| *** diogogmt has quit IRC | 00:02 | |
| *** diogogmt has joined #openstack-telemetry | 00:03 | |
| *** thorst has joined #openstack-telemetry | 00:13 | |
| *** jwcroppe has quit IRC | 00:14 | |
| *** jwcroppe has joined #openstack-telemetry | 00:14 | |
| *** diogogmt has quit IRC | 00:16 | |
| *** jwcroppe has quit IRC | 00:18 | |
| *** diogogmt has joined #openstack-telemetry | 00:21 | |
| *** thorst has quit IRC | 00:21 | |
| *** jefrite has joined #openstack-telemetry | 00:27 | |
| *** julim has quit IRC | 00:27 | |
| *** g3ek has quit IRC | 00:45 | |
| *** g3ek has joined #openstack-telemetry | 00:50 | |
| *** thorst has joined #openstack-telemetry | 00:59 | |
| *** khushbu_ has quit IRC | 00:59 | |
| *** cheneydc has joined #openstack-telemetry | 01:04 | |
| *** cheneydc has quit IRC | 01:07 | |
| openstackgerrit | liusheng proposed openstack/ceilometer: Add a tool to clean the legacy alarm tables https://review.openstack.org/312399 | 01:11 |
|---|---|---|
| *** openstack has joined #openstack-telemetry | 01:26 | |
| *** g3ek has quit IRC | 01:35 | |
| openstackgerrit | gordon chung proposed openstack/gnocchi: simplify model loading https://review.openstack.org/335248 | 01:36 |
| *** g3ek has joined #openstack-telemetry | 01:41 | |
| *** thorst has quit IRC | 01:43 | |
| *** thorst has joined #openstack-telemetry | 01:44 | |
| *** thorst has quit IRC | 01:52 | |
| openstackgerrit | gordon chung proposed openstack/gnocchi: simplify model loading https://review.openstack.org/335248 | 02:19 |
| openstackgerrit | gordon chung proposed openstack/gnocchi: simplify model loading https://review.openstack.org/335248 | 02:20 |
| *** khushbu_ has joined #openstack-telemetry | 02:21 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: refactor service to be less pipeline dependent https://review.openstack.org/323615 | 02:21 |
| *** g3ek has quit IRC | 02:21 | |
| *** g3ek has joined #openstack-telemetry | 02:30 | |
| *** khushbu_ has quit IRC | 02:39 | |
| *** khushbu has joined #openstack-telemetry | 02:40 | |
| *** sheel has joined #openstack-telemetry | 02:48 | |
| *** thorst has joined #openstack-telemetry | 02:50 | |
| *** thorst has quit IRC | 02:57 | |
| *** cheneydc has joined #openstack-telemetry | 03:03 | |
| *** jwcroppe has joined #openstack-telemetry | 03:12 | |
| *** g3ek has quit IRC | 03:14 | |
| *** g3ek has joined #openstack-telemetry | 03:20 | |
| openstackgerrit | Fei Long Wang proposed openstack/aodh: Add pre-signed support for zaqar notifier https://review.openstack.org/335289 | 03:33 |
| flwang | hi team, i need some suggestion about the unit test for this patch https://review.openstack.org/335289 | 03:34 |
| flwang | liusheng: ^ | 03:34 |
| flwang | any comment will be appreciated | 03:34 |
| openstackgerrit | Merged openstack/gnocchi: Use pbr WSGI script to build gnocchi-api https://review.openstack.org/320519 | 03:39 |
| *** jwcroppe has quit IRC | 03:41 | |
| *** thorst has joined #openstack-telemetry | 03:55 | |
| *** g3ek has quit IRC | 04:02 | |
| flwang | anybody around? | 04:02 |
| *** thorst has quit IRC | 04:03 | |
| *** g3ek has joined #openstack-telemetry | 04:04 | |
| *** links has joined #openstack-telemetry | 04:23 | |
| *** jwcroppe has joined #openstack-telemetry | 04:30 | |
| *** g3ek has quit IRC | 04:53 | |
| *** rcernin has joined #openstack-telemetry | 04:53 | |
| *** khushbu has quit IRC | 04:58 | |
| *** g3ek has joined #openstack-telemetry | 04:58 | |
| *** pcaruana has quit IRC | 04:58 | |
| *** khushbu_ has joined #openstack-telemetry | 04:59 | |
| *** flwang1 has quit IRC | 05:00 | |
| *** thorst has joined #openstack-telemetry | 05:01 | |
| *** M00nr41n has quit IRC | 05:06 | |
| *** jwcroppe has quit IRC | 05:07 | |
| *** thorst has quit IRC | 05:08 | |
| *** _nadya_ has joined #openstack-telemetry | 05:13 | |
| *** jwcroppe has joined #openstack-telemetry | 05:27 | |
