| *** amcrn has quit IRC | 00:00 | |
| *** tkatarki has quit IRC | 00:00 | |
| *** tkatarki has joined #openstack-trove | 00:02 | |
| *** ramashri has quit IRC | 00:03 | |
| *** georgelorch has quit IRC | 00:08 | |
| *** ramashri has joined #openstack-trove | 00:10 | |
| *** ViswaV has quit IRC | 00:11 | |
| *** tkatarki has quit IRC | 00:14 | |
| *** tkatarki has joined #openstack-trove | 00:14 | |
| *** annashen has joined #openstack-trove | 00:25 | |
| *** georgelorch has joined #openstack-trove | 00:37 | |
| *** vkmc has quit IRC | 00:41 | |
| *** Barker has joined #openstack-trove | 00:49 | |
| *** grapex has joined #openstack-trove | 00:50 | |
| *** Barker has quit IRC | 00:52 | |
| *** annashen has quit IRC | 00:53 | |
| *** grapex has quit IRC | 00:54 | |
| *** jasonb365 has joined #openstack-trove | 00:55 | |
| *** ViswaV has joined #openstack-trove | 00:57 | |
| *** ViswaV has quit IRC | 00:59 | |
| *** todd_dsm has joined #openstack-trove | 01:00 | |
| *** ViswaV has joined #openstack-trove | 01:00 | |
| *** tkatarki has quit IRC | 01:05 | |
| *** ramashri has quit IRC | 01:09 | |
| *** Riddhi has joined #openstack-trove | 01:10 | |
| *** ramashri has joined #openstack-trove | 01:10 | |
| *** RicoLin has joined #openstack-trove | 01:23 | |
| *** eghobo has quit IRC | 01:25 | |
| *** eghobo has joined #openstack-trove | 01:26 | |
| *** topshare has joined #openstack-trove | 01:29 | |
| *** shakamunyi has quit IRC | 01:29 | |
| *** annashen has joined #openstack-trove | 01:30 | |
| *** annashen has quit IRC | 01:32 | |
| *** topshare has quit IRC | 01:33 | |
| *** jasonb365 has quit IRC | 01:34 | |
| *** shakamunyi has joined #openstack-trove | 01:36 | |
| *** ViswaV has quit IRC | 01:40 | |
| *** topshare has joined #openstack-trove | 01:50 | |
| *** todd_dsm has quit IRC | 01:52 | |
| *** eguz has joined #openstack-trove | 01:54 | |
| *** eguz has quit IRC | 01:54 | |
| *** todd_dsm has joined #openstack-trove | 01:55 | |
| *** Riddhi has quit IRC | 01:56 | |
| *** nosnos has joined #openstack-trove | 01:56 | |
| *** eghobo has quit IRC | 01:58 | |
| *** topshare has quit IRC | 01:59 | |
| *** topshare has joined #openstack-trove | 02:03 | |
| *** topshare has quit IRC | 02:03 | |
| *** todd_dsm has quit IRC | 02:15 | |
| *** ramashri has quit IRC | 02:18 | |
| *** todd_dsm has joined #openstack-trove | 02:19 | |
| *** todd_dsm has quit IRC | 02:20 | |
| *** jasonb365 has joined #openstack-trove | 02:25 | |
| *** edmund has joined #openstack-trove | 02:26 | |
| *** haomaiwang has joined #openstack-trove | 02:27 | |
| *** haomaiwang has quit IRC | 02:34 | |
| *** haomaiwang has joined #openstack-trove | 02:35 | |
| *** haomai___ has joined #openstack-trove | 02:43 | |
| *** harlowja is now known as harlowja_away | 02:44 | |
| *** haomaiwang has quit IRC | 02:46 | |
| *** nosnos has quit IRC | 03:23 | |
| *** Riddhi has joined #openstack-trove | 03:24 | |
| *** nosnos has joined #openstack-trove | 03:24 | |
| *** nosnos has quit IRC | 03:28 | |
| *** harlowja_away has quit IRC | 03:29 | |
| *** annashen has joined #openstack-trove | 03:39 | |
| *** grapex has joined #openstack-trove | 03:51 | |
| *** nosnos has joined #openstack-trove | 03:53 | |
| *** grapex has quit IRC | 03:56 | |
| *** eghobo has joined #openstack-trove | 04:04 | |
| *** eghobo has quit IRC | 04:12 | |
| *** georgelorch has quit IRC | 04:14 | |
| *** ramishra has joined #openstack-trove | 04:17 | |
| *** jasonb365 has quit IRC | 04:19 | |
| *** Riddhi has quit IRC | 04:26 | |
| *** todd_dsm has joined #openstack-trove | 04:26 | |
| *** georgelorch has joined #openstack-trove | 04:29 | |
| *** eghobo has joined #openstack-trove | 04:42 | |
| *** ramishra has quit IRC | 04:46 | |
| *** ramishra has joined #openstack-trove | 04:46 | |
| *** ramishra has quit IRC | 04:48 | |
| *** ramishra has joined #openstack-trove | 04:49 | |
| zhiyan | amrith: he! still around? | 04:50 |
|---|---|---|
| *** ajayaa has joined #openstack-trove | 05:02 | |
| *** juantwo has quit IRC | 05:07 | |
| *** ViswaV has joined #openstack-trove | 05:32 | |
| *** ViswaV_ has joined #openstack-trove | 05:33 | |
| *** ViswaV has quit IRC | 05:36 | |
| *** todd_dsm has quit IRC | 05:51 | |
| *** grapex has joined #openstack-trove | 05:52 | |
| *** haomai___ has quit IRC | 05:53 | |
| *** grapex has quit IRC | 05:57 | |
| *** k4n0 has joined #openstack-trove | 05:59 | |
| *** vigneshvar has joined #openstack-trove | 06:01 | |
| *** ajayaa has quit IRC | 06:05 | |
| *** ramishra has quit IRC | 06:11 | |
| *** ramishra has joined #openstack-trove | 06:12 | |
| *** flaper87|afk is now known as flaper87 | 06:15 | |
| *** Longgeek has joined #openstack-trove | 06:20 | |
| *** haomaiwang has joined #openstack-trove | 06:20 | |
| *** julienvey has joined #openstack-trove | 06:24 | |
| *** ajayaa has joined #openstack-trove | 06:26 | |
| *** ramishra has quit IRC | 06:41 | |
| *** dkehn_ has quit IRC | 06:43 | |
| *** dkehnx has quit IRC | 06:43 | |
| *** eghobo has quit IRC | 06:44 | |
| *** dkehn has joined #openstack-trove | 06:44 | |
| *** dkehn_ has joined #openstack-trove | 06:45 | |
| *** grapex has joined #openstack-trove | 06:53 | |
| *** grapex has quit IRC | 06:57 | |
| *** nosnos has quit IRC | 07:04 | |
| *** nosnos_ has joined #openstack-trove | 07:04 | |
| *** freyes has quit IRC | 07:09 | |
| *** freyes has joined #openstack-trove | 07:10 | |
| *** sgotliv has joined #openstack-trove | 07:14 | |
| *** dkehn_ has quit IRC | 07:31 | |
| *** dkehn has quit IRC | 07:32 | |
| *** ViswaV_ has quit IRC | 07:35 | |
| *** ViswaV has joined #openstack-trove | 07:45 | |
| *** ViswaV has quit IRC | 07:54 | |
| *** shakayumi has joined #openstack-trove | 08:01 | |
| *** mikehn_ has joined #openstack-trove | 08:03 | |
| *** blipp_ has joined #openstack-trove | 08:04 | |
| *** imsplitb1t has joined #openstack-trove | 08:05 | |
| *** dmakogon_ has joined #openstack-trove | 08:06 | |
| *** annashen has quit IRC | 08:08 | |
| *** SergeyLukjanov2 has joined #openstack-trove | 08:09 | |
| *** shakamunyi has quit IRC | 08:10 | |
| *** mikehn has quit IRC | 08:10 | |
| *** RicoLin has quit IRC | 08:10 | |
| *** blipp has quit IRC | 08:10 | |
| *** denis_makogon has quit IRC | 08:10 | |
| *** imsplitbit has quit IRC | 08:10 | |
| *** SergeyLukjanov has quit IRC | 08:10 | |
| *** SergeyLukjanov2 is now known as SergeyLukjanov | 08:10 | |
| *** iartarisi has joined #openstack-trove | 08:14 | |
| *** haomaiwang has quit IRC | 08:14 | |
| *** haomaiwang has joined #openstack-trove | 08:15 | |
| *** RicoLin has joined #openstack-trove | 08:15 | |
| *** RicoLin has quit IRC | 08:15 | |
| *** RicoLin has joined #openstack-trove | 08:16 | |
| *** boblebauce has joined #openstack-trove | 08:17 | |
| *** flaper87 is now known as flaper87|afk | 08:17 | |
| *** Longgeek has quit IRC | 08:20 | |
| *** Longgeek has joined #openstack-trove | 08:21 | |
| *** haomaiw__ has joined #openstack-trove | 08:24 | |
| *** Longgeek has quit IRC | 08:25 | |
| *** haomaiwang has quit IRC | 08:27 | |
| *** haomaiw__ has quit IRC | 08:29 | |
| *** haomaiwang has joined #openstack-trove | 08:35 | |
| *** shakayumi has quit IRC | 08:40 | |
| *** dkehn has joined #openstack-trove | 08:42 | |
| *** Longgeek has joined #openstack-trove | 08:43 | |
| *** nosnos_ has quit IRC | 08:45 | |
| *** dkehn_ has joined #openstack-trove | 08:45 | |
| *** nosnos has joined #openstack-trove | 08:46 | |
| *** nosnos has quit IRC | 08:50 | |
| *** isviridov|away is now known as isviridov | 08:51 | |
| *** grapex has joined #openstack-trove | 08:54 | |
| *** grapex has quit IRC | 08:59 | |
