| *** mattgriffin has quit IRC | 00:09 | |
| *** juantwo has quit IRC | 00:15 | |
| *** harlowja_ has quit IRC | 00:25 | |
| *** harlowja has joined #openstack-trove | 00:28 | |
| *** juantwo has joined #openstack-trove | 00:29 | |
| *** juantwo has quit IRC | 00:29 | |
| *** juantwo has joined #openstack-trove | 00:29 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-troveclient: Updated from global requirements https://review.openstack.org/120610 | 00:31 |
|---|---|---|
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/trove: Updated from global requirements https://review.openstack.org/119818 | 00:32 |
| *** ViswaV has quit IRC | 00:35 | |
| *** mattgriffin has joined #openstack-trove | 00:35 | |
| *** Riddhi has joined #openstack-trove | 00:37 | |
| *** IanGovett has quit IRC | 01:02 | |
| *** newb_ has quit IRC | 01:17 | |
| *** saurabhs has left #openstack-trove | 01:26 | |
| *** juantwo has quit IRC | 01:33 | |
| *** grapex_ has quit IRC | 01:34 | |
| *** rhodgin has joined #openstack-trove | 01:42 | |
| openstackgerrit | Peter Stachowski proposed a change to openstack/trove: Document Trove configuration options https://review.openstack.org/118759 | 01:44 |
| *** Riddhi has quit IRC | 01:44 | |
| *** nosnos has joined #openstack-trove | 01:47 | |
| stevelle | amcrn: ty, will look at that next | 01:48 |
| *** haomaiwang has quit IRC | 01:54 | |
| *** haomaiwang has joined #openstack-trove | 01:54 | |
| *** juantwo has joined #openstack-trove | 01:56 | |
| *** juantwo has quit IRC | 01:58 | |
| *** juantwo has joined #openstack-trove | 01:58 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Custom Flavors For Config Servers and Routers https://review.openstack.org/120672 | 02:03 |
| *** jasonb365 has joined #openstack-trove | 02:04 | |
| *** ramashri has quit IRC | 02:10 | |
| *** haomaiwang has quit IRC | 02:12 | |
| *** haomaiwang has joined #openstack-trove | 02:13 | |
| amcrn | stevelle: i haven't tested this, but see https://ask.openstack.org/en/question/608/launching-instance-and-volume/?answer=673#post-id-673 | 02:21 |
| amcrn | bbiab | 02:22 |
| *** amcrn has quit IRC | 02:22 | |
| *** johnma has quit IRC | 02:38 | |
| *** vkmc has quit IRC | 02:40 | |
| *** rushiagr_away is now known as rushiagr | 02:53 | |
| *** ramishra has joined #openstack-trove | 03:04 | |
| *** rushiagr is now known as rushiagr_away | 03:14 | |
| *** mrsnivvel has joined #openstack-trove | 03:20 | |
| *** Longgeek has joined #openstack-trove | 03:29 | |
| *** Longgeek_ has joined #openstack-trove | 03:32 | |
| *** Longgeek has quit IRC | 03:36 | |
| *** rushiagr_away is now known as rushiagr | 03:46 | |
| *** haomaiwang has quit IRC | 03:54 | |
| *** haomaiwa_ has joined #openstack-trove | 03:54 | |
| *** ramishra has quit IRC | 03:56 | |
| *** ramishra has joined #openstack-trove | 03:57 | |
| *** ramishra has quit IRC | 03:58 | |
| *** ramishra has joined #openstack-trove | 03:58 | |
| *** haomai___ has joined #openstack-trove | 04:00 | |
| *** haomai___ has quit IRC | 04:00 | |
| *** haomaiw__ has joined #openstack-trove | 04:01 | |
| *** ramashri has joined #openstack-trove | 04:01 | |
| *** haomaiwa_ has quit IRC | 04:02 | |
| *** harlowja is now known as harlowja_away | 04:05 | |
| *** haomaiwang has joined #openstack-trove | 04:06 | |
| *** ramashri has quit IRC | 04:08 | |
| *** ramashri has joined #openstack-trove | 04:09 | |
| *** sgotliv has joined #openstack-trove | 04:10 | |
| *** haomaiw__ has quit IRC | 04:10 | |
| *** jasonb365 has quit IRC | 04:19 | |
| *** ramishra has quit IRC | 04:23 | |
| *** ramishra has joined #openstack-trove | 04:24 | |
| *** mattgriffin has quit IRC | 04:31 | |
| *** johnma has joined #openstack-trove | 04:32 | |
| sgotliv | SlickNik, are you still awake? | 05:12 |
| SlickNik | sgotliv: Yes, still around | 05:12 |
| SlickNik | What's up? | 05:12 |
| sgotliv | Can you take a look on the RPC patch | 05:12 |
| sgotliv | sending a link | 05:13 |
| sgotliv | https://review.openstack.org/#/c/94484/ | 05:13 |
| sgotliv | It seems like there is a delay of an hour in the end of the log.... | 05:13 |
| sgotliv | which I can't explain | 05:13 |
| sgotliv | and therefore build is marked as failed | 05:14 |
| sgotliv | although all running tests were skipped or passed | 05:14 |
| sgotliv | pep8 thing is on me :-) | 05:14 |
| SlickNik | sgotliv — will need to look into why the unit tests are not running. | 05:18 |
| SlickNik | sgotliv: Looks like the fake mode tests ran correctly, but the unit tests that run using testr timed out. | 05:18 |
| sgotliv | ok, let me check, go to rest | 05:19 |
| sgotliv | thanks for explanation | 05:19 |
| SlickNik | sgotliv: This seems like the relevant part: | 05:19 |
| SlickNik | http://paste.openstack.org/show/109943/ | 05:19 |
| SlickNik | sgotliv: unit tests are run using: "python setup.py testr --slowest" | 05:20 |
| sgotliv | I saw it, I just didn't know about the second part of tests | 05:20 |
| SlickNik | And for some reason, it looks like that is hanging | 05:20 |
| SlickNik | causing "Build timed out (after 50 minutes). Marking the build as failed." | 05:20 |
| sgotliv | no problem, I'll resolve it, now when you explained me that | 05:20 |
| sgotliv | right | 05:20 |
| SlickNik | Cool, thanks so much for looking at this! | 05:21 |
| sgotliv | This is a good and only way to learn quick | 05:21 |