| *** rcernin has quit IRC | 05:33 | |
| *** g3ek has quit IRC | 05:40 | |
| *** _nadya_ has quit IRC | 05:46 | |
| *** _nadya_ has joined #openstack-telemetry | 05:47 | |
| *** g3ek has joined #openstack-telemetry | 05:48 | |
| *** _nadya_ has quit IRC | 05:51 | |
| *** khushbu_ has quit IRC | 05:57 | |
| *** khushbu has joined #openstack-telemetry | 05:57 | |
| *** yprokule has joined #openstack-telemetry | 05:58 | |
| *** thorst has joined #openstack-telemetry | 06:05 | |
| *** rcernin has joined #openstack-telemetry | 06:06 | |
| *** M00nr41n has joined #openstack-telemetry | 06:08 | |
| *** thorst has quit IRC | 06:12 | |
| *** pcaruana has joined #openstack-telemetry | 06:16 | |
| *** pcaruana is now known as pcaruana|afk| | 06:19 | |
| *** M00nr41n has quit IRC | 06:22 | |
| *** M00nr41n has joined #openstack-telemetry | 06:23 | |
| *** g3ek has quit IRC | 06:27 | |
| *** tomoiaga1 has joined #openstack-telemetry | 06:28 | |
| *** _nadya_ has joined #openstack-telemetry | 06:32 | |
| *** M00nr41n has quit IRC | 06:33 | |
| *** M00nr41n has joined #openstack-telemetry | 06:34 | |
| *** _nadya_ has quit IRC | 06:36 | |
| *** g3ek has joined #openstack-telemetry | 06:37 | |
| *** _nadya_ has joined #openstack-telemetry | 06:37 | |
| *** pcaruana|afk| is now known as pcaruana | 06:49 | |
| *** belmoreira has joined #openstack-telemetry | 06:56 | |
| *** _nadya_ has quit IRC | 07:02 | |
| *** _nadya_ has joined #openstack-telemetry | 07:04 | |
| *** _nadya_ has quit IRC | 07:05 | |
| *** tesseract- has joined #openstack-telemetry | 07:09 | |
| *** amoralej|off is now known as amoralej | 07:10 | |
| *** thorst has joined #openstack-telemetry | 07:10 | |
| *** thorst has quit IRC | 07:17 | |
| *** g3ek has quit IRC | 07:21 | |
| *** jwcroppe has quit IRC | 07:23 | |
| *** g3ek has joined #openstack-telemetry | 07:26 | |
| *** flwang1 has joined #openstack-telemetry | 07:26 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/aodh: Imported Translations from Zanata https://review.openstack.org/335344 | 07:26 |
| *** liusheng has quit IRC | 07:34 | |
| *** liusheng has joined #openstack-telemetry | 07:35 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: ceph: uses only one ioctx https://review.openstack.org/333965 | 07:35 |
| *** khushbu has quit IRC | 07:38 | |
| *** khushbu has joined #openstack-telemetry | 08:03 | |
| *** g3ek has quit IRC | 08:10 | |
| *** jwcroppe has joined #openstack-telemetry | 08:12 | |
| *** thorst has joined #openstack-telemetry | 08:15 | |
| *** g3ek has joined #openstack-telemetry | 08:18 | |
| *** thorst has quit IRC | 08:23 | |
| *** _nadya_ has joined #openstack-telemetry | 08:23 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Add grenade plugin https://review.openstack.org/306329 | 08:25 |
| *** kaszkiet has quit IRC | 08:26 | |
| *** jwcroppe has quit IRC | 08:28 | |
| *** flwang1 has quit IRC | 08:39 | |
| *** cheneydc has quit IRC | 08:41 | |
| *** yassine__ has joined #openstack-telemetry | 08:51 | |
| *** khushbu has quit IRC | 09:00 | |
| *** g3ek has quit IRC | 09:01 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Add grenade plugin https://review.openstack.org/306329 | 09:02 |
| *** chlong has quit IRC | 09:04 | |
| *** nijaba has quit IRC | 09:06 | |
| *** nijaba has joined #openstack-telemetry | 09:07 | |
| *** nijaba has quit IRC | 09:07 | |
| *** nijaba has joined #openstack-telemetry | 09:07 | |
| *** g3ek has joined #openstack-telemetry | 09:07 | |
| *** pece has joined #openstack-telemetry | 09:08 | |
| *** chlong has joined #openstack-telemetry | 09:17 | |
| *** thorst has joined #openstack-telemetry | 09:20 | |
| openstackgerrit | Merged openstack/gnocchi: simplify model loading https://review.openstack.org/335248 | 09:21 |
| *** flwang1 has joined #openstack-telemetry | 09:26 | |
| *** thorst has quit IRC | 09:28 | |