| *** shakayumi has joined #openstack-trove | 09:06 | |
| *** shakayumi has quit IRC | 09:10 | |
| *** nosnos has joined #openstack-trove | 09:11 | |
| *** blipp_ has quit IRC | 09:12 | |
| *** haomaiwang has quit IRC | 09:25 | |
| *** blipp has joined #openstack-trove | 09:26 | |
| *** blipp has quit IRC | 09:32 | |
| *** blipp has joined #openstack-trove | 09:32 | |
| *** haomaiwang has joined #openstack-trove | 09:37 | |
| *** Longgeek has quit IRC | 09:50 | |
| *** sgotliv has quit IRC | 09:52 | |
| *** rushiagr_away is now known as rushiagr | 09:58 | |
| *** shakayumi has joined #openstack-trove | 10:06 | |
| amrith | zhiyan, yes ... | 10:09 |
| *** shakayumi has quit IRC | 10:11 | |
| *** Longgeek has joined #openstack-trove | 10:16 | |
| *** sgotliv has joined #openstack-trove | 10:27 | |
| *** rushiagr is now known as rushiagr_away | 10:30 | |
| *** Longgeek has quit IRC | 10:32 | |
| *** sgotliv has quit IRC | 10:40 | |
| *** flaper87|afk is now known as flaper87 | 10:47 | |
| *** tkatarki has joined #openstack-trove | 10:48 | |
| *** sgotliv has joined #openstack-trove | 10:54 | |
| *** Longgeek has joined #openstack-trove | 11:00 | |
| *** haomaiwang has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-trove | 11:01 | |
| *** tkatarki has quit IRC | 11:02 | |
| *** haomaiwang has quit IRC | 11:05 | |
| *** dmakogon_ is now known as denis_makogon | 11:05 | |
| *** haomaiwang has joined #openstack-trove | 11:05 | |
| denis_makogon | iccha, glucas i filed another one bug for snapshotting, see https://bugs.launchpad.net/trove/+bug/1362062 | 11:06 |
| *** haomaiwa_ has joined #openstack-trove | 11:06 | |
| *** shakayumi has joined #openstack-trove | 11:07 | |
| *** haomaiwang has quit IRC | 11:10 | |
| *** shakayumi has quit IRC | 11:11 | |
| *** flaper87 is now known as flaper87|afk | 11:14 | |
| *** nosnos has quit IRC | 11:20 | |
| *** tkatarki has joined #openstack-trove | 11:20 | |
| *** nosnos has joined #openstack-trove | 11:20 | |
| *** flaper87|afk is now known as flaper87 | 11:23 | |
| *** IanGovett has joined #openstack-trove | 11:24 | |
| *** tkatarki has quit IRC | 11:24 | |
| *** nosnos has quit IRC | 11:25 | |
| *** tkatarki has joined #openstack-trove | 11:32 | |
| *** ramishra has joined #openstack-trove | 11:33 | |
| *** tkatarki has quit IRC | 11:42 | |
| *** flaper87 is now known as flaper87|afk | 11:42 | |
| *** flaper87|afk is now known as flaper87 | 11:44 | |
| *** ramishra has quit IRC | 11:53 | |
| *** ramishra has joined #openstack-trove | 11:54 | |
| *** grapex has joined #openstack-trove | 11:55 | |
| *** ramishra has quit IRC | 11:58 | |
| *** grapex has quit IRC | 12:00 | |
| *** achampion has quit IRC | 12:03 | |
| *** achampion has joined #openstack-trove | 12:03 | |
| *** tkatarki has joined #openstack-trove | 12:03 | |
| *** shakayumi has joined #openstack-trove | 12:08 | |
| *** juantwo has joined #openstack-trove | 12:08 | |
| *** ajayaa has quit IRC | 12:08 | |
| *** vkmc has joined #openstack-trove | 12:09 | |
| *** juantwo has quit IRC | 12:10 | |
| *** juantwo has joined #openstack-trove | 12:10 | |
| *** shakayumi has quit IRC | 12:12 | |
| *** flaper87 is now known as flaper87|afk | 12:14 | |
| *** tkatarki has quit IRC | 12:17 | |
| *** tkatarki has joined #openstack-trove | 12:18 | |
| *** flaper87|afk is now known as flaper87 | 12:18 | |
| *** sgotliv has quit IRC | 12:19 | |
| *** sgotliv has joined #openstack-trove | 12:20 | |
| *** Isotopp_ is now known as Isotopp | 12:23 | |
| *** ajayaa has joined #openstack-trove | 12:29 | |
| *** tkatarki has quit IRC | 12:30 | |
| *** tkatarki has joined #openstack-trove | 12:31 | |
| *** tkatarki has quit IRC | 12:32 | |
| *** PierreRambaud has quit IRC | 12:36 | |
| *** GoT has joined #openstack-trove | 12:36 | |
| vigneshvar | amrith: hi. | 12:38 |
| *** todd_dsm has joined #openstack-trove | 12:39 | |
| vigneshvar | amrith: i just need few minutes of your time when you are free . To discuss on https://review.openstack.org/#/c/112532/ | 12:43 |
| vigneshvar | amrith: will try to discuss with nick when he is available | 12:43 |
| *** grapex has joined #openstack-trove | 12:56 | |
| *** topshare has joined #openstack-trove | 12:57 | |
| *** grapex has quit IRC | 13:01 | |
| *** achampion has quit IRC | 13:01 | |
| *** jasondotstar has joined #openstack-trove | 13:02 | |
| *** shakayumi has joined #openstack-trove | 13:08 | |
| zhiyan | amrith: hey! | 13:11 |
| zhiyan | amrith: when you ok could pls help review osprofiler trove integration change ? =) | 13:11 |
| zhiyan | amrith: https://review.openstack.org/#/c/116653/ ^^ | 13:12 |
| *** ajayaa has quit IRC | 13:13 | |
| *** shakayumi has quit IRC | 13:13 | |
| *** Barker has joined #openstack-trove | 13:26 | |
| *** GoT is now known as PierreRambaud | 13:26 | |
| *** ajayaa has joined #openstack-trove | 13:32 | |
| *** pdmars has joined #openstack-trove | 13:38 | |
| amrith | vigneshvar, ping ... | 13:38 |
| amrith | zhiyan, ping ... | 13:38 |
| zhiyan | amrith: hi =) | 13:39 |
| *** shakayumi has joined #openstack-trove | 13:40 | |
| amrith | zhiyan, I will look at the patch. I'm kind-of unsure about it, first what value it brings, second whether there is sufficient information for me to try it out and actually see the thing in action. | 13:40 |
| amrith | zhiyan, I will definitely try and look at the patch today. I'm on the hook for some changes in the next day or so, as soon as I can get those submitted I promise to look at this and other reviews. | 13:40 |
| zhiyan | amrith: sure, pls try it. then i believe you can see the benefits ^ | 13:41 |
| *** vigneshvar has quit IRC | 13:41 | |
| zhiyan | amrith: and thanks | 13:41 |
| *** shakayumi is now known as shakamunyi | 13:41 | |
| zhiyan | amrith: if you got any blocker, pls let ping me, i can give a quick feedback to help try it out normally | 13:42 |
| *** jasonb365 has joined #openstack-trove | 13:43 | |
| *** robertmyers has joined #openstack-trove | 13:44 | |
| amrith | zhiyan, is there some documentation I should be reading | 13:44 |
| zhiyan | amrith: yep, probably i think you can take a glance at the bp/idea proposal in oslo-spec | 13:45 |
| zhiyan | 1 sec | 13:45 |
| zhiyan | amrith: https://review.openstack.org/#/c/103825/ ^ | 13:46 |
| zhiyan | amrith: btw, currently Glance and Cinder projects have already integrated with it. | 13:47 |
| zhiyan | amrith: the change for other projects is wip | 13:47 |
| zhiyan | like trove =) and heat | 13:47 |
| *** mattgriffin has joined #openstack-trove | 13:48 | |
| *** robertmyers has quit IRC | 13:48 | |
| *** tomblank has joined #openstack-trove | 13:54 | |
| *** achampion has joined #openstack-trove | 13:54 | |
| *** robertmyers has joined #openstack-trove | 13:56 | |
| *** shakamunyi has quit IRC | 13:58 | |
| *** edmund has quit IRC | 13:59 | |
| *** ajayaa has quit IRC | 14:03 | |
| *** ajayaa has joined #openstack-trove | 14:04 | |
| *** topshare has quit IRC | 14:05 | |
| *** topshare has joined #openstack-trove | 14:07 | |
| *** shakamunyi has joined #openstack-trove | 14:11 | |
| *** ajayaa has quit IRC | 14:16 | |
| *** iccha_ has joined #openstack-trove | 14:19 | |
| iccha_ | glucas: i have some bandwidth so let me know if i can help with any of the replication bugs | 14:19 |
| *** jcru has joined #openstack-trove | 14:20 | |
| *** shakamunyi has quit IRC | 14:22 | |
| *** shakamunyi has joined #openstack-trove | 14:23 | |
| *** iccha_ has quit IRC | 14:35 | |
| *** Riddhi has joined #openstack-trove | 14:39 | |
| *** robertmy_ has joined #openstack-trove | 14:39 | |
| *** shakamunyi has quit IRC | 14:40 | |