| sgotliv | and thanks for the explanation | 05:21 |
| SlickNik | Not a problem — I hope to start reviewing the code tomorrow (and will encourage other folks to start looking at it as well). | 05:23 |
| SlickNik | brb | 05:23 |
| sgotliv | no problem, I hope to make it ready till then | 05:24 |
| sgotliv | but review should start anyway | 05:24 |
| sgotliv | because its important to discover potential issues as soon as possible | 05:24 |
| *** ajayaa has joined #openstack-trove | 05:32 | |
| *** amcrn has joined #openstack-trove | 05:44 | |
| *** flaper87|afk is now known as flaper87 | 05:58 | |
| *** vigneshvar has joined #openstack-trove | 05:59 | |
| *** prasoon has joined #openstack-trove | 06:00 | |
| *** vigneshvar_ has joined #openstack-trove | 06:01 | |
| *** vigneshvar_ has quit IRC | 06:03 | |
| *** ramishra has quit IRC | 06:09 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/trove: Imported Translations from Transifex https://review.openstack.org/117695 | 06:09 |
| *** achampion has quit IRC | 06:11 | |
| *** Longgeek_ has quit IRC | 06:17 | |
| *** Longgeek has joined #openstack-trove | 06:18 | |
| *** Longgeek has quit IRC | 06:20 | |
| *** ramishra has joined #openstack-trove | 06:23 | |
| *** ramishra has quit IRC | 06:27 | |
| *** ramishra has joined #openstack-trove | 06:27 | |
| *** ViswaV has joined #openstack-trove | 06:31 | |
| *** ViswaV_ has joined #openstack-trove | 06:33 | |
| *** k4n0 has joined #openstack-trove | 06:34 | |
| *** ViswaV has quit IRC | 06:36 | |
| *** ramashri has quit IRC | 06:40 | |
| *** ramashri has joined #openstack-trove | 06:40 | |
| *** Longgeek has joined #openstack-trove | 06:52 | |
| *** ramashri has quit IRC | 06:57 | |
| *** ramishra has quit IRC | 06:58 | |
| *** ramashri has joined #openstack-trove | 07:02 | |
| *** haomaiwang has quit IRC | 07:10 | |
| *** haomaiwa_ has joined #openstack-trove | 07:10 | |
| *** todd_dsm has joined #openstack-trove | 07:12 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Custom Flavors For Config Servers and Routers https://review.openstack.org/120672 | 07:14 |
| *** ramashri has quit IRC | 07:15 | |
| *** haomaiwa_ has quit IRC | 07:17 | |
| *** haomaiw__ has joined #openstack-trove | 07:17 | |
| *** amcrn has quit IRC | 07:23 | |
| openstackgerrit | A change was merged to openstack/trove: Register postgres_group in trove config https://review.openstack.org/120265 | 07:25 |
| *** haomaiwang has joined #openstack-trove | 07:31 | |
| *** ramishra has joined #openstack-trove | 07:31 | |
| *** ramishra has quit IRC | 07:32 | |
| *** ramishra has joined #openstack-trove | 07:33 | |
| *** haomaiw__ has quit IRC | 07:35 | |
| *** boblebauce has joined #openstack-trove | 08:06 | |
| *** vigneshvar has quit IRC | 08:11 | |
| sgotliv | cp16net, ping | 08:14 |
| *** juantwo_ has joined #openstack-trove | 08:16 | |
| *** juantwo has quit IRC | 08:20 | |
| *** vigneshvar has joined #openstack-trove | 08:24 | |
| *** julienvey has joined #openstack-trove | 08:29 | |
| *** iartarisi has joined #openstack-trove | 08:33 | |
| openstackgerrit | Sergey Gotliv proposed a change to openstack/trove: Updates RPC API to use oslo.messaging https://review.openstack.org/94484 | 08:41 |
| *** sunrenjie6 has joined #openstack-trove | 08:46 | |
| *** ViswaV_ has quit IRC | 08:48 | |
| openstackgerrit | Ionut Artarisi proposed a change to openstack/trove: Add support to detect SUSE https://review.openstack.org/108394 | 08:58 |
| openstackgerrit | Ionut Artarisi proposed a change to openstack/trove: Add SUSE support in mysql datastore https://review.openstack.org/108703 | 08:58 |
| openstackgerrit | Ionut Artarisi proposed a change to openstack/trove: Add support to detect SUSE https://review.openstack.org/108394 | 08:58 |
| openstackgerrit | Ionut Artarisi proposed a change to openstack/trove: Add SUSE support in mysql datastore https://review.openstack.org/108703 | 08:58 |
| *** ramishra has quit IRC | 09:25 | |
| *** ramishra has joined #openstack-trove | 09:25 | |
| *** ramishra has quit IRC | 09:27 | |
| *** ramishra has joined #openstack-trove | 09:30 | |
| *** sunrenjie6 has quit IRC | 09:35 | |
| *** haomaiwang has quit IRC | 09:42 | |
| *** haomaiwang has joined #openstack-trove | 09:42 | |
| openstackgerrit | A change was merged to openstack/trove: Marks mysql slave read-only https://review.openstack.org/118386 | 09:50 |
| *** haomaiwang has quit IRC | 10:02 | |
| *** haomaiwang has joined #openstack-trove | 10:03 | |
| *** d0ugal has quit IRC | 10:08 | |
| *** d0ugal has joined #openstack-trove | 10:08 | |
| *** boblebauce has quit IRC | 10:24 | |
| *** vkmc has joined #openstack-trove | 10:36 | |
| *** ramishra has quit IRC | 10:39 | |
| *** ramishra has joined #openstack-trove | 10:40 | |
| *** newb_ has joined #openstack-trove | 10:43 | |
| *** PeterSchwaller has quit IRC | 10:59 | |
| *** PeterSchwaller has joined #openstack-trove | 11:00 | |
| *** julienvey has quit IRC | 11:14 | |
| *** tomblank has quit IRC | 11:15 | |
| *** julienvey has joined #openstack-trove | 11:21 | |
| *** julienvey has quit IRC | 11:23 | |
| *** julienvey has joined #openstack-trove | 11:23 | |
| *** julienve_ has joined #openstack-trove | 11:24 | |
| *** julienvey has quit IRC | 11:24 | |
| *** prasoon has quit IRC | 11:37 | |
| *** ramishra has quit IRC | 11:39 | |