| openstackgerrit | Merged openstack/ceilometer: Add a tool to clean the legacy alarm tables https://review.openstack.org/312399 | 09:28 |
| openstackgerrit | Merged openstack/gnocchi: _carbonara: use tooz heartbeat management https://review.openstack.org/330982 | 09:43 |
| openstackgerrit | Merged openstack/gnocchi: devstack: Fix requirement typo https://review.openstack.org/335007 | 09:50 |
| *** g3ek has quit IRC | 09:52 | |
| *** g3ek has joined #openstack-telemetry | 09:57 | |
| openstackgerrit | Merged openstack/gnocchi: sqlalchemy: fix MySQL error handling in list_resources https://review.openstack.org/331020 | 10:04 |
| *** thorst has joined #openstack-telemetry | 10:25 | |
| *** thorst has quit IRC | 10:33 | |
| *** g3ek has quit IRC | 10:40 | |
| *** jefrite has quit IRC | 10:46 | |
| *** g3ek has joined #openstack-telemetry | 10:48 | |
| *** g3ek has quit IRC | 11:32 | |
| *** g3ek has joined #openstack-telemetry | 11:38 | |
| *** thorst has joined #openstack-telemetry | 11:42 | |
| *** jwcroppe has joined #openstack-telemetry | 11:58 | |
| *** _nadya_ has quit IRC | 12:00 | |
| *** pece has quit IRC | 12:05 | |
| *** jwcroppe has quit IRC | 12:05 | |
| *** vishwanathj has joined #openstack-telemetry | 12:13 | |
| sileht | jd__, I think I'm blocked for grenade because we don't follow openstack branches naming convention, and the devstack-vm-scripts is not flexible enought to allow to configure the BASE BRANCH and the TARGET BRANCH | 12:17 |
| *** ljxiash has joined #openstack-telemetry | 12:17 | |
| *** kelv has quit IRC | 12:20 | |
| *** g3ek has quit IRC | 12:23 | |
| jd__ | sileht: can't we hack it? | 12:24 |
| sileht | jd__, Not really, we need a new feature in devstack-vm-gate-wrap to support our scheme | 12:26 |
| jd__ | does not sound that hard | 12:26 |
| sileht | jd__, I hate this project, it's untestable locally | 12:26 |
| jd__ | sileht: do we have any other options? | 12:27 |
| sileht | I don't think so | 12:27 |
| *** pradk has joined #openstack-telemetry | 12:28 | |
| *** g3ek has joined #openstack-telemetry | 12:29 | |
| *** vishwanathj has quit IRC | 12:37 | |
| *** gordc has joined #openstack-telemetry | 12:37 | |
| *** diogogmt has quit IRC | 12:38 | |
| *** amoralej is now known as amoralej|lunch | 12:41 | |
| *** peristeri has joined #openstack-telemetry | 12:47 | |
| sileht | jd__, let's see if infra have a better idea: https://review.openstack.org/335479 | 12:49 |
| *** diogogmt has joined #openstack-telemetry | 12:52 | |
| openstackgerrit | Merged openstack/ceilometer: Config: no need to set default=None https://review.openstack.org/334815 | 12:53 |
| *** M00nr41n has quit IRC | 13:11 | |
| *** c64cosmin has joined #openstack-telemetry | 13:27 | |
| *** _nadya_ has joined #openstack-telemetry | 13:28 | |
| *** diogogmt has quit IRC | 13:32 | |
| *** links has quit IRC | 13:36 | |
| *** amoralej|lunch is now known as amoralej | 13:37 | |
| *** gordc has quit IRC | 13:38 | |
| *** Qiming has quit IRC | 13:44 | |
| *** Qiming has joined #openstack-telemetry | 13:44 | |
| *** r-mibu has quit IRC | 13:48 | |
| *** r-mibu has joined #openstack-telemetry | 13:48 | |
| *** thumpba has joined #openstack-telemetry | 13:50 | |
| *** thumpba has quit IRC | 13:51 | |
| *** KrishR has joined #openstack-telemetry | 13:51 | |
| *** thumpba has joined #openstack-telemetry | 13:51 | |
| openstackgerrit | Julien Danjou proposed openstack/aodh: sqlalchemy: allow to upgrade schema from Ceilometer Liberty https://review.openstack.org/335508 | 13:51 |
| *** ametts has joined #openstack-telemetry | 13:56 | |
| *** liamji has joined #openstack-telemetry | 13:57 | |
| *** pece has joined #openstack-telemetry | 14:01 | |
| *** c64cosmin has quit IRC | 14:02 | |
| *** khushbu_ has joined #openstack-telemetry | 14:03 | |
| *** jwcroppe has joined #openstack-telemetry | 14:04 | |
| *** g3ek has quit IRC | 14:05 | |