| *** jmontemayor has joined #openstack-trove | 14:40 | |
| *** juantwo has quit IRC | 14:40 | |
| *** juantwo has joined #openstack-trove | 14:40 | |
| *** kevinconway has joined #openstack-trove | 14:40 | |
| *** tomblank has quit IRC | 14:41 | |
| *** robertmyers has quit IRC | 14:42 | |
| *** juantwo has quit IRC | 14:43 | |
| *** juantwo has joined #openstack-trove | 14:43 | |
| *** grapex has joined #openstack-trove | 14:44 | |
| denis_makogon | iccha, i guess we're free to take any of those | 14:44 |
| denis_makogon | glucas, iccha are you there ? | 14:45 |
| denis_makogon | glucas, iccha, i know the way in which we can deal with users/schemes, volume/flavors whe attempting to create a slave | 14:47 |
| denis_makogon | glucas, iccha, we can provide another one client call that would be used to create a slave, but under the hood this call is nothing else than simple POST against /instances route (create call) | 14:49 |
| *** jmontemayor has quit IRC | 14:49 | |
| denis_makogon | this would give an ability to reduce amount of additional validation logic to Trove API | 14:50 |
| *** tomblank has joined #openstack-trove | 14:50 | |
| glucas | denis_makogon: you mean a variation of create without all the usual options? | 14:50 |
| denis_makogon | glucas, yes | 14:50 |
| *** jmontemayor has joined #openstack-trove | 14:51 | |
| denis_makogon | glucas, it'll work for all of us and will eliminate lots of filed bugs | 14:53 |
| denis_makogon | glucas, sorry, if this idea was proposed at mid-cycle meeting, i wan't aware of it since there's no report | 14:53 |
| *** openstackgerrit has joined #openstack-trove | 14:54 | |
| *** todd_dsm has quit IRC | 14:54 | |
| *** mattgriffin has quit IRC | 14:54 | |
| *** flaper87 is now known as flaper87|afk | 14:56 | |
| *** flaper87|afk is now known as flaper87 | 14:56 | |
| *** Longgeek has quit IRC | 14:56 | |
| *** vigneshvar has joined #openstack-trove | 14:56 | |
| *** mattgriffin has joined #openstack-trove | 14:58 | |
| glucas | denis_makogon: no, that wasn't discussed. | 15:00 |
| denis_makogon | glucas, great =) | 15:01 |
| *** edmund has joined #openstack-trove | 15:02 | |
| glucas | denis_makogon: would have to look at how different the set of args would be. We did discuss that it is a legitimate use case to make a slave with a smaller volume or flavor than than the master. | 15:02 |
| glucas | e.g. read slaves won't have all the same logs, won't need as much ram, etc | 15:02 |
| denis_makogon | glucas, agreed, user should be able to pass at least different flavor, not sure about volume | 15:03 |
| glucas | but we do want to prevent creating users or schemas as part of a slave create | 15:03 |
| *** topshare has quit IRC | 15:05 | |
| openstackgerrit | amrith proposed a change to openstack/trove: Update contributing.rst to include guidelines on Code Review https://review.openstack.org/115778 | 15:05 |
| glucas | iccha, vgnbkr, SlickNik: ^^ be curious what others think about a 'replica-create' command with different arg validation that calls down to the existing create impl. | 15:06 |
| denis_makogon | glucas, my proposal perfectly fits into it =) | 15:06 |
| *** topshare has joined #openstack-trove | 15:07 | |
| *** topshare has quit IRC | 15:07 | |
| *** topshare has joined #openstack-trove | 15:08 | |
| *** dkehn has quit IRC | 15:16 | |
| *** tkatarki has joined #openstack-trove | 15:17 | |
| *** georgelorch has quit IRC | 15:18 | |
| *** georgelorch has joined #openstack-trove | 15:18 | |
| denis_makogon | glucas, i found another one bug | 15:19 |
| denis_makogon | glucas, what happening when we're deleting root master ? | 15:19 |
| *** dkehn__ has joined #openstack-trove | 15:19 | |
| denis_makogon | glucas, according to given code, correct me if i'm wrong, we would have N detached slaves that are useless, right ? | 15:19 |
| *** spiffxp- has quit IRC | 15:19 | |
| *** achampion has quit IRC | 15:22 | |
| *** peoplemerge has quit IRC | 15:22 | |
| *** spiffxp_ has joined #openstack-trove | 15:22 | |
| *** dkehn__ is now known as dkehnx | 15:22 | |
| *** Longgeek has joined #openstack-trove | 15:22 | |
| *** achampion has joined #openstack-trove | 15:22 | |
| *** peoplemerge has joined #openstack-trove | 15:22 | |
| *** Longgeek has quit IRC | 15:22 | |
| amrith | vigneshvar, yt? | 15:23 |
| *** kevinconway has quit IRC | 15:23 | |
| vigneshvar | amrith: yes | 15:25 |
| amrith | vigneshvar, hello. | 15:25 |
| amrith | I'm looking at your chkin | 15:25 |
| vigneshvar | amrith: hi. Ya i guess i got deviated | 15:26 |
| amrith | according to this chkin, you are adding HTTP_PROXY. | 15:26 |
| amrith | should it be added for ALL sudo's? | 15:26 |
| vigneshvar | amrith: i have thought about it but felt it must added to places where only network related stuff is going on | 15:27 |
| vigneshvar | amrith: Like nick said i think my patch does not satisfy the bug | 15:28 |
| amrith | I see | 15:28 |
| amrith | yes | 15:28 |
| amrith | I think I understand the issue that SlickNik is raising | 15:28 |
| vigneshvar | amrith: should i open another bug, and apply the current patch | 15:29 |
| amrith | no, you don't need to open another bug | 15:30 |
| amrith | Let me review the other commit (94892). i read it a long time ago and don't know exactly what that fix proposed | 15:30 |
| *** juice has quit IRC | 15:30 | |
| amrith | I'll ping you after. | 15:30 |
| vigneshvar | amrith: sure thanks for your time | 15:31 |
| amrith | no problem. happy to help | 15:31 |
| vigneshvar | amrith: just a doubt, like http_proxy variable, no_proxy nbeeds to be set by the user himself ? or should that be automated | 15:32 |
| amrith | Good question, I don't know for sure but I think the setting should be automated. | 15:32 |
| vigneshvar | amrith: ok sure i will look from that askpect thanks | 15:32 |
| amrith | will check what is being attempted and ping you. | 15:32 |
| vigneshvar | amrith: ok :) | 15:33 |
| amrith | One thing I'd suggest is that you review the old chkins that the other person proposed. | 15:33 |
| *** annashen has joined #openstack-trove | 15:33 | |
| glucas | denis_makogon: I wouldn't call that a bug really... But there is currently no special handling for this case. | 15:33 |
| amrith | maybe that'll give you an indication of what he/she was thinking. | 15:33 |
| denis_makogon | glucas, take a look at how we deal with incremental backups while deleting one from the chain, so i think that this is a bug | 15:34 |
| *** openstack has joined #openstack-trove | 16:18 | |
| *** ViswaV has joined #openstack-trove | 16:21 | |
| *** eghobo has joined #openstack-trove | 16:22 | |
| *** eghobo has quit IRC | 16:23 | |
| *** eghobo has joined #openstack-trove | 16:23 | |
| *** georgelorch has quit IRC | 16:25 | |
| *** annashen has quit IRC | 16:26 | |
| *** jasonb365 has quit IRC | 16:28 | |
| iccha | glucas: SlickNik do you want to sometime today touchbase on the high priorty replication bugs and see if there is an order we would like to work on them? also if we want to distribute the work | 16:32 |
| openstackgerrit | Denis M. proposed a change to openstack/trove: Apply recursive initial master lookup while replica provisioning https://review.openstack.org/117298 | 16:33 |
| *** boblebauce has left #openstack-trove | 16:33 | |
| glucas | denis_makogon: backups are a bit different, right? You're not doing a cascading delete of running instances that may have clients connected. | 16:33 |
| iccha | denis_makogon: should we even allow recurrsive calls? | 16:33 |
| denis_makogon | glucas, but if you're deleting master - you'r getting into trouble =) | 16:33 |