| *** IanGovett has joined #openstack-trove | 11:40 | |
| *** Riddhi has joined #openstack-trove | 11:48 | |
| *** boblebauce has joined #openstack-trove | 11:52 | |
| *** Riddhi has quit IRC | 12:03 | |
| *** Riddhi has joined #openstack-trove | 12:04 | |
| *** jcru has joined #openstack-trove | 12:10 | |
| *** nosnos has quit IRC | 12:22 | |
| *** boblebauce has quit IRC | 12:22 | |
| *** nosnos has joined #openstack-trove | 12:22 | |
| *** grapex has joined #openstack-trove | 12:27 | |
| *** nosnos has quit IRC | 12:27 | |
| *** amrith is now known as _amrith_ | 12:28 | |
| *** grapex_ has joined #openstack-trove | 12:29 | |
| *** grapex has quit IRC | 12:32 | |
| *** tomblank has joined #openstack-trove | 12:36 | |
| *** prasoon has joined #openstack-trove | 12:43 | |
| *** radez_g0n3 is now known as radez | 12:47 | |
| *** prasoon has quit IRC | 12:48 | |
| *** grapex_ has quit IRC | 12:48 | |
| openstackgerrit | Riddhi Shah proposed a change to openstack/trove: Allow users the ability to update an instance name https://review.openstack.org/92701 | 12:50 |
| *** Riddhi has quit IRC | 12:55 | |
| *** miqui has joined #openstack-trove | 12:55 | |
| *** prasoon has joined #openstack-trove | 13:01 | |
| *** achampion has joined #openstack-trove | 13:08 | |
| *** boblebauce has joined #openstack-trove | 13:09 | |
| *** ramashri has joined #openstack-trove | 13:10 | |
| *** newb_ has quit IRC | 13:12 | |
| *** Ugallu has joined #openstack-trove | 13:13 | |
| *** ramashri has quit IRC | 13:15 | |
| *** ramashri has joined #openstack-trove | 13:17 | |
| *** prasoon has quit IRC | 13:30 | |
| *** juantwo_ has quit IRC | 13:37 | |
| *** juantwo has joined #openstack-trove | 13:37 | |
| *** Longgeek has quit IRC | 13:41 | |
| *** IanGovett has quit IRC | 13:43 | |
| *** IanGovett has joined #openstack-trove | 13:43 | |
| *** _amrith_ is now known as amrith | 13:49 | |
| *** newb_ has joined #openstack-trove | 13:51 | |
| *** newb_ is now known as newb | 13:52 | |
| *** IanGovett1 has joined #openstack-trove | 13:58 | |
| *** IanGovett has quit IRC | 13:59 | |
| *** Longgeek has joined #openstack-trove | 14:05 | |
| *** rhodgin has quit IRC | 14:06 | |
| *** robertmyers has joined #openstack-trove | 14:09 | |
| *** mattgriffin has joined #openstack-trove | 14:20 | |
| *** jasonb365 has joined #openstack-trove | 14:30 | |
| *** vigneshvar has quit IRC | 14:31 | |
| *** tomblank_ has joined #openstack-trove | 14:36 | |
| *** tomblank has quit IRC | 14:38 | |
| *** tomblank_ is now known as tomblank | 14:38 | |
| openstackgerrit | Peter Stachowski proposed a change to openstack/trove: Document Trove configuration options https://review.openstack.org/118759 | 14:40 |
| *** grapex has joined #openstack-trove | 14:42 | |
| *** kevinconway has joined #openstack-trove | 14:48 | |
| *** prasoon has joined #openstack-trove | 14:49 | |
| *** rhodgin has joined #openstack-trove | 14:50 | |
| *** ajayaa has quit IRC | 14:57 | |
| *** Riddhi has joined #openstack-trove | 15:00 | |
| *** k4n0 has quit IRC | 15:01 | |
| *** Longgeek has quit IRC | 15:07 | |
| *** boblebauce has quit IRC | 15:07 | |
| *** julienve_ has quit IRC | 15:10 | |
| *** juantwo_ has joined #openstack-trove | 15:13 | |
| *** juantwo has quit IRC | 15:16 | |
| *** prasoon has quit IRC | 15:19 | |
| *** juantwo_ has quit IRC | 15:21 | |
| *** juantwo has joined #openstack-trove | 15:21 | |
| *** rhodgin has quit IRC | 15:27 | |
| *** todd_dsm has quit IRC | 15:36 | |
| *** todd_dsm has joined #openstack-trove | 15:43 | |
| *** ramashri has quit IRC | 15:48 | |
| *** ruebenramirez has joined #openstack-trove | 15:48 | |
| *** julienvey has joined #openstack-trove | 15:49 | |
| *** Barker has joined #openstack-trove | 16:03 | |
| openstackgerrit | iccha-sethi proposed a change to openstack/trove: Delete backup created during replication https://review.openstack.org/119169 | 16:09 |
| *** rushiagr is now known as rushiagr_away | 16:21 | |
| *** julienvey has quit IRC | 16:21 | |
| *** amrith is now known as _amrith_ | 16:22 | |
| *** rwsu has quit IRC | 16:24 | |
| *** PierreRambaud has quit IRC | 16:26 | |
| *** rwsu has joined #openstack-trove | 16:29 | |
| *** rwsu has quit IRC | 16:33 | |
| *** rwsu has joined #openstack-trove | 16:35 | |
| *** iartarisi has quit IRC | 16:38 | |
| *** ramashri has joined #openstack-trove | 16:39 | |
| openstackgerrit | Riddhi Shah proposed a change to openstack/trove: Allow users the ability to update an instance name https://review.openstack.org/92701 | 16:40 |
| *** rushiagr_away is now known as rushiagr | 16:46 | |
| *** ViswaV has joined #openstack-trove | 16:50 | |
| openstackgerrit | A change was merged to openstack/trove: Fix issue with intermittent test failures in test_pkg.py https://review.openstack.org/119483 | 16:50 |
| *** tomblank has quit IRC | 16:52 | |
| *** julienvey has joined #openstack-trove | 16:56 | |
| *** ViswaV has quit IRC | 17:01 | |
| *** ViswaV has joined #openstack-trove | 17:04 | |
| *** ViswaV_ has joined #openstack-trove | 17:07 | |
| *** mattgriffin has quit IRC | 17:09 | |
| *** harlowja_away is now known as harlowja | 17:09 | |
| *** ViswaV has quit IRC | 17:10 | |
| *** mattgriffin has joined #openstack-trove | 17:12 | |
| *** Barker has quit IRC | 17:14 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/trove: Updated from global requirements https://review.openstack.org/119818 | 17:21 |
