| *** IanGovett has quit IRC | 00:16 | |
| *** ashleighfarnham has quit IRC | 00:23 | |
| *** ashleighfarnham has joined #openstack-trove | 00:23 | |
| *** ashleighfarnham has quit IRC | 00:28 | |
| *** amrith is now known as _amrith_ | 00:45 | |
| *** _amrith_ is now known as amrith | 00:50 | |
| *** exploreshaifali has quit IRC | 01:06 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove-specs: Add backup and restore to the Redis datastore https://review.openstack.org/182910 | 01:14 |
|---|---|---|
| *** erkules_ has joined #openstack-trove | 01:22 | |
| *** erkules has quit IRC | 01:25 | |
| *** Longgeek has quit IRC | 01:38 | |
| amrith | sharikapongubala, sorry I just saw this. maybe we can chat in the morning. I agree with SlickNik's comment from 1730. if there's an element which handles the variable (datastore independent) I would be fine with one variable. It could decide what to do; for a file, for a directory, for a URL. | 02:02 |
| *** eghobo has quit IRC | 02:16 | |
| *** Longgeek has joined #openstack-trove | 02:28 | |
| *** Longgeek has quit IRC | 02:41 | |
| *** Longgeek_ has joined #openstack-trove | 02:41 | |
| *** bhunter71 has quit IRC | 02:53 | |
| *** Longgeek_ has quit IRC | 03:04 | |
| *** Longgeek has joined #openstack-trove | 03:06 | |
| *** sgotliv has joined #openstack-trove | 03:26 | |
| *** sgotliv has quit IRC | 03:31 | |
| *** coolsvap|afk is now known as coolsvap | 03:42 | |
| *** sgotliv has joined #openstack-trove | 03:46 | |
| *** Longgeek has quit IRC | 03:50 | |
| *** Longgeek has joined #openstack-trove | 03:55 | |
| *** Longgeek has quit IRC | 04:00 | |
| *** eghobo has joined #openstack-trove | 04:04 | |
| *** eghobo_ has joined #openstack-trove | 04:07 | |
| *** eghobo has quit IRC | 04:10 | |
| *** eghobo has joined #openstack-trove | 04:39 | |
| *** eghobo_ has quit IRC | 04:40 | |
| *** Longgeek has joined #openstack-trove | 04:57 | |
| *** WonChon has quit IRC | 04:59 | |
| *** WonChon has joined #openstack-trove | 05:02 | |
| *** Longgeek has quit IRC | 05:08 | |
| *** Longgeek has joined #openstack-trove | 06:05 | |
| *** masco has joined #openstack-trove | 06:05 | |
| *** Longgeek has quit IRC | 06:22 | |
| *** WonChon has quit IRC | 06:29 | |
| *** WonChon has joined #openstack-trove | 06:33 | |
| *** eghobo has quit IRC | 06:40 | |
| *** john_f_ has quit IRC | 07:40 | |
| *** john_f_ has joined #openstack-trove | 07:41 | |
| *** Longgeek has joined #openstack-trove | 07:49 | |
| *** k4n0 has joined #openstack-trove | 07:51 | |
| *** Longgeek has quit IRC | 08:00 | |
| *** links has joined #openstack-trove | 08:02 | |
| *** exploreshaifali has joined #openstack-trove | 08:18 | |
| openstackgerrit | Merged openstack/python-troveclient: Updated coverage related options to project https://review.openstack.org/181365 | 09:27 |
| *** Longgeek has joined #openstack-trove | 09:36 | |
| *** tosky has joined #openstack-trove | 09:38 | |
| *** Longgeek has quit IRC | 09:44 | |
| *** IanGovett has joined #openstack-trove | 09:46 | |
| *** haomaiwang has joined #openstack-trove | 10:42 | |
| *** haomaiwa_ has joined #openstack-trove | 10:45 | |
| *** haomaiwang has quit IRC | 10:45 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove: Use datastore specific timeouts if they exists https://review.openstack.org/164640 | 10:58 |
| *** haomaiwa_ has quit IRC | 11:15 | |
| *** Longgeek has joined #openstack-trove | 11:29 | |
| *** chlong has joined #openstack-trove | 11:48 | |
| *** Longgeek has quit IRC | 12:01 | |
| *** exploreshaifali has quit IRC | 12:04 | |
| *** pmackinn has joined #openstack-trove | 12:07 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove: Use datastore specific timeouts if they exists https://review.openstack.org/164640 | 12:15 |
| *** coolsvap is now known as coolsvap|afk | 12:22 | |
| *** bhunter71 has joined #openstack-trove | 12:29 | |
| *** masco has quit IRC | 12:32 | |
| openstackgerrit | Sushil Kumar proposed openstack/python-troveclient: Added more unit-tests for troveclient/client.py https://review.openstack.org/182771 | 12:51 |