| denis_makogon | iccha, sure | 16:34 |
| denis_makogon | iccha, about backup - of course, because incremental sequence matters | 16:34 |
| denis_makogon | iccha, same thing is for replication set, if we're deleting master, we should delete all slaves once and for all | 16:35 |
| denis_makogon | glucas, iccha, SlickNik, you can count on me, i'd like to help with those bugs | 16:35 |
| openstackgerrit | Denis M. proposed a change to openstack/trove: Apply recursive initial master lookup while replica provisioning https://review.openstack.org/117298 | 16:37 |
| openstackgerrit | Denis M. proposed a change to openstack/trove: Fix master status validation while replica provisioning https://review.openstack.org/117255 | 16:37 |
| *** georgelorch has joined #openstack-trove | 16:37 | |
| jedix | Is there a way to remove a datastore from trove with trove-manage? | 16:37 |
| iccha | awesome thanks denis_makogon | 16:37 |
| iccha | denis_makogon: i dont know if we should delete them or dettach them, can u add ur bugs to etherpad denis | 16:38 |
| *** ViswaV has quit IRC | 16:39 | |
| iccha | glucas: also i know ur changing name to replica but since api call is merged we cant change it in reponse | 16:40 |
| iccha | *repsonse | 16:40 |
| denis_makogon | iccha, i didn't file it yet, just wanted to ask glucas before doing that | 16:40 |
| *** kevinconway has joined #openstack-trove | 16:41 | |
| denis_makogon | iccha, i'm leaving, gonna be too late today, so i'm going to miss meeting =( | 16:41 |
| denis_makogon | iccha, API is not released yet | 16:41 |
| denis_makogon | iccha, so we can change everything we want before H3 | 16:42 |
| denis_makogon | J3 | 16:42 |
| denis_makogon | cu, guys, later | 16:42 |
| iccha | gotcha thanks denis_makogon . yeah we dont have in todays meeting agenda uyet. whenevr u re free add your thoughts to etherpad. maybe new subtitle needs to be prioritized | 16:42 |
| *** rushiagr_away is now known as rushiagr | 16:43 | |
| *** kevinconway has quit IRC | 16:45 | |
| iccha | the api did get merged denis_makogon glucas https://review.openstack.org/#/c/105758/ | 16:45 |
| *** georgelorch has quit IRC | 16:45 | |
| jedix | How do I specify a volume size for trove create? I'm getting "ERROR: Volume size was not specified." | 16:46 |
| iccha | jedix: are u using the client | 16:46 |
| *** todd_dsm has joined #openstack-trove | 16:46 | |
| iccha | jedix: if yes --size 1 | 16:47 |
| jedix | iccha: I am | 16:47 |
| jedix | what is the unit of size? | 16:47 |
| jedix | sigh, status == error. | 16:48 |
| iccha | jedix: size is in gb | 16:48 |
| iccha | jedix: do u have cinder running for volumes | 16:48 |
| glucas | jedix: --size <size> Size of the instance disk volume in GB. | 16:48 |
| jedix | iccha: yes. | 16:49 |
| jedix | Checking for db [id=d76cd864-840a-4f87-bd87-5839fdf0c706, compute_instance_id=None] | 16:49 |
| iccha | jedix: restart service cinder-volume restart | 16:49 |
| jedix | Cannot find compute instance None for instance | 16:49 |
| jedix | iccha: how do I restart an instance in ERROR state? it's saying it's not ready when I run trove restart <id> | 16:51 |
| iccha | jedix: yeah u cannot restart an instance in error status | 16:51 |
| iccha | jedix: u ll have to delete it and create a new one | 16:51 |
| jedix | iccha: okay, thanks | 16:52 |
| jedix | iccha: is there a way to verify cinder connectivity? | 16:52 |
| jedix | Cannot find compute instance None for instance | 16:52 |
| jedix | (again) | 16:52 |
| iccha | jedix: u can look at logs | 16:53 |
| iccha | and see where it is erroring | 16:53 |
| jedix | that's what the log says.. | 16:55 |
| jedix | cannot find compute instance None | 16:55 |
| iccha | jedix: can u look at nova-compute logs | 16:56 |
| iccha | grep for error | 16:56 |
| jedix | iccha: no errors there.. | 16:57 |
| iccha | jedix: can u look cinder logs | 16:57 |
| *** vigneshvar has quit IRC | 16:57 | |
| *** jcru has quit IRC | 16:57 | |
| *** tomblank has quit IRC | 16:57 | |
| jedix | oh wow.. yeah, there's errors there. | 16:58 |
| *** georgelorch has joined #openstack-trove | 16:58 | |
| *** jcru has joined #openstack-trove | 16:59 | |
| *** jcru has quit IRC | 17:04 | |
| *** harlowja has joined #openstack-trove | 17:04 | |
| openstackgerrit | Greg Lucas proposed a change to openstack/trove: Add detach-replica support https://review.openstack.org/101981 | 17:04 |
| *** ViswaV has joined #openstack-trove | 17:06 | |
| *** tomblank has joined #openstack-trove | 17:08 | |
| *** georgelorch has quit IRC | 17:09 | |
| *** jmontemayor has quit IRC | 17:12 | |
| *** mattgriffin has quit IRC | 17:15 | |
| *** ViswaV has quit IRC | 17:19 | |
| glucas | iccha: Your earlier comment about slave vs. replica -- are you saying that because we have a merged commit we can no longer make an API change? I was expecting that we could continue to do so for Juno since the code is unreleased (and noone could actually have been using this API until a few days ago, if at all...) | 17:20 |
| *** jmontemayor has joined #openstack-trove | 17:21 | |
| *** georgelorch has joined #openstack-trove | 17:21 | |
| openstackgerrit | Greg Lucas proposed a change to openstack/trove: Add detach-replica support https://review.openstack.org/101981 | 17:25 |
| *** saurabhs has joined #openstack-trove | 17:25 | |
| *** arborism has joined #openstack-trove | 17:26 | |
| *** arborism has quit IRC | 17:26 | |
| *** amcrn has joined #openstack-trove | 17:27 | |
| *** jcru has joined #openstack-trove | 17:28 | |
| *** tomblank has quit IRC | 17:29 | |
| *** annashen has joined #openstack-trove | 17:32 | |
| *** jcru has quit IRC | 17:32 | |
| SlickNik | iccha: +1 about quickly catching up on where we are with the bugs. Is after this morning's Trove meeting a good time for you guys? | 17:39 |
| SlickNik | ^replication related bugs. | 17:39 |
| SlickNik | also glucas / iccha: I think we're okay to fast follow with a quick API change if we feel like that's the right thing to do — since this code hasn't yet been released. | 17:40 |
| *** mattgriffin has joined #openstack-trove | 17:49 | |
| *** tomblank has joined #openstack-trove | 17:51 | |
| openstackgerrit | amrith proposed a change to openstack/trove: In some cases, guest agents may leave temporary config files https://review.openstack.org/112672 | 17:52 |
| *** ViswaV has joined #openstack-trove | 17:56 | |
| *** ViswaV_ has joined #openstack-trove | 17:57 | |
| *** ranjitha has joined #openstack-trove | 17:57 | |
| iccha | SlickNik: yes lets sync up after trove meeting | 17:58 |
| *** kevinconway has joined #openstack-trove | 17:58 | |
| *** jcru has joined #openstack-trove | 18:00 | |
| *** ViswaV has quit IRC | 18:00 | |
| *** jasonb365 has joined #openstack-trove | 18:02 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Clusters API Implementation https://review.openstack.org/113675 | 18:03 |
| *** vigneshvar has joined #openstack-trove | 18:03 | |
| openstackgerrit | Steve Leon proposed a change to openstack/trove: Make storage strategy available for trove API and TM https://review.openstack.org/86242 | 18:06 |
| *** Barker has quit IRC | 18:07 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/python-troveclient: Clusters troveclient Implementation https://review.openstack.org/113710 | 18:08 |
| *** Barker has joined #openstack-trove | 18:10 | |
| *** robertmy_ is now known as robertmyers | 18:11 | |
| *** sriram_tesora has joined #openstack-trove | 18:14 | |
| *** rushiagr is now known as rushiagr_away | 18:16 | |
| openstackgerrit | Mat Lowery proposed a change to openstack/trove: Clusters TaskManager Implementation https://review.openstack.org/113692 | 18:21 |