| *** g3ek has joined #openstack-telemetry | 14:07 | |
| *** jwcroppe has quit IRC | 14:09 | |
| *** caisan has joined #openstack-telemetry | 14:26 | |
| *** tomoiaga1 has quit IRC | 14:33 | |
| *** khushbu_ has quit IRC | 14:38 | |
| *** khushbu has joined #openstack-telemetry | 14:46 | |
| openstackgerrit | Julien Danjou proposed openstack/aodh: sqlalchemy: allow to upgrade schema from Ceilometer Liberty https://review.openstack.org/335508 | 14:52 |
| *** g3ek has quit IRC | 14:55 | |
| *** g3ek has joined #openstack-telemetry | 15:03 | |
| *** thumpba has quit IRC | 15:04 | |
| *** tesseract- has quit IRC | 15:12 | |
| *** rcernin has quit IRC | 15:14 | |
| *** sheel has quit IRC | 15:15 | |
| *** catintheroof has joined #openstack-telemetry | 15:16 | |
| *** pcaruana has quit IRC | 15:17 | |
| jd__ | devstack, such a joke | 15:17 |
| jd__ | can't stack | 15:17 |
| *** gordc has joined #openstack-telemetry | 15:25 | |
| sileht | jd__, it won't work because applications don't work out of the box... | 15:26 |
| jd__ | clearly | 15:29 |
| jd__ | I just enabled Keystone, it does not work | 15:29 |
| jd__ | *awesome* | 15:29 |
| sileht | jd__, do you get the new 'I can't discover the identify version' ? | 15:30 |
| jd__ | sileht: YES | 15:30 |
| jd__ | exactly | 15:30 |
| jd__ | no clue why | 15:30 |
| sileht | jd__, I have noticed that if I have a openrc loaded it doesn't work anymore | 15:31 |
| *** rcernin has joined #openstack-telemetry | 15:31 | |
| sileht | jd__, I have logout/login or start a new clean shell to have stack.sh working | 15:31 |
| sileht | have/have to | 15:31 |
| *** pcaruana has joined #openstack-telemetry | 15:32 | |
| *** khushbu has quit IRC | 15:34 | |
| jd__ | let's try | 15:34 |
| *** thumpba has joined #openstack-telemetry | 15:41 | |
| *** khushbu has joined #openstack-telemetry | 15:43 | |
| *** belmoreira has quit IRC | 15:43 | |
| *** g3ek has quit IRC | 15:44 | |
| *** jmccarthy has joined #openstack-telemetry | 15:45 | |
| *** jmccarthy has left #openstack-telemetry | 15:46 | |
| jd__ | sileht: didn't help :((( | 15:48 |
| sileht | jd__, keystone logs said nothing ? | 15:48 |
| sileht | jd__, keystone client now say 'I can't discover the identify version' when server answer 500... | 15:49 |
| *** sheel has joined #openstack-telemetry | 15:49 | |
| jd__ | sileht: ah right I got something in keystone log if I try manuall | 15:50 |
| jd__ | y | 15:50 |
| jd__ | again a requirements issue | 15:50 |
| jd__ | lolololol | 15:50 |
| * jd__ is losing it | 15:50 | |
| *** g3ek has joined #openstack-telemetry | 15:54 | |
| *** diogogmt has joined #openstack-telemetry | 15:55 | |
| gordc | jd__: sileht: is python struct the best way to binary serialise a delimited string? | 15:56 |
| gordc | i don't need any fancy json/object stuff | 15:56 |
| jd__ | gordc: yes | 16:05 |
| gordc | jd__: kk. just testing out some serialisation techniques | 16:06 |
| jd__ | gordc: sure | 16:07 |
| gordc | if i change the format. i can make the serialisation a bit quicker (and support appends) by maybe 50%. but it's about 250% slower to deserialise because it needs to rebuild all the timestamps (which i don't store) | 16:09 |
| gordc | i'm just going to see if appends helps enough to ignore the slower deserialisation. | 16:09 |
| *** ljxiash has quit IRC | 16:11 | |
| *** ljxiash has joined #openstack-telemetry | 16:12 | |
| *** yprokule has quit IRC | 16:14 | |
| *** khushbu has quit IRC | 16:16 | |
| *** caisan has quit IRC | 16:21 | |
| *** Tamayo has joined #openstack-telemetry | 16:24 | |
| gordc | sileht: did you fix this issue yesterday? ContextualVersionConflict: (oslo.utils 3.8.0 (/usr/lib/python2.7/site-packages), Requirement.parse('oslo.utils>=3.11.0'), set(['oslo.middleware'])) | 16:33 |
| gordc | jd__: ^ | 16:33 |
| sileht | gordc, yes | 16:33 |
| sileht | gordc, that is a pip upgrade bug | 16:34 |
