd__ | so you're good :) | 15:44 |
| nprivalova | ok, will add | 15:45 |
| openstackgerrit | Eoghan Glynn proposed a change to openstack/ceilometer: Add new rate-based disk and network pipelines https://review.openstack.org/66658 | 15:48 |
| nprivalova | eglynn, sorry I still didn't answer your comments in etherpad. Will do ASAP | 15:49 |
| eglynn | nprivalova: np! | 15:49 |
| tongli | @eglynn, ping | 15:58 |
| eglynn | tongli: 'sup? | 15:58 |
| tongli | @eglynn, trying to start implementing a BP, and had few emails exchanges with Julien and copied you on, | 15:59 |
| tongli | @eglynn, wonder if you have time to take a look and give some thoughts on? | 15:59 |
| eglynn | tongli: k, will do shortly | 15:59 |
| tongli | @eglynn, great. thanks. | 16:00 |
| *** SergeyLukjanov_ is now known as SergeyLukjanov | 16:13 | |
| *** _ruhe is now known as ruhe | 16:14 | |
| *** shengjiemin has quit IRC | 16:31 | |
| *** boris-42 has joined #openstack-ceilometer | 16:49 | |
| openstackgerrit | Balazs Gibizer proposed a change to openstack/ceilometer: Implements metadata query for complex query feature https://review.openstack.org/66891 | 16:51 |
| openstackgerrit | Balazs Gibizer proposed a change to openstack/ceilometer: Implements "not" operator for complex query https://review.openstack.org/66892 | 16:51 |
| *** yassine has quit IRC | 16:54 | |
| *** yassine has joined #openstack-ceilometer | 16:54 | |
| *** ildikov_ has quit IRC | 16:55 | |
| *** ildikov_ has joined #openstack-ceilometer | 17:11 | |
| openstackgerrit | Eoghan Glynn proposed a change to openstack/ceilometer: Add new rate-based disk and network pipelines https://review.openstack.org/66658 | 17:15 |
| *** ildikov_ is now known as ildikov_afk | 17:26 | |
| *** shengjiemin has joined #openstack-ceilometer | 17:32 | |
| *** kwhitney1 has joined #openstack-ceilometer | 17:39 | |
| *** shengjiemin has quit IRC | 17:41 | |
| *** ruhe is now known as ruhe_away | 17:49 | |
| *** herndon has quit IRC | 17:55 | |
| *** eglynn has quit IRC | 17:58 | |
| *** ruhe_away is now known as _ruhe | 17:58 | |
| *** Alexei_9871 has quit IRC | 17:59 | |
| *** nadya_ has joined #openstack-ceilometer | 18:11 | |
| *** nadya_ has quit IRC | 18:24 | |
| *** openstackgerrit has quit IRC | 18:27 | |
| *** openstackgerrit has joined #openstack-ceilometer | 18:27 | |
| openstackgerrit | ZhiQiang Fan proposed a change to openstack/ceilometer: Use DEFAULT section for dispatcher in doc https://review.openstack.org/66908 | 18:41 |
| *** nadya_ has joined #openstack-ceilometer | 18:49 | |
| *** julienvey_ has quit IRC | 18:50 | |
| *** nadya_ has quit IRC | 19:18 | |
| *** jhurlbert has joined #openstack-ceilometer | 19:19 | |
| *** nadya_ has joined #openstack-ceilometer | 19:23 | |
| *** herndon has joined #openstack-ceilometer | 19:23 | |
| *** _ruhe is now known as ruhe | 19:24 | |
| *** jaypipes has joined #openstack-ceilometer | 19:41 | |
| *** julienvey_ has joined #openstack-ceilometer | 20:10 | |
| *** thomasem has quit IRC | 20:18 | |
| *** prad has quit IRC | 20:19 | |
| *** prad has joined #openstack-ceilometer | 20:23 | |
| *** sayali has joined #openstack-ceilometer | 20:27 | |
| *** sayali has quit IRC | 20:28 | |
| *** ruhe is now known as _ruhe | 20:28 | |
| *** yassine has quit IRC | 20:35 | |
| *** yassine has joined #openstack-ceilometer | 20:35 | |
| *** yassine has quit IRC | 20:36 | |
| nadya_ | gordc, are you around? | 20:42 |
| *** ildikov_afk is now known as ildikov_ | 20:49 | |
| gordc | nadya_: sorry about that, was working on another machine. | 20:56 |
| openstackgerrit | John Herndon proposed a change to openstack/ceilometer: Return trait type from Event api https://review.openstack.org/66936 | 20:57 |
| *** eglynn has joined #openstack-ceilometer | 21:00 | |
| nadya_ | gordc, hi :)! I wanted to ask about list queries | 21:00 |
| *** julienvey_ has quit IRC | 21:01 | |
| gordc | nadya_: sure. go for it. i'll answer in between meeting | 21:01 |