| *** RicoLin has quit IRC | 18:21 | |
| openstackgerrit | Ramashri Umale proposed a change to openstack/trove: Clusters Guest Implementation https://review.openstack.org/113728 | 18:24 |
| *** ViswaV_ has quit IRC | 18:32 | |
| *** esp has left #openstack-trove | 18:33 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Clusters API Implementation https://review.openstack.org/113675 | 18:33 |
| iccha | SlickNik: looks like the blueprint associated with this review dispappeared https://review.openstack.org/#/c/107117/ | 18:34 |
| iccha | cant find it | 18:34 |
| openstackgerrit | Simon Chang proposed a change to openstack/trove: Fixed database migration script issues https://review.openstack.org/113582 | 18:34 |
| SlickNik | iccha: huh. That's very weird. | 18:34 |
| iccha | SlickNik: ok found it https://blueprints.launchpad.net/trove/+spec/per-datastore-volume-support it needs to be targetted | 18:38 |
| iccha | its already merged | 18:38 |
| iccha | just wanted to make sure it is included | 18:39 |
| *** vigneshvar has quit IRC | 18:39 | |
| *** annashen has quit IRC | 18:40 | |
| iccha | SlickNik: glucas around to talk about https://etherpad.openstack.org/p/trove-replication-bugs ? | 18:41 |
| glucas | dougshelley66: ^^ | 18:41 |
| dougshelley66 | glucas, i'm here | 18:42 |
| SlickNik | iccha: yes, one sec. | 18:42 |
| *** zigo has quit IRC | 18:43 | |
| *** Barker has quit IRC | 18:43 | |
| *** annashen has joined #openstack-trove | 18:44 | |
| SlickNik | iccha: when I try to create a bp, it tells me that one with the same name exists. But if I search for it, nothing comes up. | 18:45 |
| iccha | SlickNik: yeah i had the same issue. finnaly dug up link from my emails | 18:46 |
| *** harlowja has quit IRC | 18:46 | |
| SlickNik | Here it is: https://blueprints.launchpad.net/trove/+spec/per-datastore-volume-support | 18:46 |
| *** zigo has joined #openstack-trove | 18:46 | |
| *** harlowja has joined #openstack-trove | 18:46 | |
| iccha | SlickNik: thanks, wanted to make sure its approved and targetted since the code is already merged | 18:47 |
| iccha | btw SlickNik tomblank is our new manager. forgot to introduce during the meeting :) | 18:47 |
| SlickNik | Yup, thanks for that. Approved, and targeted appropriately now. | 18:48 |
| SlickNik | Hi tomblank! | 18:48 |
| tomblank | hey SlickNik…. | 18:48 |
| dougshelley66 | Hi tomblank | 18:48 |
| openstackgerrit | amrith proposed a change to openstack/trove: In some cases, guest agents may leave temporary config files https://review.openstack.org/112672 | 18:48 |
| tomblank | Hey dougshelley66 and all | 18:49 |
| *** julienvey has quit IRC | 18:51 | |
| *** sgotliv has joined #openstack-trove | 18:53 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Clusters API Implementation https://review.openstack.org/113675 | 18:54 |
| *** ViswaV has joined #openstack-trove | 18:54 | |
| *** todd_dsm has quit IRC | 19:02 | |
| *** georgelorch has quit IRC | 19:04 | |
| openstackgerrit | Mat Lowery proposed a change to openstack/trove: Clusters TaskManager Implementation https://review.openstack.org/113692 | 19:05 |
| *** annashen has quit IRC | 19:08 | |
| *** Barker has joined #openstack-trove | 19:09 | |
| iccha | glucas: vgnbkr is there any of the bugs assined to yall you would like me to pick up ? | 19:10 |
| openstackgerrit | Ramashri Umale proposed a change to openstack/trove: Clusters Guest Implementation https://review.openstack.org/113728 | 19:10 |
| dougshelley66 | iccha, yes | 19:10 |
| dougshelley66 | can you take the metadata one | 19:11 |
| dougshelley66 | https://bugs.launchpad.net/trove/+bug/1359363 | 19:11 |
| iccha | sure dougshelley66 | 19:11 |
| dougshelley66 | thx | 19:11 |
| iccha | dougshelley66: also i spoke to some of my managers and they said deleting backup is not a priority for now if we want to focus on other bugs | 19:13 |
| *** denis_makogon_ has joined #openstack-trove | 19:13 | |
| dougshelley66 | iccha, i would have thought for a cloud provider not leaving the backup around would be important but thanks for the info | 19:14 |
| denis_makogon_ | glucas, iccha https://bugs.launchpad.net/trove/+bug/1362299 | 19:14 |
| denis_makogon_ | SlickNik, around? | 19:15 |
| vgnbkr | Apparently it also reflects poorly in Horizon having all these useless "backups" showing up. | 19:15 |
| iccha | dougshelley66: that doesnt mean we dont do it :) just in comparison to marking slave read only and replication password bugs | 19:15 |
| dougshelley66 | iccha, ok thanks for the clarification | 19:16 |
| iccha | vgnbkr: dougshelley66 i am all for doing it though | 19:16 |
| denis_makogon_ | vgnbkr, we can add some sort of flag/description to backup entity to mark it as "replication use only" | 19:16 |
| dougshelley66 | iccha, would you like to take the delete backup one? | 19:17 |
| vgnbkr | denis_makogon_: Sure, but wouldn't it be just as easy to delete them? | 19:17 |
| iccha | sure dougshelley66 | 19:17 |
| denis_makogon_ | vgnbkr, sure | 19:17 |
| denis_makogon_ | iccha, dougshelley66, one comment about deleting backup | 19:18 |
| denis_makogon_ | iccha, dougshelley66, i'd suggest to remove use of user quotas, since backup appears to be alive very short period of time | 19:18 |
| denis_makogon_ | glucas, would you take this one https://bugs.launchpad.net/trove/+bug/1362299 ? | 19:19 |
| iccha | denis_makogon_: vgnbkr dougshelley66 i say we we forbid deletes on a master unless user has detached all slaves | 19:20 |
| denis_makogon_ | if noone has objections i'd like to take this one - Snapshot on incremental backup on existing backup | 19:20 |
| *** georgelorch has joined #openstack-trove | 19:21 | |
| denis_makogon_ | iccha, sound ok for me | 19:21 |
| *** PeterSchwaller has joined #openstack-trove | 19:22 | |
| denis_makogon_ | iccha, glucas, dougshelley66, what about API call for "replica-delete <master_id> ? | 19:22 |
| vgnbkr | I suggest keeping user quotas. The snapshots, though transient, could be huge and exist for weeks. | 19:22 |
| denis_makogon_ | vgnbkr, snapshot exists only if it's not deleted | 19:23 |
| denis_makogon_ | vgnbkr, i guess we've planned to drop snapshot once slave is up, right? | 19:24 |
| vgnbkr | denis_makogon_: right, but it could take weeks to take a backup then restore it. | 19:24 |
| *** grapex has quit IRC | 19:24 | |
| denis_makogon_ | vgnbkr, in this case, completely agreed | 19:24 |
| *** grapex has joined #openstack-trove | 19:24 | |
| iccha | denis_makogon_: vgnbkr do we want to make it optional to keep the backup around then? | 19:25 |
| *** PeterSchwaller__ has quit IRC | 19:25 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Clusters API Implementation https://review.openstack.org/113675 | 19:25 |
| vgnbkr | denis_makogon_: I've already been giving thought to the incremental snapshot and it is my idea. Do you mind if I see it through? | 19:25 |
| denis_makogon_ | does only once consider on adding tests for mysql bulk load and do testing for backups/replication ? | 19:26 |
| denis_makogon_ | *any one | 19:26 |
| denis_makogon_ | vgnbkr, of course | 19:26 |
| dougshelley66 | denis_makogon, SlickNik, iccha, vgnbkr, glucas we should make a decision around the replica_* apis | 19:26 |
| vgnbkr | iccha: I have no problem with keeping the snapshot/backup around, but wonder if it would really be useful. | 19:26 |
| vgnbkr | denis_makogon_: thanks. | 19:26 |
| denis_makogon_ | vgnbkr, np | 19:27 |
| iccha | denis_makogon_: vgnbkr we are talking about a lot of things parallel. do we want to quickly scan the etherpad in order? | 19:27 |