| *** radez_g0n3 is now known as radez | 12:59 | |
| openstackgerrit | Merged openstack/trove: Update modules to turn on dangling mock detection https://review.openstack.org/180757 | 13:06 |
| *** Longgeek has joined #openstack-trove | 13:15 | |
| *** masco has joined #openstack-trove | 13:19 | |
| *** bhunter71 has quit IRC | 13:25 | |
| *** IanGovett has quit IRC | 13:25 | |
| *** IanGovett has joined #openstack-trove | 13:26 | |
| *** IanGovett1 has joined #openstack-trove | 13:40 | |
| *** IanGovett has quit IRC | 13:43 | |
| *** Longgeek has quit IRC | 13:48 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove: Added more unit-tests to Vertica-Cluster-Strategy https://review.openstack.org/178236 | 14:10 |
| *** masco has quit IRC | 14:20 | |
| *** k4n0 has quit IRC | 14:26 | |
| *** shakamunyi has joined #openstack-trove | 14:29 | |
| *** shayneburgess_ has quit IRC | 14:37 | |
| *** jodah has quit IRC | 14:37 | |
| *** jodah has joined #openstack-trove | 14:39 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove: Fix leaked mocks in the 'guestagent/test_api' module https://review.openstack.org/183067 | 14:41 |
| *** shayneburgess_ has joined #openstack-trove | 14:41 | |
| *** sriram_tesora has joined #openstack-trove | 14:42 | |
| *** barra204 has joined #openstack-trove | 14:43 | |
| *** bhunter71 has joined #openstack-trove | 14:45 | |
| *** Longgeek has joined #openstack-trove | 14:45 | |
| *** pmackinn_ has joined #openstack-trove | 15:00 | |
| *** pmackinn has quit IRC | 15:04 | |
| *** Longgeek has quit IRC | 15:04 | |
| *** Longgeek_ has joined #openstack-trove | 15:05 | |
| *** Longgeek_ has quit IRC | 15:05 | |
| *** Longgeek has joined #openstack-trove | 15:05 | |
| *** erkules_ is now known as erkules | 15:20 | |
| *** erkules has joined #openstack-trove | 15:20 | |
| *** links has quit IRC | 15:24 | |
| peterstac | SlickNik, quick question about the mock detector consolidation ... | 15:25 |
| *** haomaiwang has joined #openstack-trove | 15:28 | |
| *** haomaiwang has quit IRC | 15:29 | |
| *** haomaiwang has joined #openstack-trove | 15:29 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove: Fixed the unmocked entry in taskmanager unit-tests https://review.openstack.org/183084 | 15:35 |
| openstackgerrit | Sushil Kumar proposed openstack/trove: Fix leaked mocks in the 'guestagent/test_api' module https://review.openstack.org/183067 | 15:43 |
| *** amrith is now known as _amrith_ | 15:59 | |
| *** pmackinn_ has quit IRC | 16:13 | |
| openstackgerrit | Sushil Kumar proposed openstack/python-troveclient: Added more unit-tests for troveclient/client.py https://review.openstack.org/182771 | 16:15 |
| openstackgerrit | Sushil Kumar proposed openstack/python-troveclient: Added more unit-tests for troveclient/client.py https://review.openstack.org/182771 | 16:24 |
| SlickNik | peterstac: Around? Just saw your note… | 16:25 |
| peterstac | SlickNik: Just wanted some clarification - I was in the middle of consolidating the mock stuff over the weekend when I noticed that you had already done it | 16:27 |
| peterstac | So I abandoned my patchsets | 16:27 |
| peterstac | Now I notice patchsets going up just fixing one file again, so I'm confused ... | 16:27 |
| *** shakamunyi has quit IRC | 16:27 | |
| *** barra204 has quit IRC | 16:28 | |
| peterstac | What's our strategy going forward? Are we still using the etherpad? https://etherpad.openstack.org/p/mock-leak-cleanup | 16:29 |
| peterstac | Or are we just fixing them only when we find an issue in a file? | 16:29 |
| SlickNik | I haven't had a chance to look into reviews since last evening, but I'd prefer them to be consolidated if they're just enabling the detector by changing the base class. | 16:30 |
| SlickNik | Otherwise we are just going to have a proliferation of patches. | 16:31 |
| SlickNik | And many commits with just 1/2 line changes for the same thing. | 16:31 |