| grapex | amcrn mat-lowery: I have a question about the BUILD_PENDING status | 17:24 |
| grapex | I don't understand why it was added. | 17:24 |
| mat-lowery | grapex: Before clusters, guest got the prepare message and went straight to RUNNING. With clusters, the guest can only do so much until it needs more information to complete the cluster setup. | 17:25 |
| grapex | mat-lowery: I'm looking at this commit: https://github.com/openstack/trove/commit/06196fcf67b27f0308381da192da5cc8ae65b157#diff-a999957f2c43a14679e10d319425ba3cR281 | 17:26 |
| grapex | I can't see where BUILD_PENDING is getting set | 17:26 |
| *** julienvey has quit IRC | 17:27 | |
| grapex | mat-lowery: So the mongo service needed more information then- was it running normally or stopped? | 17:27 |
| mat-lowery | Ah. Due to the way we chopped up the cluster Gerrit changes, it happens in another Gerrit change: https://github.com/openstack/trove/commit/74b436e1d2708c0137437ff1bf5b846a89bda6ed#diff-f2d2373085354cab72f3eae86a8c92deR98 | 17:27 |
| *** tomblank has joined #openstack-trove | 17:28 | |
| mat-lowery | Depending on the "role" or "type" of the guest, its service may actually be running or not running. mongos doesn't start until it gets the config server IPs | 17:29 |
| mat-lowery | but in terms of what it advertises to TM, they all say BUILD_PENDING until the TM tells them that cluster setup is done. | 17:30 |
| mat-lowery | if I remember correctly | 17:30 |
| grapex | mat-lowery: So in build pending you don't want it sending messages back to Trove for example if it can't ping mongo | 17:30 |
| grapex | since you know it's not finished yet | 17:30 |
| mat-lowery | right | 17:31 |
| *** ViswaV_ has quit IRC | 17:31 | |
| grapex | I think this should be renamed- the build is done, it's installed mongo. It's just needs some info | 17:32 |
| grapex | But that's not a big deal | 17:32 |
| grapex | So I have another question- what is the Trove instance status at this point in time? | 17:32 |
| grapex | When it's in BUILD_PENDING? | 17:32 |
| grapex | Is it ACTIVE or something else? | 17:33 |
| mat-lowery | I think it translates to BUILD. Because to the user, it's still building. It's just the TM that needs the intermediate stage. | 17:33 |
| grapex | Where is that code? | 17:34 |
| mat-lowery | Let me get a link | 17:34 |
| grapex | Oh I remember | 17:34 |
| grapex | there's the api_status property of the service status | 17:34 |
| *** ViswaV has joined #openstack-trove | 17:34 | |
| grapex | that is what gets returned | 17:34 |
| grapex | Ok, so it goes to BUILD | 17:34 |
| grapex | I think then I'm ok, my only other qualm is I feel there's a loss of encapsulation to the instance code with these changes we should address before too much time passes | 17:35 |
| grapex | Back to this: | 17:35 |
| grapex | https://github.com/openstack/trove/commit/06196fcf67b27f0308381da192da5cc8ae65b157#diff-a4d09d28bd2b650c2327f5d8d81be3a9R89 | 17:35 |
| grapex | if you look at "_all_instances_ready", it's loading up instance_ids and querying their guts from the database | 17:36 |
| grapex | it should be loading up SimpleInstances subclasses from trove.instance.models and checking the status property | 17:36 |
| grapex | if the status property can't handle this new state I think we need t a new property to describe this as what's happening is different | 17:36 |
| grapex | we also might want to think of making a new status that isn't "BUILD" but something else | 17:36 |
| grapex | mat-lowery: trove.instance.models.load_any_instance can be used to fetch the server | 17:38 |
| grapex | and the clustering code can check its properties | 17:38 |
| *** _amrith_ is now known as amrith | 17:39 | |
| *** ramashri has quit IRC | 17:39 | |
| mat-lowery | Sorry. Was on a call. "there's the api_status property of the service status" < -- yes. | 17:41 |
| mat-lowery | I came up with BUILD_PENDING. I thought it was pretty apt. :) | 17:42 |
| grapex | Ah, it's ok. That's an opinion I guess. :p | 17:42 |
| grapex | The bigger issue is how clustering keeps grabbing the guts of instance models | 17:42 |
| mat-lowery | looking | 17:44 |
| grapex | this cluster code is going to break some stuff, particularly at Rax, and since there's not a great way to test except manually that means pretty soon there will be push back on major changes. So even though the fact it touches the instance models code internals is a nitpick I'd like us to fix it right away | 17:44 |
| grapex | I'll send an email | 17:44 |
| *** openstackgerrit has quit IRC | 17:46 | |
| *** openstackgerrit has joined #openstack-trove | 17:47 | |
| *** ramashri has joined #openstack-trove | 17:48 | |
| *** ViswaV has quit IRC | 17:48 | |
| mat-lowery | I don't understand the "break some stuff" but I'll wait for the email. Thanks. | 17:48 |
| grapex | mat-lowery: sent | 17:52 |
| mat-lowery | grapex: Thanks. | 17:52 |
| * SlickNik reads scrollback | 17:56 | |
| grapex | SlickNik: My email summarized it better. :p | 17:58 |
| SlickNik | grapex: yup, just looked at email as well — I agree with you that we should be checking the status using the SimpleInstance abstraction rather than going directly to the database. | 18:01 |
| SlickNik | hopefully this is something we can get fixed before the RC> | 18:01 |