| gordc | sileht: oh. uh is there a workaround? | 16:34 |
| gordc | rm devstack? | 16:35 |
| sileht | gordc, in devstack plugin I have replaced pip install by pip install -U for gnocchi | 16:35 |
| gordc | oh ok. i'll try that. | 16:35 |
| sileht | gordc, https://review.openstack.org/#/c/306329/34/devstack/plugin.sh | 16:35 |
| gordc | lol i like the note. | 16:36 |
| *** pcaruana has quit IRC | 16:36 | |
| *** khushbu has joined #openstack-telemetry | 16:37 | |
| *** g3ek has quit IRC | 16:40 | |
| gordc | dang. i don't think it's gnocchi. i'm just going to update everything i guess. | 16:44 |
| *** rcernin has quit IRC | 16:45 | |
| *** g3ek has joined #openstack-telemetry | 16:46 | |
| jd__ | 2016-06-29 16:42:58.379 | WARNING: | 16:52 |
| jd__ | 2016-06-29 16:42:58.379 | Using lib/neutron-legacy is deprecated, and it will be removed in the future | 16:52 |
| jd__ | ok | 16:52 |
| jd__ | I never intended to use it | 16:52 |
| jd__ | why do you say that to me | 16:52 |
| jd__ | source openrc | 16:53 |
| jd__ | WARNING: setting legacy OS_TENANT_NAME to support cli tools. | 16:53 |
| jd__ | ok so what? I don't care | 16:53 |
| jd__ | this is so bogus | 16:53 |
| sileht | I love warning that can't be fixed | 16:54 |
| gordc | WARNING: chew your food before you swallow. | 16:57 |
| *** krotscheck is now known as krotscheck_vaca | 17:01 | |
| *** krotscheck_vaca is now known as krot_vaca_jul19 | 17:01 | |
| *** ljxiash has quit IRC | 17:06 | |
| *** shardy has quit IRC | 17:20 | |
| *** khushbu has quit IRC | 17:21 | |
| *** g3ek has quit IRC | 17:30 | |
| *** g3ek has joined #openstack-telemetry | 17:32 | |
| *** julim has joined #openstack-telemetry | 17:38 | |
| *** amoralej is now known as amoralej|off | 17:42 | |
| *** rcernin has joined #openstack-telemetry | 17:45 | |
| *** chlong has quit IRC | 17:52 | |
| *** pece has quit IRC | 17:54 | |
| *** permalac has quit IRC | 17:57 | |
| *** julim has quit IRC | 18:16 | |
| *** g3ek has quit IRC | 18:20 | |
| *** g3ek has joined #openstack-telemetry | 18:23 | |
| *** eglynn has quit IRC | 18:30 | |
| flwang1 | jd__: gordc: is there any document about ceilometer migration based on mongo? | 18:34 |
| gordc | flwang1: you mean between stable releases? | 18:37 |
| flwang1 | gordc: yes | 18:37 |
| gordc | not really. you don't really need to do anything. just run ceilometer-upgrade | 18:37 |
| gordc | i don't know if we've modified the mongodb recently. | 18:38 |
| flwang1 | gordc: the background story is we(zaqar) are going to claim we support cold migration, but we're asked something like this "The stores used by Zaqar support dynamic schemas, which makes schema changes easier and more dangerous. What's Zaqar's mechanism to migrate database schemas now?" | 18:39 |
| flwang1 | so i think ceilometer's mongoDB has the same 'issue' | 18:39 |
| flwang1 | it's a dynamic schema, so migration is easy and 'dangerous', do you guys have any document to make sure the process is safe? though personally, i don't think it's a problem | 18:40 |
| gordc | cold migration == turn off services, upgrade mongo/services, enable services? | 18:41 |
| gordc | i don't know what the diff between a migration in mongo and migration in sql would be. you read what you expect, you write what you expect. if you read something that doesn't exist, "boom" | 18:43 |
| *** julim has joined #openstack-telemetry | 18:47 | |
| stevelle | if the schema changes that do occur are backwards compatible or the operations degrade gracefully when data is in old schemas, than you should be "safe" | 18:48 |
| stevelle | I would expect that if mongo is a long-term supported storage you need unit tests for processing old schema formats throughout | 18:48 |
| stevelle | also, versioned objects becomes pretty easy in mongo | 18:49 |
| gordc | flwang1: ^ | 18:49 |
| gordc | flwang1: if this is a tc question i'm assuming they want the answer to be 'grenade testing' | 18:49 |