| sushilkm | can we have standalone bugfixes if it is more than just enabling the detector and actually fixes bad mocks | 16:32 |
| peterstac | so how do we keep track of what's done and what isn't? | 16:32 |
| *** ashleighfarnham has joined #openstack-trove | 16:32 | |
| peterstac | sushilkm: It will generate lots more patchsets that way - doing it file-by-file | 16:34 |
| sushilkm | i think bugfixes should still go in but if it is just enabling the detector they should be merged into one big patchset, it would also help in reviewing the fixes | 16:36 |
| peterstac | sushilkm, SlickNik: I'm ok doing it on a file-by-file basis (only if changes need to be made), just think it'll be hard to track when the work is done | 16:36 |
| peterstac | sushilkm, the problem is that you don't know if there are any fixes until you add the base class | 16:37 |
| sushilkm | as u can see in my current patchset once you are fixing the problematic mocked entry it could lead to some other problem somewhere like i faced in my case when i fixed guestagent/test_api it gave me problems in upgrade tests | 16:37 |
| sushilkm | which would otherwise seem totally unrelated | 16:37 |
| sushilkm | yes so one can add it and see if any bugfixes are there push it up, else keep it in separate big patchset | 16:38 |
| peterstac | SlickNik, sushilkm: ok, so I'll skip any files that just add the detector | 16:39 |
| *** masco has joined #openstack-trove | 16:39 | |
| sushilkm | and they can be put up separately as one big patchset in one go | 16:40 |
| peterstac | sure, as long as someone is keeping track of them | 16:40 |
| sushilkm | :) | 16:41 |
| *** shayneburgess has quit IRC | 16:41 | |
| SlickNik | Frankly, I'd prefer it all done as part of one patchset (you can still use the wiki to track whose doing what). | 16:41 |
| peterstac | SlickNik: you mean all the mock changes in one patchset? | 16:42 |
| SlickNik | s/wiki/etherpad | 16:42 |
| vkmc | that sounds like a huge patchset | 16:42 |
| vkmc | harder to review and probably painful to be free of merge conflicts | 16:43 |
| pmalik | <SlickNik> My understanding is that all with change in just the base class go into a single patch set and the other ones are done by module. Is that what you mean? | 16:43 |
| sushilkm | +1 pmalik | 16:43 |
| SlickNik | vkmc: Yes, but so is something like https://review.openstack.org/#/c/182008, but it isn't that hard to review or merge. | 16:44 |
| SlickNik | just big :) | 16:44 |
| vkmc | SlickNik, certainly | 16:44 |
| SlickNik | But very mechanical | 16:45 |
| pmalik | How can multiple ppl then work on that single patch set? | 16:45 |
| SlickNik | pmalik: Same way how multiple people work on single patchsets today. | 16:46 |
| SlickNik | But I digress — I'm okay with the compromise I suggested earlier. | 16:46 |
| pmalik | The point of doing stuff by module is that ppl working on it do not step on each other's toes. | 16:46 |
| pmalik | What compromise? | 16:47 |
| SlickNik | Where we can eagerly fix / merge patchsets that actually have fixes to the mocks. | 16:47 |
| SlickNik | If they aren't fixing anything, then lump them into a big change that enables the base class switch out | 16:48 |
| pmalik | This is not as mechanical as reordering the imports. Keep in mind some files contain significant changes. You have to ensure they do not get overwritten by other ppl's changes. | 16:48 |
| SlickNik | If the modules are different people will not step on each others' toes. | 16:48 |
| SlickNik | If the modules are not different people will step on each others' toes regardless of whether it is one commit or multiple commits. | 16:48 |
| *** eghobo has joined #openstack-trove | 16:49 | |
| SlickNik | The only thing that I see that's preventing folks from stepping on each others' toes is the etherpad, which you'd probably want to use in either case. | 16:50 |
| pmalik | No, it's git and gerrit that issues a merge conflict. | 16:50 |
| *** pmackinn_ has joined #openstack-trove | 16:51 | |