| iccha | https://etherpad.openstack.org/p/trove-replication-bugs | 19:27 |
| denis_makogon_ | iccha, vgnbkr, dougshelley66 lets do this | 19:27 |
| dougshelley66 | iccha, i think we are good with the first 5 | 19:28 |
| iccha | dougshelley66: and one on time outs as well? | 19:28 |
| vgnbkr | Can everyone fill in their name on that etherpad? | 19:28 |
| denis_makogon_ | so, if everyone is ok with first 5 | 19:28 |
| dougshelley66 | so the "Different usage..." and the "Method that pulls" | 19:28 |
| dougshelley66 | they are somewhat related | 19:28 |
| * SlickNik is back and looking at the etherpad | 19:29 | |
| denis_makogon_ | dougshelley66, i agree, since "different usage" states about proposing additional polling timeout for restoring, and "method that pulls" talking about changing the type of API call (from call to cast) | 19:31 |
| iccha | for backup metadat one anyone have prefer for what kind of metadata is added? | 19:31 |
| denis_makogon_ | dougshelley66, they are all about timeouts | 19:31 |
| dougshelley66 | denis_makogon, right, that is why i puinted that out | 19:31 |
| vgnbkr | iccha: What sort of backup metadata do you require? You suggested that just having "replication" in the name or description was not sufficient, but I don't recall why. | 19:32 |
| iccha | vgnbkr: just so billing can look at a specific field in the notification | 19:33 |
| *** denis_makogon has quit IRC | 19:33 | |
| *** denis_makogon_ is now known as denis_makogon | 19:33 | |
| *** dmakogon_ has joined #openstack-trove | 19:34 | |
| denis_makogon | iccha, i've just helped me to remind what we have missed - notifications | 19:34 |
| glucas | sorry folks, was pulled away for something else.. catching up on dsicussion here / etherpad | 19:39 |
| *** todd_dsm has joined #openstack-trove | 19:40 | |
| denis_makogon | dougshelley66, SlickNik, glucas, vgnbkr, guys, what about adding API call that deletes whole replication set? | 19:41 |
| denis_makogon | for now to delete replication set we need to detach all slaves, delete them, then delete the master | 19:41 |
| iccha | denis_makogon: i think thats the safer ption | 19:41 |
| iccha | so the user is aware of the consequences of deleting the master | 19:42 |
| *** ranjitha has quit IRC | 19:42 | |
| iccha | manually detaching everything first should be the recommded behavior in my opinion | 19:42 |
| SlickNik | iccha: +1, I agree. | 19:42 |
| denis_makogon | iccha, as i can forsee, we need both cases | 19:42 |
| SlickNik | denis_makogon: For replication (at least for async mysql replication), the mental model is disparate instances sharing some data. | 19:43 |
| openstackgerrit | Mat Lowery proposed a change to openstack/trove: Clusters TaskManager Implementation https://review.openstack.org/113692 | 19:43 |
| SlickNik | There really isn't a concept of a "set" — that sort of mental modal is more suited for clustering imho. | 19:44 |
| openstackgerrit | Auston McReynolds proposed a change to openstack/python-troveclient: Clusters troveclient Implementation https://review.openstack.org/113710 | 19:44 |
| denis_makogon | SlickNik, but it's too close to read scale out replica set | 19:44 |
| dougshelley66 | ok i will type over here | 19:45 |
| dougshelley66 | so for https://bugs.launchpad.net/trove/+bug/1362062 are we going to do the huge=maxint thing as a workaround? | 19:46 |
| denis_makogon | dougshelley66, as temp solution | 19:46 |
| dougshelley66 | because i think the proper fix is larger than we have time for in J | 19:46 |
| glucas | dougshelley66: +1 | 19:46 |
| denis_makogon | i feel concerned not having one API call that eliminates master and its deps, because of need to delete 3 node replica we need to use 2 calls to detach, 3 calls to delete | 19:47 |
| SlickNik | dougshelley66: I'd be okay with that fix in Juno, but we should track the issue for Kilo. | 19:47 |
| denis_makogon | SlickNik, ++ | 19:47 |
| dougshelley66 | SlickNik, agreed | 19:48 |
| *** todd_dsm has quit IRC | 19:48 | |
| glucas | denis_makogon: I think we need the 'atomic' blocks of the API to be right, users can always script multi-step actions (and we can layer them in to the API later) | 19:48 |
| dougshelley66 | do we need a new bug for that? is the temp/workaround fixed un https://bugs.launchpad.net/trove/+bug/1362062 | 19:48 |
| glucas | denis_makogon: similar arguments were made when we originally proposed creating multiple slaves in one call. | 19:49 |
| denis_makogon | glucas, muliple create call can cause lots of troubles, but how can delete call for all of them can cause something bad ? | 19:50 |
| iccha | denis_makogon: you lose data | 19:50 |
| SlickNik | denis_makogon: There really isn't a concept of an entity called a "set" in this case, so what resource do you call the delete against? IMO it murkies the API. | 19:50 |
| *** sgotliv has quit IRC | 19:50 | |
| dougshelley66 | SlickNik, +1 | 19:50 |
| denis_makogon | ok | 19:51 |
| dougshelley66 | So do we all think we have a handle on the "High" list? | 19:51 |
| iccha | yeah looks good to me | 19:52 |
| SlickNik | going through it again right now, one sec. | 19:52 |
| *** julienvey has joined #openstack-trove | 19:52 | |
| SlickNik | Looks good to me. | 19:52 |
| dougshelley66 | SlickNik, how would you like to track the workaround and the Kilo work for it? | 19:53 |
| SlickNik | Of that list, I think I'd like the slave - replica terminology one merge before juno-3, but the rest of the bugsfixes can come before the RC is cut. | 19:53 |
| SlickNik | ^glucas, I think you're already aware, but just in case. | 19:54 |
| denis_makogon | iccha, https://bugs.launchpad.net/trove/+bug/1362299 i already filed bug for you | 19:54 |
| iccha | thanks denis_makogon | 19:54 |
| denis_makogon | iccha, np | 19:54 |
| SlickNik | dougshelley66: File a bugs for the Kilo work, and I'll triage and assign it to trove-next (until kilo is created). | 19:55 |
| jedix | Is there no way to drop the Legacy MySQL datastore? | 19:55 |
| dougshelley66 | ok so vgnbkr filed a bug for the workaround; we will then move the "Method..." item and discussion lower on the list | 19:55 |
| glucas | SlickNik: OK, I'll work on that! | 19:55 |
| denis_makogon | jedix, nope, sorry | 19:55 |
| jedix | denis_makogon: np, Is there a way to set the default one? | 19:55 |
| jedix | instead of having to do --datastore <id> every time? | 19:56 |
| denis_makogon | jedix, yes, there's a option for Trove API service, called defaul_datastore | 19:56 |
| *** julienvey has quit IRC | 19:56 | |
| denis_makogon | if it's not mentioned we don't have to use --datastore | 19:56 |
| jedix | it tells me I have to when I try to create something | 19:57 |
| SlickNik | dougshelley66: what's the bug for the workaround that vgnbkr filed? | 19:57 |
| dougshelley66 | it is on the etherpad | 19:58 |
| dougshelley66 | https://bugs.launchpad.net/trove/+bug/1362310 | 19:58 |
| *** todd_dsm has joined #openstack-trove | 19:58 | |
| denis_makogon | jedix, have you updated conf file for trove-api? | 19:58 |
| *** jmontemayor has quit IRC | 20:00 | |
| jedix | updated since when? I'm using git id c9173f22e91c1f15179a8163b6c4748d53db4f0f | 20:00 |
| jedix | denis_makogon: I did just add default_datastore postgresql to my trove.conf and that seems to have fixed my issue | 20:00 |
| jedix | (I'm using postgresql + kevinconway patch set 13) | 20:00 |
| *** jmontemayor has joined #openstack-trove | 20:00 | |
| dougshelley66 | SlickNik, since you have the other timeout one do you want to take this one | 20:01 |
| denis_makogon | jedix, that's what i've said - add default_datastore to trove.conf | 20:01 |
