| openstackgerrit | Ashleigh Farnham proposed openstack/trove: Update DatastoreNotFound status code to 404 https://review.openstack.org/151055 | 00:20 |
|---|---|---|
| *** openstackgerrit has quit IRC | 00:21 | |
| *** openstackgerrit has joined #openstack-trove | 00:21 | |
| *** david-lyle is now known as david-lyle_afk | 00:28 | |
| *** sbfox has quit IRC | 00:30 | |
| *** Riddhi has joined #openstack-trove | 00:44 | |
| *** eghobo has joined #openstack-trove | 00:47 | |
| *** amcrn has quit IRC | 00:56 | |
| *** vigneshvar has joined #openstack-trove | 00:58 | |
| *** jcru has quit IRC | 00:58 | |
| *** annashen has quit IRC | 01:03 | |
| *** ashleighfarnham has quit IRC | 01:04 | |
| *** openstackgerrit has quit IRC | 01:05 | |
| *** openstackgerrit has joined #openstack-trove | 01:05 | |
| *** rkourtz has joined #openstack-trove | 01:12 | |
| *** IanGovett has quit IRC | 01:12 | |
| *** IanGovett has joined #openstack-trove | 01:16 | |
| *** nexusz99 has joined #openstack-trove | 01:18 | |
| *** Riddhi has quit IRC | 01:24 | |
| *** IanGovett has quit IRC | 01:25 | |
| *** vigneshvar has quit IRC | 01:36 | |
| *** jcru has joined #openstack-trove | 01:37 | |
| *** jcru has quit IRC | 01:37 | |
| *** nexusz99 has quit IRC | 01:37 | |
| *** vigneshvar has joined #openstack-trove | 01:37 | |
| *** vigneshvar has quit IRC | 01:41 | |
| *** atomic77 has quit IRC | 01:54 | |
| *** sushilkm_ has quit IRC | 01:55 | |
| *** nexusz99 has joined #openstack-trove | 01:55 | |
| *** nexusz99 has quit IRC | 02:04 | |
| *** ashleighfarnham has joined #openstack-trove | 02:04 | |
| *** rkourtz has quit IRC | 02:08 | |
| *** ashleighfarnham has quit IRC | 02:10 | |
| *** rkourtz has joined #openstack-trove | 02:11 | |
| *** rwsu is now known as rwsu-afk | 02:13 | |
| *** haomaiwang has joined #openstack-trove | 02:14 | |
| *** saurabhs has quit IRC | 02:20 | |
| *** openstackgerrit has quit IRC | 02:20 | |
| *** openstackgerrit has joined #openstack-trove | 02:20 | |
| *** erkules_ has joined #openstack-trove | 02:21 | |
| *** erkules has quit IRC | 02:23 | |
| *** IanGovett has joined #openstack-trove | 02:37 | |
| *** robertmyers has joined #openstack-trove | 02:46 | |
| *** eghobo has quit IRC | 02:53 | |
| *** tomblank has joined #openstack-trove | 02:55 | |
| *** radez_g0n3 is now known as radez | 03:04 | |
| *** annashen has joined #openstack-trove | 03:08 | |
| *** IanGovett1 has joined #openstack-trove | 03:15 | |
| *** IanGovett has quit IRC | 03:18 | |
| *** openstackgerrit has quit IRC | 03:20 | |
| *** openstackgerrit has joined #openstack-trove | 03:20 | |
| *** pmackinn has quit IRC | 03:24 | |
| *** IanGovett1 has quit IRC | 03:29 | |
| *** sushilkm_ has joined #openstack-trove | 03:44 | |
| *** dkehn has quit IRC | 03:52 | |
| *** dkehn has joined #openstack-trove | 03:52 | |
| *** radez is now known as radez_g0n3 | 04:04 | |
| openstackgerrit | Anna Shen proposed openstack/trove: Add limit stanzas for mongos https://review.openstack.org/150131 | 04:09 |
| SlickNik | thansen: sorry, missed your earlier message. | 04:10 |
| thansen | SlickNik: no problem, got a few follow up questions | 04:10 |
| *** saurabhs has joined #openstack-trove | 04:11 | |
| *** saurabhs has quit IRC | 04:11 | |
| thansen | SlickNik: give me a couple minutes though, getting ready to do a deployment | 04:11 |
| SlickNik | thansen: Sure — grabbing a quick bite, but I should be around. | 04:12 |
| *** bhunter71 has quit IRC | 04:15 | |
| *** annashen has quit IRC | 04:17 | |
| *** annashen has joined #openstack-trove | 04:18 | |
| *** sbfox has joined #openstack-trove | 04:22 | |
| *** annashen has quit IRC | 04:28 | |
| *** annashen has joined #openstack-trove | 04:31 | |
| *** annashen has quit IRC | 04:31 | |
| *** annashen has joined #openstack-trove | 04:34 | |
| *** annashen has quit IRC | 04:44 | |
| *** cp16net has quit IRC | 04:45 | |
| *** cp16net has joined #openstack-trove | 04:46 | |
| *** f13o has quit IRC | 05:00 | |
| *** timbyr_ has quit IRC | 05:00 | |
| thansen | SlickNik: still around? | 05:03 |
| *** annashen has joined #openstack-trove | 05:10 | |
| *** sbfox has quit IRC | 05:13 | |
| SlickNik | thansen: am now. | 05:13 |
| *** timbyr_ has joined #openstack-trove | 05:13 | |
| *** f13o has joined #openstack-trove | 05:13 | |
| thansen | SlickNik: so basically I'm after some general information about long-term maintenance | 05:13 |
| thansen | so let's say a couple months down the road (after creating a trove instance) vulnerability is found | 05:14 |
| thansen | let me rephrase that, a trove *image* | 05:15 |
| thansen | I assume the instances are out of luck without doing a destroy and recreate? | 05:16 |