| SlickNik | Right, if you were working on the same modules (which isn't the case here) git and gerrit would give you merge conflicts that you'd have to fix before pushing up your changes in either case. | 16:52 |
| pmalik | I still don't understand how you want to sync work from multiple ppl into a single patch set without using gerrit == having a separate commit/patch set for each module. | 16:52 |
| pmalik | Also remember that that single huge patch set will then conflict with every other incoming patch set. Having separate patch sets for modules makes the merges and conflicts more granural | 16:55 |
| *** masco has quit IRC | 16:57 | |
| SlickNik | pmalik: It's not that hard to fix merge conflicts when a patch is ready to merge. We've merged patches that are bigger than this one before. Plus folks don't usually go fixing unit tests all over the place so merge conflicts should be minimal in this case. | 16:58 |
| SlickNik | You should _not_ be changing trove code for this anyhow. | 16:58 |
| pmalik | Also by having multiple ppl commiting into a single patch set you offload the whole conflict and merge management on them rather than having gerrit/git doing it. Keep in mind that those packages are not 100% independent. Some fixes end up requiring changes to other packages too. That's what I am trying to say. This is not as mechanical as with reordering imports (unless the only change is to the base class and there is n | 16:58 |
| pmalik | othing else to fix). | 16:58 |
| SlickNik | pmalik: if there are merge conflicts — a human has to resolve them in either case. Gerrit will only tell you that you have to do it. | 17:03 |
| SlickNik | And in this case if folks are working per module, there ideally shouldn't be merge conflicts. | 17:04 |
| *** _crobertsrh is now known as crobertsrh | 17:09 | |
| pmalik | <SlickNik> As can be seen from https://review.openstack.org/#/c/183067/ it's not always the case though... ;-) | 17:11 |
| *** shayneburgess has joined #openstack-trove | 17:26 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove: Added dangling mock detection to 'guestagent' https://review.openstack.org/180825 | 17:31 |
| *** bhunter71 has quit IRC | 17:44 | |
| *** bhunter71 has joined #openstack-trove | 17:45 | |
| *** bhunter71 has quit IRC | 17:49 | |
| *** _amrith_ is now known as amrith | 17:50 | |
| *** bhunter71 has joined #openstack-trove | 17:51 | |
| *** _mattf is now known as mattf | 18:02 | |
| *** exploreshaifali has joined #openstack-trove | 18:14 | |
| *** sriram_tesora has quit IRC | 18:26 | |
| *** exploreshaifali has quit IRC | 18:30 | |
| *** IanGovett1 has quit IRC | 18:44 | |
| *** IanGovett has joined #openstack-trove | 18:51 | |
| vkmc | hey guys, is there an spec about clustering somewhere? | 18:51 |
| SlickNik | vkmc: About mysql clustering? Or about the API? | 18:52 |
| vkmc | about how we implement clustering :) | 18:52 |
| SlickNik | pmalik / peterstac: I'm trying to understand why a lot of the mock detector tests change the order of imports — do you guys know why this is needed? | 18:52 |
| openstackgerrit | Matthew Van Dijk proposed openstack/trove-specs: MongoDB database management commands https://review.openstack.org/183166 | 18:53 |
| SlickNik | vkmc: https://wiki.openstack.org/wiki/Trove/Replication-And-Clustering-With-Nodes-5 | 18:53 |
| vkmc | SlickNik, thanks! | 18:53 |
| SlickNik | This is still in wiki format since it landed in Juno before we went to the rst specs format. | 18:54 |
| peterstac | SlickNik: I don't think I changed any import order, at least not intentionally (if so I'll revert it) | 18:54 |
| vkmc | yeah, its fine :) I checked the code but I was looking something more docs-ish | 18:55 |
| SlickNik | peterstac: I don't care about import order :) — just wanted to know if there's a reason we need to do this intentionally. | 18:55 |
| peterstac | SlickNik: no, not that I'm aware of :) | 18:56 |
| SlickNik | sounds good, thanks! | 18:57 |
| pmalik | I guess the import order is now better handled by https://review.openstack.org/#/c/181950/ (once it merges) | 18:58 |
