| *** boris-42 has quit IRC | 01:23 | |
| *** vivekd has joined #magnetodb | 02:46 | |
| *** jeromatron has joined #magnetodb | 04:06 | |
| *** rushiagr_away is now known as rushiagr | 05:12 | |
| *** jeromatron has quit IRC | 05:15 | |
| *** ajayaa has joined #magnetodb | 05:28 | |
| *** k4n0 has joined #magnetodb | 06:34 | |
| *** romainh has joined #magnetodb | 08:27 | |
| ajayaa | isviridov, https://review.openstack.org/#/c/126335/3/specs/kilo/magnetodb-notifications.rst | 08:51 |
|---|---|---|
| ajayaa | isviridov_away ^^ | 08:51 |
| *** vivekd has quit IRC | 08:54 | |
| *** isviridov_away is now known as isviridov | 08:59 | |
| isviridov | Hello ajayaa, cool! | 09:00 |
| isviridov | Let me take a look | 09:00 |
| isviridov | ajayaa : what is resource id in samples? | 09:05 |
| ajayaa | That is a unique field per resource. In our case it would represent a unique table in mdb. | 09:06 |
| ajayaa | As of now I don't think UUIDs are being sent in notifications. | 09:06 |
| ajayaa | But we will send them I guess. | 09:06 |
| ajayaa | isviridov ^^ | 09:07 |
| isviridov | I see now. How does multitenancy work here? | 09:15 |
| isviridov | It seems like we are using samples where events should be used. | 09:15 |
| isviridov | I mean http://docs.openstack.org/developer/ceilometer/events.html#events-vs-samples | 09:16 |
| isviridov | With samples we have some changing data for specific resource. F.e. cpu usage for instanse, of size for table. | 09:17 |
| isviridov | Probably we can handle table count per tenant as sample, where tenant is resource. | 09:18 |
| isviridov | What do you think? | 09:19 |
| ajayaa | isviridov, I was using cinder as a reference. When a volume is created, a notification is generated from cinder which is caught by ceilometer. Then a sample is created from this notification. | 09:22 |
| ajayaa | isviridov, If we just keep count of tables, we won't be able to store number of indices per table. | 09:23 |
| ajayaa | isviridov, I just came across events. I am not sure whether an event or sample is fit for our usecase. | 09:25 |
| *** boris-42 has joined #magnetodb | 09:46 | |
| isviridov | ajayaa : ok, I see. Not sure that cinder is the best example, because some things was implemented to provide bacis monitoring of availability | 09:51 |
| isviridov | I mean 'volume' meter | 09:51 |
| isviridov | I think we have to follow heat or sahara, as far as in both cases they are publishing data about abstrac entities like stack or cluster. | 09:52 |
| isviridov | Let me comment in your review. | 09:52 |
| ajayaa | isviridov, It is easy to get count of tables per tenant from samples. You can pass a query parameter to ceilometer client. | 09:52 |
| isviridov | ajayaa : yeap, I've found it. Thank you | 09:53 |
| isviridov | Why it is -1 by jenkins? Any ideas? | 09:53 |
| ajayaa | isviridov, nope. I will take a look. | 09:57 |
| *** openstackgerrit has joined #magnetodb | 10:03 | |
| isviridov | ajayaa : added several comments, actually the idea is to do it now for only two types events table.create and table.delete in order to count the tables and track activity. | 10:11 |
| isviridov | I also suggest removing index count, as ambigous and keep it simple on the first stage. | 10:12 |
| ajayaa | isviridov, I think the two type of events table.create and table.delete could be stored under the same meter. | 10:12 |
| isviridov | ajayaa : what the meter? | 10:12 |
| isviridov | table? | 10:12 |
| ajayaa | yes. | 10:12 |
| ajayaa | That way when you see two samples per resource_id per meter, you can infer that the table is deleted. | 10:13 |
| ajayaa | I need to check whether we can express this in ceilometer query so that we get count of tables per tenant. | 10:15 |
| isviridov | I'm looking at it as monitoring system also, not only backend for billing. I personally would like to track table connected events https://github.com/stackforge/magnetodb/blob/master/magnetodb/notifier/__init__.py#L40-L45 | 10:17 |
| isviridov | But they looks like events, not samles :) It is needed to clarify how to put to ceilo as well. | 10:17 |
| isviridov | ajayaa : yes, please do so. Thanks | 10:18 |
| ajayaa | isviridov, It is convenient to add two type of meter. table.create and table.delete. That way we would count table.create samples and subtract count of table.delete samples to get number of tables per tenant. | 10:20 |
| ajayaa | You were right. ;) | 10:21 |