| grapex | SlickNik: Yeah. My main fear is since there's no way to automatically test clustering, in the future if we try to change it well be too scared | 18:02 |
| grapex | however | 18:02 |
| grapex | I think I can make it possible to test clustering in fake mode (i.e. the run_tests.py) | 18:02 |
| grapex | It'll take me awhile | 18:02 |
| *** ViswaV has joined #openstack-trove | 18:02 | |
| grapex | mat-lowery ran into the nasty side of the event simulator. He's the first person so far to see just how fake it is. :) | 18:02 |
| grapex | But hopefully next week I can make it possible to run some kind of test with clustering in quick tests mode at least | 18:02 |
| *** Barker has joined #openstack-trove | 18:03 | |
| mat-lowery | grapex: It's on my TODO list to get back with you about cluster int tests and fake mode. Do you have a rough plan on your changes you can share that will allow fake mode to work with clusters? | 18:05 |
| grapex | mat-lowery: yeah- the rough draft is I'm going to change the event simulator to actually use threads but only let one thing run at a time | 18:07 |
| grapex | that way we get everything to run in the same determinisitic way, regardless of who runs it | 18:07 |
| grapex | but are actually making use of eventlet coroutines so things you would think of as threads actually get their own stacks | 18:07 |
| grapex | and we avoid the problem you hit | 18:07 |
| grapex | I worked on it a few hours last week but got confused | 18:08 |
| grapex | I think sometime soon though it'll click and I'll be able to submit something that's somewhat clean | 18:08 |
| grapex | and will be an improvement over what we have today | 18:08 |
| *** rushiagr is now known as rushiagr_away | 18:09 | |
| mat-lowery | grapex: Sweet. Feel free to share anything you have in any state. And let me know how I can help. | 18:10 |
| grapex | mat-lowery: will do. Thanks | 18:10 |
| grapex | and sorry you ran into that issue- it's the second time it ever bit anybody (the first time was me) | 18:10 |
| amrith | Riddhi, yt? | 18:10 |
| mat-lowery | grapex: np. learned a lot | 18:11 |
| amrith | Riddhi, when you get a chance, I would like to chat with you about 92701. | 18:11 |
| amrith | please ping me when you have a second. thx | 18:11 |
| amrith | esp ... have a second for a quick consult on oslo.messaging? | 18:12 |
| *** rushiagr_away is now known as rushiagr | 18:14 | |
| *** ViswaV has quit IRC | 18:22 | |
| *** Barker has quit IRC | 18:22 | |
| *** ViswaV has joined #openstack-trove | 18:23 | |
| Riddhi | hey amrith | 18:26 |
| Riddhi | yt? | 18:27 |
| amrith | yes | 18:27 |
| amrith | hi Riddhi | 18:27 |
| amrith | i had a couple of comments inline in the review | 18:27 |
| amrith | the one I wanted to speak with you about is this one | 18:27 |
| Riddhi | yeah, so sorry, the blueprint in the patch isint pointing to the right one | 18:28 |
| Riddhi | i guess the name changes | 18:28 |
| Riddhi | *changed | 18:28 |
| *** vigneshvar has joined #openstack-trove | 18:28 | |
| Riddhi | its this one: https://blueprints.launchpad.net/trove/+spec/update-instance-name | 18:28 |
| amrith | ok, thanks. I found another old one which we should probably update/delete/something else | 18:29 |
| amrith | I will read this new BP. | 18:29 |
| amrith | I'm curious about the change in instance/models.py and taskmanager/models.py where the call to update_db() has been removed. why is that? | 18:30 |
| openstackgerrit | Morgan Jones proposed a change to openstack/trove: Use unique passwords for replication user https://review.openstack.org/120874 | 18:30 |
| Riddhi | ahh..yeah, so in if you see here - https://review.openstack.org/#/c/92701/23/trove/instance/service.py | 18:31 |
| * amrith looking | 18:31 | |
| Riddhi | we do the update db in the methods itsle | 18:31 |
| Riddhi | *itself | 18:31 |
| * iccha1 looking at marith looking | 18:32 | |
| iccha1 | *amrith | 18:32 |
| * amrith feels self-conscious and strats fnmuling kbd | 18:32 | |
| amrith | fumbling | 18:32 |
| Riddhi | :P | 18:32 |
| iccha1 | amrith: so its in task manager so we do the delete only once | 18:32 |
| *** amcrn has joined #openstack-trove | 18:32 | |
| openstackgerrit | Ramashri Umale proposed a change to openstack/trove: Strengthens the regex for mongodb json functions https://review.openstack.org/120877 | 18:33 |
| Riddhi | so that we can pass the update args (config , name) together | 18:33 |
| *** ajayaa has joined #openstack-trove | 18:37 | |
| * amrith is puzzled | 18:37 | |
| *** Ugallu has quit IRC | 18:39 | |
| amrith | so, if I understand this correctly, the intent behind this change is the following | 18:40 |
| amrith | if you did the update_db() in assign_configuration() and unassign_configuration() there would still need to be (potentially) one more update_db for the name change | 18:41 |
| amrith | and to save on the one update for the name change (in edit) we moved the update_db() calls into edit and update instead of remaining in the model? | 18:41 |
| amrith | is that a fair (but uncharitable) description? | 18:42 |
| amrith | iccha1, Riddhi, am I following so far? | 18:43 |
| Riddhi | yeah, you got that right | 18:44 |
| amrith | so here is why I am puzzled. | 18:44 |
| iccha1 | its also so that the change in the db is atomic | 18:44 |