| nadya_ | gordc, is it queryable for sql now? I see in the code "Unknown metadata type. key ... will not be queryable" | 21:02 |
| gordc | nadya_: if i remember correctly. it should work. when i built metadata query, i built the keys as 'a.b.c[0]' something along that format... something similar to how mongo builds keys | 21:04 |
| gordc | that said, since i apparently didn't add a test for it, i can't confirm that. | 21:04 |
| gordc | nadya_: i think regardless if it works for sql now. we should think about whether we need it and see if this migration makes it easier or harder to accomplish that. | 21:07 |
| *** SergeyLukjanov is now known as SergeyLukjanov_ | 21:08 | |
| *** thomasem has joined #openstack-ceilometer | 21:20 | |
| *** boris-42 has quit IRC | 21:36 | |
| *** boris-42 has joined #openstack-ceilometer | 21:40 | |
| nadya_ | gordc, sorry we were on meeting channel :) So, as I see it to make list queries simple we need use dict_to_keyval instead of recursive_keypairs | 21:45 |
| *** eglynn has quit IRC | 21:45 | |
| nadya_ | gordc, but we can't | 21:45 |
| nadya_ | gordc, and one more thing. I realized that flatten_metadata during request data (not POST) is needed only for Mongo now. Because HBase and SQL do it's own 'flatten' during record_metering_data | 21:46 |
| gordc | nadya_: is a safe alternative to keep metadata tables as is? | 21:47 |
| gordc | nadya_: if i recall, dict_to_keyval is only use to build metadata tables... which are really just a secondary table to resource_metadata value in meter table. | 21:48 |
| nadya_ | gordc, as I understand Mongo and SQL keeps original data. But additionally SQL stores metadata table | 21:49 |
| gordc | yes. so maybe just keep the metadata just for querying? although i can see that being an issue later where people ask why we have resource_metadata stored in two places. | 21:50 |
| nadya_ | gordc, yep, it looks incorrect | 21:50 |
| *** tongli has quit IRC | 21:51 | |
| nadya_ | actually, I thought about removing flatten_metadata at all. just add the code to mongo get_meter | 21:52 |
| nadya_ | gordc, but it's too risky | 21:52 |
| gordc | nadya_: that might be a plan. i haven't looked into that. it'd be nice to have a standard interface and let the backends flatten as they like (although i don't know if that works in real life) | 21:53 |
| gordc | nadya_: i'd probably need to think about it for a bit...but you're right. list query will be hard using recursive_keypairs. | 21:53 |
| *** asalkeld has joined #openstack-ceilometer | 21:54 | |
| openstackgerrit | Pradeep Kilambi proposed a change to openstack/ceilometer: Fix to handle null threshold_rule values https://review.openstack.org/66757 | 21:54 |
| nadya_ | gordc, will think about this once more too. Thank you for discussion! need to go now | 21:55 |
| gordc | nadya_: cool cool. ttyl. | 21:55 |
| *** nadya_ has quit IRC | 21:55 | |
| *** jdob has quit IRC | 21:55 | |
| *** prad has quit IRC | 21:59 | |
| *** EmilienM has quit IRC | 22:10 | |
| *** EmilienM has joined #openstack-ceilometer | 22:12 | |
| *** thomasem has quit IRC | 22:19 | |
| *** jmckind has quit IRC | 22:25 | |
| *** jaypipes has quit IRC | 22:25 | |
| *** herndon has quit IRC | 22:59 | |
| *** herndon has joined #openstack-ceilometer | 23:10 | |
| openstackgerrit | gordon chung proposed a change to openstack/ceilometer: ERRORs in ceilometer-acentral log after succesful tempest run https://review.openstack.org/66973 | 23:14 |
| *** elextro_ has quit IRC | 23:15 | |
| *** dragondm has quit IRC | 23:17 | |
| *** dragondm has joined #openstack-ceilometer | 23:21 | |
| *** xuhanp has joined #openstack-ceilometer | 23:42 | |
| *** gordc has quit IRC | 23:48 | |
| openstackgerrit | Cyril Roelandt proposed a change to openstack/python-ceilometerclient: Sync with oslo https://review.openstack.org/65716 | 23:54 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!