| denis_makogon | SlickNik, are there any plans to write meetup report ? | 20:01 |
| SlickNik | dougshelley66: Sure I can take care of this as well. | 20:01 |
| dougshelley66 | SlickNik, thx | 20:01 |
| dougshelley66 | SlickNik, while i have you on the line :) can you triage this one: https://bugs.launchpad.net/trove/+bug/1361389 | 20:01 |
| *** tomblank_ has joined #openstack-trove | 20:02 | |
| jedix | denis_makogon: oh, I was looking for trove-api <command> | 20:02 |
| SlickNik | denis_makogon: I'm in the process of writing one — plan to send it out this evening. | 20:02 |
| jedix | denis_makogon: thanks! | 20:02 |
| denis_makogon | jedix, np | 20:02 |
| *** tomblank has quit IRC | 20:02 | |
| *** tomblank_ is now known as tomblank | 20:02 | |
| denis_makogon | SlickNik, sound awesome | 20:02 |
| denis_makogon | *sounds | 20:02 |
| *** sgotliv has joined #openstack-trove | 20:02 | |
| SlickNik | dougshelley66: We should probably involve the folks from RAX in the conversation for that — since they made the call to not delete the tables on downgrade in the first place, I want to make sure we're not overlooking anything. | 20:03 |
| dougshelley66 | SlickNik, ok makes sense - schang should talk to cp16net | 20:04 |
| SlickNik | ^ cp16net / grapex | 20:04 |
| *** kevinconway has quit IRC | 20:04 | |
| *** jmontemayor has quit IRC | 20:05 | |
| *** jmontemayor has joined #openstack-trove | 20:08 | |
| *** kevinconway has joined #openstack-trove | 20:08 | |
| *** esp has joined #openstack-trove | 20:10 | |
| *** todd_dsm has quit IRC | 20:12 | |
| *** ViswaV has quit IRC | 20:15 | |
| openstackgerrit | Greg Lucas proposed a change to openstack/trove: Add detach-replica support https://review.openstack.org/101981 | 20:16 |
| cp16net | sure | 20:18 |
| cp16net | i'm here | 20:18 |
| cp16net | i'm catching up... | 20:18 |
| schang | cp16net: This bug was created and stating config tables should be dropped on downgrade https://bugs.launchpad.net/trove/+bug/1361389. | 20:18 |
| schang | cp16net: The main reason behind the dropping table proposal is that the config tables may contain corrupted data, which may cause subsequent issues on upgrades and therefore should be dropped on downgrade. | 20:19 |
| schang | cp16net: I notice you implemented the original downgrade code. Can you provide insights? Thanks! | 20:19 |
| cp16net | OH... | 20:19 |
| cp16net | yeah i recall why we decided that | 20:20 |
| cp16net | we thought at the time if we needed to backout of the configurations we didnt want to delete all the user's config groups they created | 20:21 |
| cp16net | seems a bit destructive | 20:21 |
| *** todd_dsm has joined #openstack-trove | 20:22 | |
| schang | cp16net: But the other downgrade scripts are dropping the tables being created, and some of those tables too contain user created data (e.g. 012_backup.py). Your thoughts? | 20:27 |
| *** sgotliv has quit IRC | 20:27 | |
| openstackgerrit | Riddhi Shah proposed a change to openstack/trove: Associate flavor types with datastore versions https://review.openstack.org/109824 | 20:27 |
| cp16net | schang: i think now that configurations has been used for a while and deployed i doubt that there is a need to rollback like it was when we first merged the feature | 20:28 |
| schang | cp16net: so I guess it's OK to proceed with dropping the config tables on downgrade as proposed? | 20:32 |
| cp16net | i dont see an issue with it. (we wont be dropping those tables) :) | 20:32 |
| denis_makogon | SlickNik, could you please take a look ? https://bugs.launchpad.net/trove/+bug/1362318 | 20:33 |
| cp16net | schang: if there was another person that poked at the configs alot was amcrn | 20:34 |
| amcrn | schang: we've avoided dropping the tables for quite a few schema upgrades; we pivoted on our thinking probably in the late teens (015_ and onward) | 20:35 |
| *** pdmars has quit IRC | 20:39 | |
| *** georgelorch has quit IRC | 20:46 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Clusters API Implementation https://review.openstack.org/113675 | 20:47 |
| schang | amcrn: I just double checked, the create_tables in upgrade for all 31 migration scripts are paired with drop_tables in downgrade. Am I missing something? | 20:48 |
| amcrn | schang: we haven't added a new table in a long time :) | 20:49 |
| amcrn | if you've got corrupted data, just drop the table manually | 20:49 |
| schang | amcrn: also, the dropping doesn't seem to be new (015_ and onward), it's pretty consistant across the board. | 20:49 |
| *** ranjitha has joined #openstack-trove | 20:50 | |
| amcrn | schang: for example, in our clusters in-flight patch, it doesn't make sense that on a downgrade you drop the entire table; there's no way to reconstruct the data | 20:51 |
| vgnbkr | Either way, the upgrade needs to work. If you don't want to have the downgrade delete tables, then the upgrade needs to do more than just ignore exceptions. | 20:52 |
| *** julienvey has joined #openstack-trove | 20:53 | |
| vgnbkr | The upgrade needs to validate the existence of the tables, their shape, and the validity of any data existing in the tables. | 20:54 |
| *** julienvey has quit IRC | 20:54 | |
| *** julienvey has joined #openstack-trove | 20:54 | |
| *** juantwo has quit IRC | 20:55 | |
| amcrn | vgnbkr: if you want to write a 300 line upgrade script, go for it ;) | 20:56 |
| amcrn | fairly impossible considering the existing constraints aren't named | 20:56 |
| amcrn | you'd have to write some pretty in-depth logic to assert everything | 20:56 |
| schang | amcrn: We also have been adding tables recently ... the most recent one is at 028, where the new_agent_heartbeats table is added, and that migration script also drops the table on downgrade. | 20:57 |
| amcrn | schang: heartbeat data can be dropped and rebuilt within 3 seconds | 20:57 |
| amcrn | or whatever the heartbeat push interval is (i don't remember) | 20:58 |
| *** ViswaV has joined #openstack-trove | 20:59 | |
| *** julienvey has quit IRC | 20:59 | |
| vgnbkr | amcrn: Right, that's why we suggested deleting the tables. How is it acceptable to have upgrade methods that silently fail, or leave corrupted data generated by errors in future versions? | 21:00 |
| amcrn | it's probably also worth mentioning that i've never met a single person that has done a downgrade in production | 21:00 |
| vgnbkr | Great, then let's get rid of the downgrade functionality and have the upgrade methods work properly. | 21:00 |
| amcrn | vgnbkr: if you want to wrap every statement separately in an upgrade, that's fine. | 21:00 |
| amcrn | but your comment suggested you'd actually assert the state | 21:01 |
| *** georgelorch has joined #openstack-trove | 21:01 | |
| amcrn | the former is reasonable, the latter is nearly impossible (when considering you've got postgres + mysql + sqlite to contend with) | 21:01 |
| vgnbkr | So what is the usecase for someone doing a downgrade, then an upgrade, but not expecting things to have been reset? | 21:03 |
| *** ViswaV has quit IRC | 21:04 | |
| *** ViswaV_ has joined #openstack-trove | 21:04 | |
| *** sriram_tesora has quit IRC | 21:04 | |
| *** ViswaV_ has quit IRC | 21:04 | |
| amcrn | you're probably talking to the wrong person, because i think the whole downgrade idea is an anti-pattern | 21:04 |
| amcrn | just because the tables drop and the rows drop doesn't mean the computes and volumes and security-groups and <etc> magically disappear | 21:05 |
| amcrn | unraveling indexes and constraints on a downgrade are relatively safe, but anything else is asking for trouble | 21:05 |