| amrith | if I look at unassign_configuration() it alread has the possibility of doing many update_db() calls. and there's a problem in the way it is doing that (I think). | 18:45 |
| amrith | observe the code doing the unassigns that produce a TypeError. | 18:45 |
| amrith | lookign for a line # for you | 18:45 |
| amrith | one second | 18:45 |
| amrith | ~ lien 1044 | 18:45 |
| amrith | taskmanager/models.py | 18:45 |
| amrith | so, when you call unassign, it goes one item at a time | 18:46 |
| amrith | and looks to see if that is in the default configuration | 18:46 |
| amrith | with me so far? | 18:46 |
| amrith | maybe this is the easiest way to point to this section of code: https://review.openstack.org/#/c/92701/23/trove/taskmanager/models.py,cm line 1040-1047 | 18:48 |
| *** sgotliv has quit IRC | 18:48 | |
| amrith | iccha1, Riddhi ? | 18:51 |
| Riddhi | yeah thats the case in unassign config, but update_db has been removed from assign config in this case | 18:51 |
| amrith | well, not really | 18:52 |
| amrith | it is still there (see line 1045) | 18:52 |
| amrith | so you could be calling update_db from there as well | 18:53 |
| amrith | yes? | 18:53 |
| amrith | potentially multiple times | 18:53 |
| *** vigneshvar has quit IRC | 18:53 | |
| amrith | or is that rare? | 18:53 |
| Riddhi | sorry, 1045 line in taskmanager/models? | 18:54 |
| amrith | yessir | 18:54 |
| Riddhi | its still part of unassign right()? so wait..the question you have is about the removal of update_db in trove/instance/models in the assign_config method right? | 18:56 |
| amrith | not really, it is about the relocation of the update_db from two places (models.py and models.py) to two other places (service.py) | 18:57 |
| amrith | what I'm basically trying to understand is the rationale behind the change | 18:57 |
| amrith | because as I see it, it was being done in models (along with other update_db) calls. | 18:58 |
| amrith | so why move it | 18:58 |
| amrith | and iccha said it becomes atomic | 18:58 |
| amrith | but it doesn't become atomic because there are still update_db calls in taskmanager/models.py | 18:58 |
| amrith | hence my confusion. | 18:58 |
| amrith | I'm sure the code works | 18:58 |
| amrith | I would like to know what the 'requirement' is. | 18:59 |
| amrith | if it is atomicity, I submit to you that it isn't. | 18:59 |
| amrith | Riddhi, iccha1 ^^ | 18:59 |
| amrith | also, I don't quite know how to delete an instance name. | 19:00 |
| amrith | the code in service.py seems setup to handle it | 19:00 |
| amrith | but I don't know how one would generate a None instance name ;) | 19:00 |
| *** rushiagr is now known as rushiagr_away | 19:00 | |
| *** ajayaa has quit IRC | 19:01 | |
| iccha1 | patch only changes it amrith it cant delete it | 19:01 |
| iccha1 | amrith: so what do u siggest we remove out the update_db changes? | 19:01 |
| amrith | iccha1, re: delete instance name, I thought the bp said that would be supported. maybe it was the wrong bp. | 19:02 |
| amrith | re: removing the update_db changes. Absolutely not, I just wanted to understand them. | 19:02 |
| amrith | it seemed like update_db() in the old place should've worked. I didn't know why it was being changed. | 19:03 |
| *** zacksh has quit IRC | 19:03 | |
| iccha1 | so here is the complication amrith | 19:03 |
| iccha1 | in one of the prev patches peter had commented that the patch was not particularly atomic | 19:03 |
| iccha1 | for our usecase | 19:03 |
| iccha1 | it is hard for patch to be atomic cause of the way config groups work in trove | 19:03 |
| amrith | and it still isn't ;) | 19:03 |
| iccha1 | yes | 19:03 |
| iccha1 | the attempt to combine some of it was because we wanted to atleast do what we can to move towards it | 19:04 |
| Riddhi | yep and i do see your point about the multiple update db calls happening either way and there not being much of value add ...the only one being this: when you want to update config and name, u can pass the kwargs at once to update_db | 19:04 |
| Riddhi | hence being close to atomic..:P | 19:04 |
| amrith | I get it. your code does what you wanted it to. Looks good to me. | 19:05 |
| amrith | thanks for being patient ;) | 19:05 |
| iccha1 | the only other solution is not to allow config group patch calls | 19:05 |
| *** russellb has quit IRC | 19:05 | |
| amrith | iccha1, what's that? | 19:05 |
| *** vigneshvar has joined #openstack-trove | 19:06 | |
| *** adam_g has quit IRC | 19:06 | |
| Riddhi | i think she means not to include config group updation thru patch calls | 19:06 |
| iccha1 | amrith: we as trove do want to move towards patch call instead of put and maybe even support json patch in future | 19:06 |
| *** russellb has joined #openstack-trove | 19:07 | |
| amrith | ok, so, would this work? | 19:07 |
| iccha1 | but i think this is a trade off we have to make with config groups :) | 19:07 |
| amrith | the only db change that unassign_configuration() has to do is potentially set RESTART_REQUIRED | 19:07 |
| amrith | you could easily make unassign_configuration() return that information to the caller | 19:07 |
| amrith | and the caller can only be edit or update | 19:07 |
| amrith | and both of them could include that in their **kwargs to update_db() | 19:08 |