| *** IanGovett1 has joined #openstack-trove | 18:58 | |
| *** IanGovett has quit IRC | 19:01 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove: Added dangling mock detection to 'guestagent' https://review.openstack.org/180825 | 19:01 |
| *** jeraldv has joined #openstack-trove | 19:02 | |
| openstackgerrit | Saurabh Surana proposed openstack/trove: accepting network and availability zone for instances in cluster https://review.openstack.org/179443 | 19:07 |
| *** eghobo has quit IRC | 19:41 | |
| *** radez is now known as radez_g0n3 | 19:49 | |
| *** openstackgerrit has quit IRC | 19:52 | |
| *** openstackgerrit has joined #openstack-trove | 19:52 | |
| *** amrith is now known as _amrith_ | 20:13 | |
| vkmc | guys, is there a roadmap? | 20:21 |
| vkmc | the main wiki seems outdated | 20:21 |
| *** bhunter71 has quit IRC | 20:23 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove-specs: Trove Instance Metadata https://review.openstack.org/175042 | 20:25 |
| *** eghobo has joined #openstack-trove | 20:27 | |
| *** georgelorch has quit IRC | 20:29 | |
| openstackgerrit | Mayuri Ganguly proposed openstack/trove: Associate flavor types with datastore versions https://review.openstack.org/109824 | 20:30 |
| *** JamesK has joined #openstack-trove | 20:32 | |
| *** mattf is now known as _mattf | 20:33 | |
| vkmc | dougshelley66, pmalik, peterstac, do somebody know if we have an updated roadmap? | 20:34 |
| *** JamesK has quit IRC | 20:34 | |
| dougshelley66 | vkmc, not that i know of. It was discussed at mid-cycle and a couple of folks took an action item to work towards one. How far out would you expect the roadmap to go? | 20:36 |
| vkmc | hmm, at least L-1/L-2? | 20:37 |
| dougshelley66 | i think the first Trove Work Session at summit is to talk about liberty blueprints | 20:47 |
| dougshelley66 | which i assume would cover L-1 and L-2 | 20:47 |
| vkmc | ok :) | 20:48 |
| vkmc | being so close of the summit | 20:48 |
| vkmc | I guess we could work on creating the roadmap for then | 20:48 |
| vkmc | I mean, after those sessions | 20:48 |
| *** sgotliv has quit IRC | 20:53 | |
| *** JamesK has joined #openstack-trove | 20:54 | |
| *** georgelorch has joined #openstack-trove | 20:56 | |
| *** georgelorch has quit IRC | 21:06 | |
| *** haomaiwang has quit IRC | 21:08 | |
| *** chlong has quit IRC | 21:08 | |
| *** haomaiwa_ has joined #openstack-trove | 21:09 | |
| *** crobertsrh is now known as _crobertsrh | 21:13 | |
| *** openstackgerrit has quit IRC | 21:22 | |
| *** openstackgerrit has joined #openstack-trove | 21:22 | |
| *** sgotliv has joined #openstack-trove | 21:28 | |
| *** _amrith_ is now known as amrith | 22:02 | |
| *** eghobo has quit IRC | 22:09 | |
| *** eghobo has joined #openstack-trove | 22:10 | |
| *** bhunter71 has joined #openstack-trove | 22:11 | |
| *** georgelorch has joined #openstack-trove | 22:14 | |
| *** sgotliv has quit IRC | 22:30 | |
| *** tosky has quit IRC | 22:35 | |
| *** shakamunyi has joined #openstack-trove | 22:47 | |
| *** barra204 has joined #openstack-trove | 22:47 | |
| *** pmackinn_ has quit IRC | 23:00 | |
| *** ashleighfarnham has quit IRC | 23:00 | |
| *** pmackinn_ has joined #openstack-trove | 23:01 | |
| *** barra204 has quit IRC | 23:04 | |
| *** shakamunyi has quit IRC | 23:04 | |
| *** pmackinn_ has quit IRC | 23:05 | |
| *** ashleighfarnham has joined #openstack-trove | 23:07 | |
| *** chlong has joined #openstack-trove | 23:10 | |
| *** shakamunyi has joined #openstack-trove | 23:17 | |
| *** barra204 has joined #openstack-trove | 23:17 | |
| *** cp16net has quit IRC | 23:17 | |
| *** Longgeek has quit IRC | 23:30 | |
| *** Longgeek has joined #openstack-trove | 23:30 | |
| amrith | SlickNik, yt? | 23:36 |
| *** ashleighfarnham has quit IRC | 23:48 | |
| *** shayneburgess has quit IRC | 23:49 | |
| openstackgerrit | Mayuri Ganguly proposed openstack/trove: Changing cluster status to none when instances are in error state https://review.openstack.org/183359 | 23:52 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!