| vgnbkr | I agree, so why would we have upgrade implemented improperly to support a questionable feature? | 21:06 |
| amcrn | vgnbkr: because someone ran into this problem in their deployment | 21:06 |
| amcrn | we can't state "in trove, we don't subscribe to the notion of a downgrade at all, best of luck" | 21:07 |
| openstackgerrit | Mat Lowery proposed a change to openstack/trove: Clusters TaskManager Implementation https://review.openstack.org/113692 | 21:07 |
| *** georgelorch has quit IRC | 21:07 | |
| vgnbkr | That's OK, too, but upgrade needs to work. | 21:08 |
| amcrn | vgnbkr: if you submitted a patch that wrapped every single side-effecting statement in an upgrade() with its own try/catch, i'd +2 it | 21:08 |
| amcrn | i'm sure someone could come up with a more clever way of writing it; something more pythonic | 21:09 |
| kevinconway | which review are you two talking about? | 21:09 |
| vgnbkr | But that wouldn't solve the problem. The issue isn't about which line causes the exception, rather making sure that the initial upgrade works. | 21:10 |
| amcrn | to be curious, what's your vested interest in this? are you actually going to ever run a downgrade? | 21:10 |
| amcrn | and if so, why. | 21:10 |
| vgnbkr | No, I agree with you that downgrade is stupid. I just think upgrade should work in all cases. | 21:11 |
| amcrn | if you never downgrade, an upgrade will always work, no? | 21:12 |
| amcrn | or are you arguing that we're possibly swallowing things in say 019_ and continuing onto 020_ | 21:12 |
| vgnbkr | No, that was my original point. Eating the exception means that an error would leave the database in an unknown state and silently move on. | 21:12 |
| vgnbkr | Right. | 21:13 |
| amcrn | well, because we're stuck supporting the "downgrade then upgrade back" scenario, not sure how to fix it. | 21:13 |
| *** tomblank has quit IRC | 21:13 | |
| amcrn | since the migration table does not have a history | 21:13 |
| amcrn | it's a single row | 21:13 |
| vgnbkr | If we want to say "downgrade at your own risk", OK, but upgrade must work in all cases in the initial upgrade. | 21:13 |
| *** robertmyers has quit IRC | 21:13 | |
| vgnbkr | Well, work or throw an exception to stop the upgrade process - we can't just silently move on. | 21:14 |
| amcrn | vgnbkr: the problem is, at a point in time, you can't tell whether a person is on 015_ because they downgraded, or because that's what the last version was when they last ran db_sync | 21:14 |
| *** kevinconway has left #openstack-trove | 21:15 | |
| amcrn | if migrate_version had a prior_version column, then we could easily guarantee an upgrade works as expected | 21:16 |
| vgnbkr | OK, so then at a minimum, explicitly check for the existence of the tables - relying on exceptions isn't the way to do it. | 21:17 |
| *** kevinconway has joined #openstack-trove | 21:18 | |
| *** georgelorch has joined #openstack-trove | 21:19 | |
| *** jmontemayor has quit IRC | 21:19 | |
| amcrn | vgnbkr: seems like a reasonable compromise | 21:20 |
| openstackgerrit | Ramashri Umale proposed a change to openstack/trove: Clusters Guest Implementation https://review.openstack.org/113728 | 21:22 |
| vgnbkr | OK, simon will not drop the tables, but won't execute the CREATE statements if the tables already exist. | 21:23 |
| vgnbkr | If the upgrade does not execute CREATE because the table already exists, it should log a message. Wondering if it should be INFO or WARN? | 21:23 |
| *** jmontemayor has joined #openstack-trove | 21:24 | |
| amcrn | my hunch: warn | 21:24 |
| vgnbkr | Mine, too. Thanks. | 21:24 |
| schang | ok then, I'll submit a new patch. | 21:24 |
| *** julienvey has joined #openstack-trove | 21:26 | |
| *** achampion has quit IRC | 21:27 | |
| *** ViswaV has joined #openstack-trove | 21:27 | |
| *** jasondotstar has quit IRC | 21:28 | |
| *** achampion has joined #openstack-trove | 21:30 | |
| *** IanGovett has quit IRC | 21:39 | |
| openstackgerrit | Riddhi Shah proposed a change to openstack/python-troveclient: Drop config group option in update instance patch call https://review.openstack.org/116369 | 21:41 |
| *** ViswaV has quit IRC | 21:41 | |
| denis_makogon | amcrn, hey, are you there ? | 21:42 |
| amcrn | denis_makogon: what's up | 21:43 |
| *** georgelorch has quit IRC | 21:51 | |
| * SlickNik reads scrollback | 21:51 | |
| *** georgelorch has joined #openstack-trove | 21:51 | |
| *** Riddhi has quit IRC | 21:51 | |
| *** Riddhi has joined #openstack-trove | 21:53 | |
| *** achampion has quit IRC | 21:54 | |
| *** georgelorch has quit IRC | 21:56 | |
| schang | SlickNik: I think https://bugs.launchpad.net/trove/+bug/1361389 can be triaged. I've put the discussions summary in the comment of the bug. | 22:02 |
| *** tkatarki has quit IRC | 22:05 | |
| *** ranjitha has quit IRC | 22:06 | |
| *** Riddhi has quit IRC | 22:08 | |
| *** tkatarki has joined #openstack-trove | 22:08 | |
| *** georgelorch has joined #openstack-trove | 22:12 | |
| *** julienvey has quit IRC | 22:13 | |
| *** harlowja_ has joined #openstack-trove | 22:13 | |
| *** Barker has quit IRC | 22:14 | |
| *** harlowja_ has quit IRC | 22:15 | |
| *** harlowja_ has joined #openstack-trove | 22:15 | |
| *** harlowja has quit IRC | 22:16 | |
| openstackgerrit | Denis M. proposed a change to openstack/trove: Support full backup/restore for Cassandra https://review.openstack.org/77551 | 22:18 |
| *** tomblank has joined #openstack-trove | 22:21 | |
| *** arist has quit IRC | 22:23 | |
| *** arist has joined #openstack-trove | 22:23 | |
| *** ViswaV has joined #openstack-trove | 22:30 | |
| *** ViswaV_ has joined #openstack-trove | 22:31 | |
| denis_makogon | amrith, hey, could you please address my comment here https://review.openstack.org/#/c/113155/3 ? | 22:34 |
| *** ViswaV has quit IRC | 22:34 | |
| *** sriram_tesora has joined #openstack-trove | 22:35 | |
| *** tkatarki has quit IRC | 22:36 | |
| *** jcru has quit IRC | 22:37 | |
| *** tkatarki has joined #openstack-trove | 22:37 | |
| *** denis_makogon has quit IRC | 22:42 | |
| *** tkatarki has quit IRC | 22:43 | |
| *** tkatarki has joined #openstack-trove | 22:45 | |
| *** jasonb365 has quit IRC | 22:50 | |
| *** kevinconway has quit IRC | 22:50 | |
| *** tkatarki has quit IRC | 23:01 | |
| *** tkatarki has joined #openstack-trove | 23:01 | |
| *** tkatarki has quit IRC | 23:04 | |
| *** mattgriffin has quit IRC | 23:04 | |
| openstackgerrit | Mat Lowery proposed a change to openstack/trove: Clusters TaskManager Implementation https://review.openstack.org/113692 | 23:05 |
| *** Riddhi has joined #openstack-trove | 23:06 | |
| openstackgerrit | Ramashri Umale proposed a change to openstack/trove: Clusters Guest Implementation https://review.openstack.org/113728 | 23:10 |
| *** tkatarki has joined #openstack-trove | 23:11 | |
| *** tkatarki has quit IRC | 23:18 | |
| *** edmund has quit IRC | 23:18 | |
| *** tkatarki has joined #openstack-trove | 23:21 | |
| *** tkatarki has quit IRC | 23:24 | |
| *** tkatarki has joined #openstack-trove | 23:25 | |
| *** tkatarki has quit IRC | 23:25 | |
| openstackgerrit | Simon Chang proposed a change to openstack/trove: Test tables existence on v20 upgrade https://review.openstack.org/117291 | 23:31 |
| *** jmontemayor has quit IRC | 23:39 | |
| *** juantwo has joined #openstack-trove | 23:45 | |
| *** RicoLin has joined #openstack-trove | 23:50 | |
| *** ViswaV_ has quit IRC | 23:53 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Clusters API Implementation https://review.openstack.org/113675 | 23:54 |
| *** esp has left #openstack-trove | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!