| amrith | that way, you would have a truly atomic (or at least in a single place) update_db() call. | 19:08 |
| amrith | unassign_configuration() would still update_overrides() which is fine. | 19:08 |
| amrith | you'd have the contract be that the caller (edit/update) do the update_db(). | 19:09 |
| *** tomblank_ has joined #openstack-trove | 19:09 | |
| amrith | and you'd have all the information you need there. | 19:09 |
| amrith | now, that is only if this 'atomicity' is really importat | 19:09 |
| amrith | important | 19:09 |
| amrith | if not, this change should suffice. | 19:09 |
| *** freyes has quit IRC | 19:09 | |
| amrith | in any event, I now understand why this change was made. thanks! | 19:09 |
| *** freyes has joined #openstack-trove | 19:11 | |
| *** tomblank has quit IRC | 19:11 | |
| *** tomblank_ is now known as tomblank | 19:11 | |
| *** iccha1 has quit IRC | 19:12 | |
| Riddhi | i agree with making it more cleaner amrith: can it be part of this bug filed by you - https://bugs.launchpad.net/trove/+bug/1364701 | 19:13 |
| Riddhi | we can make it more 'atomic' during the cleanup | 19:15 |
| *** russellb has quit IRC | 19:15 | |
| Riddhi | i dont think 'atomicity' as of now is high priority with this change | 19:16 |
| Riddhi | but there has been a tiny attempt in that direction:P | 19:16 |
| *** freyes has quit IRC | 19:19 | |
| amrith | Thanks Riddhi, that would be fine (I think). | 19:22 |
| *** vigneshvar has quit IRC | 19:28 | |
| *** juice_ has joined #openstack-trove | 19:36 | |
| *** juice has quit IRC | 19:40 | |
| *** IanGovett1 has quit IRC | 19:40 | |
| *** IanGovett1 has joined #openstack-trove | 19:40 | |
| *** juice_ is now known as juice | 19:41 | |
| *** vigneshvar has joined #openstack-trove | 19:45 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/trove: Updated from global requirements https://review.openstack.org/119818 | 19:46 |
| *** zacksh has joined #openstack-trove | 19:52 | |
| *** kfox1111_ has joined #openstack-trove | 19:54 | |
| kfox1111_ | trying to kick the tires with a trunk trove. it wont connect to amqp though. I did set rabbit_host port, hosts, userid/password,virtual_host and ha_queues the same as nova and nova's working. any ideas what I may be missing? | 19:55 |
| *** iccha_ has joined #openstack-trove | 19:57 | |
| iccha_ | sorry was having issues with irc amcrn | 19:57 |
| iccha_ | i mean amrith | 19:57 |
| *** amrith is now known as notamcrn | 19:58 | |
| notamcrn | iccha_, thanks! | 19:59 |
| *** notamcrn is now known as amrith | 19:59 | |
| iccha_ | lol | 19:59 |
| *** ramashri has quit IRC | 20:04 | |
| Riddhi | slicknik: i dint hear back from you about the freeze exception on https://review.openstack.org/#/c/110197/ | 20:09 |
| Riddhi | SlickNik:^^ | 20:09 |
| Riddhi | would this be possible? | 20:09 |
| kfox1111_ | odd. rdo set up nova and friends to use rabbitmq user/pass of guest. i did the same for trove, and am getting a bad password error. | 20:15 |
| *** flaper87 is now known as flaper87|afk | 20:24 | |
| SlickNik | kfox1111_: Trove still uses an older version of the messaging code from olso incubator, so the config values to use are slightly different IIRC. | 20:28 |
| SlickNik | kfox1111_: The format to use is described in probably best described by the sample confs. eg: https://github.com/openstack/trove/blob/master/etc/trove/trove.conf.sample#L18-42 | 20:29 |
| kfox1111_ | yeah. I have rabbit_host,rabbit_port,rabbit_hosts,rabbit_use_ssl,rabbit_userid=guest,rabbit_password=guest,rabbit_virtual_host=/ | 20:31 |
| kfox1111_ | in the logs I see "AMQPLAIN login refused: user 'guest' - invalid credentials" | 20:32 |
| kfox1111_ | its talking to the right host/port. I see the right username. its masking out the pw, but if userid's right, I'd expect password to be. | 20:33 |
| openstackgerrit | iccha-sethi proposed a change to openstack/trove: Delete backup created during replication https://review.openstack.org/119169 | 20:37 |
| *** kevinconway has quit IRC | 20:39 | |
| kfox1111_ | Any other ideas? | 20:39 |
| *** radez is now known as radez_g0n3 | 20:40 | |
| openstackgerrit | amrith proposed a change to openstack/trove: Fix NoSuchOptError on Couchbase create https://review.openstack.org/119636 | 20:41 |
| *** ViswaV has quit IRC | 20:45 | |
| *** ramashri has joined #openstack-trove | 20:53 | |
| *** ramashri has quit IRC | 20:54 | |
| *** ramashri has joined #openstack-trove | 20:55 | |
| *** tomblank has quit IRC | 20:55 | |
| SlickNik | kfox1111_: Not quite sure what may be going on. Couple of things to look at: | 20:56 |
| SlickNik | 1. Looks at the log to see that the cfg values loaded on startup have the correct connection parameters, and | 20:56 |
| SlickNik | 2. perhaps try a manual connection from that server using those creds to see if that's successful. | 20:56 |
| SlickNik | Look* | 20:56 |
| amcrn | kfox1111_: ensure the control_exchange is set correctly as well | 20:57 |
| kfox1111_ | the creds work since nova works. | 20:58 |
| kfox1111_ | I'm not seeing much by way of rabbit in the logs when verbose and debug is turned up. | 20:58 |
| kfox1111_ | how do I check for the control_exhange? | 20:58 |
| SlickNik | amcrn: good call. | 20:58 |
| SlickNik | kfox1111: amcrn was referring to - https://github.com/openstack/trove/blob/master/etc/trove/trove.conf.sample#L167 | 20:58 |