| flwang1 | yep, it's a tc question | 18:50 |
| flwang1 | but the tricky part is how can i prove our test for old schema is fully throughout :D | 18:52 |
| stevelle | ¯\_(ツ)_/¯ dynamic schemaless documents | 18:53 |
| gordc | usually the grenade test is write with v1, upgrade, read with v2 | 18:53 |
| stevelle | the only hangup for ^ is ensuring that you have a representative data set in the version N-1 dataset | 18:53 |
| *** thumpba has quit IRC | 18:54 | |
| gordc | i don't think mongo migrations are any more unique than a sql migration. the only difference i would imagine is that mongo will hide some of the breakage. | 18:54 |
| flwang1 | gordc: so is there any migration testing in ceilometer against mongodb? | 18:55 |
| *** sheel has quit IRC | 18:55 | |
| gordc | stevelle: i wonder how exhaustive that testing is. i'm pretty sure ours is relatively trivial (since we spend most time on other backends) | 18:56 |
| stevelle | gordc: agreed | 18:56 |
| flwang1 | i just checked ceilometer's source code | 18:57 |
| gordc | flwang1: our grenade test is against sql | 18:57 |
| flwang1 | i didn't see any test about mongodb | 18:57 |
| flwang1 | migration | 18:57 |
| *** julim has quit IRC | 18:58 | |
| flwang1 | same thing for hbase | 18:58 |
| gordc | flwang1: probably not. we haven't changed our model in mongodb ever. aside from adding/removing indices | 18:58 |
| flwang1 | really? | 18:59 |
| gordc | hbase i'm not sure is validated anywhere. someting the mirantis folks cooked up... i'm not sure how you'd test that in our CI | 18:59 |
| flwang1 | i think any change against sql means you may need to do the similar change for mongo | 18:59 |
| gordc | flwang1: not that i know of... there's a reason the existing mongodb driver is not really scalable (unless you have a ton of machines) | 18:59 |
| flwang1 | unfortunately, we(catalyst IT) are using mongoDB | 19:00 |
| flwang1 | and yep, it's bring some troubles now and we're going to move to gnocchi | 19:00 |
| gordc | flwang1: changing sql would have no impact on whether mongo or any other driver changed | 19:00 |
| gordc | it's only if you change the incoming datamodel | 19:00 |
| flwang1 | back to my above question, why do you think ceilometer's mongo schema never been changed/ | 19:01 |
| gordc | if you look at our data models you have sample and events. they have been static for years. | 19:01 |
| gordc | you can store that in any way you want. but that's totally dependent on your backend. | 19:01 |
| *** therve has quit IRC | 19:02 | |
| flwang1 | i mean something like this https://github.com/openstack/ceilometer/blob/master/ceilometer/storage/sqlalchemy/migrate_repo/versions/042_add_raw_column.py | 19:02 |
| gordc | our sql changes were just changes to store the same data in a different way. we could do the same for mongodb but it'd be a ton of work and most of the devs actually contributing work on other drivers | 19:02 |
| gordc | you mean for events? that field is completely new and optional. not sure there's any migration needed three | 19:03 |
| gordc | from mongo pov it's nothing. if you're v1, you will never check for it. if you're v2, you will check for it, and if it's there use it, it not, carry on. | 19:04 |
| flwang1 | ok, fair enough :) | 19:06 |
| flwang1 | so i think it's still hard to explain the data migration things for dynamic schema | 19:07 |
| *** thumpba has joined #openstack-telemetry | 19:07 | |
| *** therve has joined #openstack-telemetry | 19:07 | |
| flwang1 | since you can't just say 'we didn't change anything(a lot)' to tc, I don't think they will buy in :) | 19:07 |
| *** thumpba has quit IRC | 19:07 | |
| *** thumpba has joined #openstack-telemetry | 19:08 | |
| gordc | flwang1: the basic idea of rolling upgrades is 'never remove, never modify, only add' so if you do that, your code will always work. | 19:08 |