| isviridov | Nice | 10:21 |
| * ajayaa reading more about events in ceilometer. | 10:25 | |
| *** rushiagr is now known as rushiagr_away | 10:52 | |
| openstackgerrit | Illia Khudoshyn proposed a change to stackforge/magnetodb: Raise ValidationError if attribute type is missing https://review.openstack.org/132219 | 11:43 |
| *** boris-42 has quit IRC | 12:14 | |
| openstackgerrit | Illia Khudoshyn proposed a change to stackforge/magnetodb: Raise ValidationError if attribute type is missing https://review.openstack.org/132219 | 12:21 |
| *** rushiagr_away is now known as rushiagr | 12:31 | |
| *** isviridov is now known as isviridov_away | 12:34 | |
| *** ajayaa has quit IRC | 12:34 | |
| *** miqui has joined #magnetodb | 12:48 | |
| ikhudoshyn | https://review.openstack.org/#/c/132219/ pls review | 12:55 |
| *** charlesw has joined #magnetodb | 13:28 | |
| openstackgerrit | Oleksandr Minakov proposed a change to stackforge/magnetodb: Adds monitoring API https://review.openstack.org/122330 | 13:34 |
| openstackgerrit | Oleksandr Minakov proposed a change to stackforge/magnetodb: Adds monitoring API https://review.openstack.org/122330 | 13:35 |
| *** boris-42 has joined #magnetodb | 13:38 | |
| openstackgerrit | Nuno Santos proposed a change to stackforge/magnetodb: Record and retrieve table creation timestamp https://review.openstack.org/131546 | 13:41 |
| *** denis_makogon_ has joined #magnetodb | 13:57 | |
| openstackgerrit | Oleksandr Minakov proposed a change to stackforge/magnetodb: Adds monitoring API https://review.openstack.org/122330 | 13:59 |
| *** jeromatron has joined #magnetodb | 13:59 | |
| openstackgerrit | Oleksandr Minakov proposed a change to stackforge/magnetodb: Adds monitoring API https://review.openstack.org/122330 | 14:06 |
| *** jeromatron has quit IRC | 14:21 | |
| openstackgerrit | Illia Khudoshyn proposed a change to stackforge/magnetodb: Raise ValidationError if attribute type is missing https://review.openstack.org/132219 | 14:23 |
| ikhudoshyn | charlesw: around? | 14:25 |
| *** jeromatron has joined #magnetodb | 14:27 | |
| openstackgerrit | Illia Khudoshyn proposed a change to stackforge/magnetodb: Raise ValidationError if attribute type is missing https://review.openstack.org/132219 | 14:28 |
| *** jeromatron has quit IRC | 14:30 | |
| *** jeromatron has joined #magnetodb | 14:32 | |
| openstackgerrit | Oleksandr Minakov proposed a change to stackforge/magnetodb: Adds Cassandra implemetation of monitoring API https://review.openstack.org/132267 | 14:33 |
| *** romainh has quit IRC | 14:38 | |
| *** jeromatron has quit IRC | 14:39 | |
| *** jeromatron has joined #magnetodb | 14:46 | |
| *** romainh has joined #magnetodb | 14:51 | |
| *** boris-42 has quit IRC | 14:52 | |
| *** jeromatron has quit IRC | 14:53 | |
| *** jeromatron has joined #magnetodb | 14:54 | |
| *** boris-42 has joined #magnetodb | 14:54 | |
| *** boris-42 has quit IRC | 15:11 | |
| *** k4n0 has quit IRC | 15:27 | |
| *** jeromatron has quit IRC | 15:29 | |
| openstackgerrit | Nuno Santos proposed a change to stackforge/magnetodb: Record and retrieve table creation timestamp https://review.openstack.org/131546 | 15:40 |
| *** jeromatron has joined #magnetodb | 16:16 | |
| *** denis_makogon_ has quit IRC | 17:09 | |
| *** jeromatron has quit IRC | 17:58 | |
| *** romainh has quit IRC | 18:14 | |
| *** romainh1 has joined #magnetodb | 18:14 | |
| *** rushiagr is now known as rushiagr_away | 18:32 | |
| openstackgerrit | A change was merged to stackforge/magnetodb: Raise ValidationError if attribute type is missing https://review.openstack.org/132219 | 19:16 |
| *** jeromatron has joined #magnetodb | 19:34 | |
| *** romainh1 has left #magnetodb | 20:13 | |
| *** jeromatron has quit IRC | 20:25 | |
| *** jeromatron has joined #magnetodb | 20:37 | |
| *** denis_makogon_ has joined #magnetodb | 21:18 | |
| *** nunosantos has quit IRC | 21:21 | |
| *** charlesw has quit IRC | 21:31 | |
| *** boris-42 has joined #magnetodb | 21:55 | |
| openstackgerrit | A change was merged to stackforge/magnetodb: Record and retrieve table creation timestamp https://review.openstack.org/131546 | 22:02 |
| *** jeromatron has quit IRC | 22:06 | |
| *** jeromatron has joined #magnetodb | 22:07 | |
| *** denis_makogon_ has quit IRC | 22:08 | |
| *** jeromatron has quit IRC | 22:13 | |
| *** charlesw has joined #magnetodb | 23:40 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!