| kfox1111_ | yeah. control_exchange=trove. | 20:59 |
| kfox1111_ | rabbitmqctl list_exchanges | grep trove shows nothing though. do I need to precreate the exchange somehow? | 21:00 |
| SlickNik | kfox1111_: No you shouldn't have to do that. | 21:01 |
| kfox1111_ | k. didn't think so. | 21:01 |
| *** ViswaV has joined #openstack-trove | 21:04 | |
| *** jasonb365 has quit IRC | 21:08 | |
| *** glucas has quit IRC | 21:08 | |
| *** IanGovett1 has quit IRC | 21:09 | |
| *** denis_makogon has joined #openstack-trove | 21:11 | |
| *** vigneshvar has quit IRC | 21:11 | |
| *** denis_makogon has quit IRC | 21:12 | |
| *** denis_makogon has joined #openstack-trove | 21:13 | |
| *** ruebenramirez has quit IRC | 21:13 | |
| *** amcrn_ has joined #openstack-trove | 21:16 | |
| kfox1111_ | interesting... "invalid literal for int() with base 10: '672....' seems like trove doens't like flavors that have uuid's. | 21:18 |
| kfox1111_ | heh. someone beet me to it. bug #1333852. | 21:19 |
| SlickNik | kfox1111_: Yes that's a known issue with trove. — https://bugs.launchpad.net/trove/+bug/1333852 | 21:19 |
| SlickNik | yup, you found it. | 21:20 |
| *** amcrn has quit IRC | 21:20 | |
| *** grapex has quit IRC | 21:21 | |
| kfox1111_ | am I reading the bug right? someone marked it as invalid? | 21:21 |
| *** grapex has joined #openstack-trove | 21:21 | |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Custom Flavors For Config Servers and Routers https://review.openstack.org/120672 | 21:24 |
| SlickNik | kfox1111_: Someone also targeted it to the wrong project (tempest) and it's marked invalid in that project | 21:25 |
| SlickNik | kfox1111_: It's marked in progress in trove | 21:25 |
| *** grapex has quit IRC | 21:26 | |
| *** adam_g has joined #openstack-trove | 21:26 | |
| *** adam_g has quit IRC | 21:27 | |
| *** adam_g has joined #openstack-trove | 21:27 | |
| *** amrith is now known as _amrith_ | 21:27 | |
| openstackgerrit | A change was merged to openstack/trove-integration: Load cfg params conditionally https://review.openstack.org/119880 | 21:32 |
| *** ViswaV has quit IRC | 21:35 | |
| *** robertmyers has quit IRC | 21:44 | |
| *** jcru has quit IRC | 21:50 | |
| *** achampion has quit IRC | 21:56 | |
| *** tomblank has joined #openstack-trove | 21:59 | |
| *** jdandrea has quit IRC | 21:59 | |
| kfox1111_ | SlickNik: ah. I see. thanks. | 22:02 |
| *** iccha_ has quit IRC | 22:04 | |
| *** julienvey has joined #openstack-trove | 22:04 | |
| *** julienvey has quit IRC | 22:09 | |
| *** sgotliv has joined #openstack-trove | 22:12 | |
| *** ViswaV has joined #openstack-trove | 22:13 | |
| *** amcrn_ is now known as amcrn | 22:13 | |
| *** ramashri has quit IRC | 22:15 | |
| kfox1111_ | can you get a raw patch out of the review website? | 22:19 |
| openstackgerrit | Riddhi Shah proposed a change to openstack/trove: Associate flavor types with datastore versions https://review.openstack.org/109824 | 22:20 |
| kfox1111_ | there we go. taking the uuid patch, dropping the extra uuid field and altering the db to have flavor_id be varchar seems to work so far. | 22:21 |
| *** jdandrea has joined #openstack-trove | 22:22 | |
| kfox1111_ | but it fails validation during create. :/ | 22:25 |
| openstackgerrit | Auston McReynolds proposed a change to openstack/trove: Miscellaneous Cluster Fixes https://review.openstack.org/119259 | 22:34 |
| *** ramashri has joined #openstack-trove | 22:35 | |
| *** Riddhi has quit IRC | 22:36 | |
| kfox1111_ | is access to the rabbitmq/nova_proxy_admin creds still required in the instance image? | 22:38 |
| *** openstackgerrit has quit IRC | 22:38 | |
| *** openstackgerrit_ has joined #openstack-trove | 22:38 | |
| *** openstackgerrit_ is now known as openstackgerrit | 22:40 | |
| *** mattgriffin has quit IRC | 22:56 | |
| *** achampion has joined #openstack-trove | 23:02 | |
| SlickNik | kfox1111_: Access to rabbitmq is needed on the guest, since that's how it receives messages from taskmanager / how it communicates with the conductor. | 23:03 |
| SlickNik | nova_proxy_admin creds aren't required, and there's a bug in progress to remove them completely from the codebase. | 23:04 |
| *** julienvey has joined #openstack-trove | 23:06 | |
| *** newb has quit IRC | 23:09 | |
| *** julienvey has quit IRC | 23:11 | |
| *** IanGovett has joined #openstack-trove | 23:24 | |
| kfox1111_ | SlickNik: Can you have a separate queue for api -> conductor/taskmanager and conductor/taskmanager -> guestagent? Right now rabbit is pretty locked down. Spawning a second one is almost easier then trying to make it possible to make it easily vm facing. | 23:24 |
| *** ViswaV has quit IRC | 23:28 | |
| *** ViswaV has joined #openstack-trove | 23:29 | |
| kfox1111_ | OH... I figured out why conductor was authentication denied. arg... trove.conf and trove-conductor.conf. :/ | 23:31 |
| *** _amrith_ is now known as amrith | 23:40 | |
| kfox1111_ | hmm. now Iwhen I start taskmanager it fails with empty catalog. it doesn't ever seem to try and contact keystone though... | 23:42 |
| *** ViswaV has quit IRC | 23:48 | |
| kfox1111_ | hmm... it really wanted a nova endpoint. :/ | 23:52 |
| openstackgerrit | amrith proposed a change to openstack/trove: Fix NoSuchOptError on Couchbase create https://review.openstack.org/119636 | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!