| gordc | depending on how much you change, it will grow a lot though. | 19:08 |
| flwang1 | gordc: we didn't change a lot since the init schema is very stable, well-defined | 19:09 |
| flwang1 | but i need a strong justification for tc :D | 19:09 |
| stevelle | flwang1: is there any documentation for the mongo data schemas? | 19:09 |
| flwang1 | stevelle: for zaqar? | 19:09 |
| stevelle | flwang1: yes | 19:09 |
| gordc | flwang1: i'm pretty sure it's just 'grenade'. usually it's just a bunch of checkboxes they have beside them. | 19:10 |
| *** g3ek has quit IRC | 19:10 | |
| *** thumpba_ has joined #openstack-telemetry | 19:11 | |
| flwang1 | stevelle: we have defined the schema like this https://github.com/openstack/zaqar/blob/master/zaqar/storage/mongodb/queues.py#L70 | 19:11 |
| *** thumpba has quit IRC | 19:11 | |
| flwang1 | but we're not using a rst doc to record it | 19:11 |
| flwang1 | maybe we should do that | 19:11 |
| *** liamji has quit IRC | 19:12 | |
| stevelle | flwang1: best would be to have a formal json-schema format or something | 19:12 |
| flwang1 | stevelle: that's a good idea | 19:12 |
| stevelle | but I think that's for social value rather than technical | 19:12 |
| stevelle | if you were to design a "grenade" test what would it look like? Insert data in version N-1 format, read data in version N-1, upgrade, insert more data in version N format, then read data from both versions? | 19:13 |
| stevelle | maybe you can use your existing integration testing to help with this? | 19:14 |
| *** thumpba has joined #openstack-telemetry | 19:14 | |
| flwang1 | stevelle: do you think a unit test is enough? | 19:15 |
| *** thumpba_ has quit IRC | 19:15 | |
| stevelle | flwang1: I would expect you will be asked the same question you asked before: how can i prove our test for old schema is fully through | 19:17 |
| flwang1 | haha yep | 19:17 |
| stevelle | unit testing makes that tougher to answer | 19:17 |
| flwang1 | that's why i'm here to get feedback from you guys | 19:17 |
| flwang1 | stevelle: ah, that makes sense | 19:17 |
| flwang1 | stevelle: we don't have integration test but functional test | 19:18 |
| stevelle | flwang1: possibly you can use the list of mongo tables before and after upgrade to improve confidence? | 19:18 |
| stevelle | flwang1: you might be able to use gabbi fairly easily to add some integration tests for this | 19:18 |
| *** thumpba has quit IRC | 19:19 | |
| *** thumpba has joined #openstack-telemetry | 19:19 | |
| flwang1 | stevelle: https://github.com/cdent/gabbi ? | 19:19 |
| *** g3ek has joined #openstack-telemetry | 19:19 | |
| stevelle | flwang1: correct | 19:19 |
| gordc | cdent will be very happy for the plug ;) | 19:19 |
| *** thumpba has quit IRC | 19:20 | |
| flwang1 | stevelle: cool | 19:20 |
| *** thumpba has joined #openstack-telemetry | 19:20 | |
| flwang1 | stevelle: now we're using 'from falcon import testing' | 19:20 |
| flwang1 | since we're using falcon | 19:20 |
| stevelle | flwang1: I use the same in most of my personal projects :o) | 19:21 |
| stevelle | that isn't integration testing however | 19:21 |
| gordc | stevelle: here's wip re: gnocchi+ceph tuning https://docs.google.com/presentation/d/1eHZDcgN9QJk8uEo4V-tGhthpq7aOp_m1kncBFWLwwQI/edit?usp=sharing | 19:22 |
| flwang1 | gordc: oh btw, which backend of gnocchi is the most stable one? | 19:22 |
| *** thumpba_ has joined #openstack-telemetry | 19:23 | |
| flwang1 | stevelle: so you mean falcon.testing is not integration, but gabbi is? | 19:23 |
| gordc | flwang1: i believe people are using swift but i think we're mostly working on ceph driver (small group of devs) | 19:23 |
| *** thumpba_ has quit IRC | 19:24 | |
| *** thumpba_ has joined #openstack-telemetry | 19:24 | |
| stevelle | flwang1: correct. falcon.testing mocks a request but gabbi comes in from outside | 19:24 |
| flwang1 | gordc: a silly question, can the swift backend make sure good performance? | 19:24 |
| flwang1 | stevelle: good to know | 19:24 |
| *** thumpba_ has quit IRC | 19:25 | |
| *** thumpba_ has joined #openstack-telemetry | 19:25 | |
| *** thumpba has quit IRC | 19:25 | |
| *** thumpba_ has quit IRC | 19:25 | |
| stevelle | also, substitute "databases" for "tables" in my suggestion above for improving confidence | 19:27 |
| stevelle | juggling contexts and got the wrong thing in my brain :D | 19:27 |
| gordc | flwang1: i'm playing with a new format. it'll (hopefully) shrink number of iops but only for ceph|file. swift it'll only shrink size since swift doesn't support append. | 19:28 |
| gordc | not that we know. | 19:28 |
| gordc | flwang1: i'm never played with swift driver recently though. | 19:29 |
| flwang1 | gordc: we're planning to move to gnocchi but before that we need to know which is the right one we should to go to avoid black holes | 19:30 |
| gordc | gnocchi requires quite a bit of iops to scale. so i'd probably just choose whatever object store you currently have deployed. if you don't have one, ceph then? it's the one i'm testing against :P | 19:32 |
| flwang1 | we have both ceph and swift(on the way) | 19:34 |
| gordc | ... oh? i guess you don't like ceph? we're working on both drivers (but mostly ceph). that's not to say one is more performant. | 19:36 |
| flwang1 | i like ceph :D | 19:36 |
| flwang1 | gordc: btw, as for the api deprecation, have you guys totally removed the v1 code? | 19:36 |
| gordc | v1 code for what? | 19:37 |
| flwang1 | v1 code from code repository | 19:37 |
| gordc | like ceilometer api v1? | 19:38 |
| flwang1 | oh, yes, sorry | 19:38 |
| gordc | oh. err.. yeah i imagine that's gone. i'm pretty sure it's been v2 since before we were tc approved | 19:38 |
| gordc | i don't know what the v1 api is tbhj | 19:39 |
| gordc | s/tbhj/tbh/ | 19:39 |
| flwang1 | hah, ok, cool | 19:39 |
| stevelle | There isn't any v1 api code, that much I know | 19:43 |
| *** catintheroof has quit IRC | 19:43 | |
| *** Tamayo has quit IRC | 19:47 | |
| *** Guest20454 is now known as mgagne | 19:58 | |
| *** mgagne has joined #openstack-telemetry | 19:58 | |
| *** jwcroppe has joined #openstack-telemetry | 20:11 | |
| *** julim has joined #openstack-telemetry | 20:19 | |
| *** marcin1234 has quit IRC | 20:23 | |
| *** flwang1 has quit IRC | 20:36 | |
| *** g3ek has quit IRC | 20:55 | |
| *** g3ek has joined #openstack-telemetry | 21:03 | |
| openstackgerrit | gordon chung proposed openstack/gnocchi: avoid serialising to dict https://review.openstack.org/335670 | 21:14 |
| *** thorst has quit IRC | 21:21 | |
| *** thorst has joined #openstack-telemetry | 21:46 | |
| *** gordc has quit IRC | 21:49 | |
| *** _nadya_ has quit IRC | 21:49 | |
| *** g3ek has quit IRC | 21:49 | |
| *** thorst has quit IRC | 21:50 | |
| *** _nadya_ has joined #openstack-telemetry | 21:50 | |
| *** _nadya_ has quit IRC | 21:52 | |
| *** diogogmt has quit IRC | 21:53 | |
| *** rcernin has quit IRC | 21:54 | |
| *** g3ek has joined #openstack-telemetry | 21:54 | |
| *** thorst has joined #openstack-telemetry | 22:07 | |
| *** ametts has quit IRC | 22:16 | |
| *** chlong has joined #openstack-telemetry | 22:17 | |
| *** diogogmt has joined #openstack-telemetry | 22:56 | |
| *** thorst has quit IRC | 22:58 | |
| *** thorst has joined #openstack-telemetry | 22:59 | |
| *** thorst has quit IRC | 23:07 | |
| *** KrishR has quit IRC | 23:07 | |
| *** peristeri has quit IRC | 23:23 | |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/ceilometer: Use tempest.lib's v2 images_client https://review.openstack.org/329259 | 23:30 |
| *** g3ek has quit IRC | 23:33 | |
| *** g3ek has joined #openstack-telemetry | 23:38 | |
| *** yassine__ has quit IRC | 23:40 | |
| *** khushbu has joined #openstack-telemetry | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!