| SlickNik | thansen: We (HP) basically use an out-of-band config / remote execution tool (salt) to manage patching the existing instances to mitigate vulnerabilities. | 05:19 |
| SlickNik | thansen: You could use something like ansible as well. | 05:19 |
| SlickNik | thansen: I think most folks who deploy trove have some out of band tool to help them do something like this. | 05:19 |
| thansen | SlickNik: you install a salt client in the images? | 05:19 |
| SlickNik | thansen: Yes, we do for our (HP) public cloud images. | 05:20 |
| thansen | yeah | 05:20 |
| thansen | ok, what about patching the DB itself though | 05:20 |
| thansen | that sounds dangerous to automate restarting mysql for example | 05:20 |
| SlickNik | Yes, that would have customer impact. | 05:21 |
| cp16net | it could be just as dangerous if not more to not restart it in some cases | 05:22 |
| SlickNik | You'd basically have to weigh the outcome of running vulnerable with the customer impact of the restart. | 05:22 |
| thansen | yeah, to that's a case by case management decision though (again, out of band) | 05:22 |
| SlickNik | Usually you're able to schedule a maintenance window of sorts with customers to do something like that. | 05:23 |
| thansen | understood | 05:23 |
| thansen | what exactly does db_upgrade do? | 05:23 |
| thansen | cp16net: yeah, I get it, just trying to figure out what kinds of things trove itself does vs. not | 05:25 |
| cp16net | it manages a database | 05:26 |
| SlickNik | thansen: upgrades the DB schema between versions | 05:26 |
| cp16net | not really the security of the database or image | 05:26 |
| thansen | SlickNik: you mean the actual 'trove' openstack management db? | 05:27 |
| cp16net | yeah db_upgrade is for the backend db of trove (not customers db) | 05:27 |
| SlickNik | thansen: Yes, the db that trove stores all its backend data in | 05:27 |
| thansen | ok, that all makes sense | 05:27 |
| SlickNik | thansen: We use sqlalchemy migrate to do move between versions (https://sqlalchemy-migrate.readthedocs.org/en/latest/changeset.html) | 05:28 |
| * thansen nods | 05:28 | |
| SlickNik | thansen: And the schema versions are defined at https://github.com/openstack/trove/tree/master/trove/db/sqlalchemy/migrate_repo/versions | 05:28 |
| SlickNik | Again all this is for the internal trove backend DB | 05:29 |
| thansen | right, I'm noting that down here | 05:29 |
| thansen | changing directions, if you have multiple datastore versions, the cloudinit files can't do much | 05:30 |
| thansen | ie: the database installation etc can't be defined in the cloudinit file and must be in the image already | 05:31 |
| thansen | do I understand that correctly? | 05:31 |
| SlickNik | Yes: the practice that most folks I know who run trove follow is to do exactly that, and have the datastore (at a particular version) baked in the image. | 05:33 |
| * thansen nods and notes it down | 05:33 | |
| SlickNik | Moving to a new major version can then happen using something like backup and restore. | 05:34 |
| thansen | I see | 05:34 |
| cp16net | you can bake anything you need into an image (ie configs/monitoring/super cool feature) | 05:35 |
| cp16net | if restore will work for the next magor version... | 05:35 |
| thansen | so if that's 'best practice' then setting the 'packages' param doesn't do much or serve much purpose | 05:35 |
| cp16net | well the other direction is that you can use a common image across datastores and have the guest install the packages you want | 05:36 |
| thansen | yeah | 05:37 |
| SlickNik | thansen / cp16net: Although I've rarely seen folks take that approach FWIW. | 05:37 |
| thansen | the packages param correlates to cloudinit 'packages' right? | 05:37 |
| cp16net | SlickNik: true for a deployment but development its really nice :-D | 05:38 |
| SlickNik | cp16net: True that. | 05:38 |
| SlickNik | thansen: It corresponds to whatever packages mean for the distro — apt-get pkgs in debian based distros, yum in RH based ones. | 05:39 |
| thansen | right, but it injects it into the cloudinit files right? | 05:39 |
| SlickNik | There's logic in the guest to install the packages based on the distro type (if the packages aren't already present on the image) | 05:39 |
| thansen | oh, so it's 'out of band' of cloudinit then | 05:40 |
| SlickNik | thansen: It just uses the standard package-management system to fetch the packages | 05:40 |
| SlickNik | yes out of band of cloud-init | 05:40 |
| thansen | ok | 05:41 |
| thansen | so yum and apt are the only supported installers? | 05:41 |
| SlickNik | Yes, I believe that's currently the case. | 05:41 |
| *** Longgeek has joined #openstack-trove | 05:43 | |
| SlickNik | https://github.com/openstack/trove/blob/master/trove/guestagent/pkg.py | 05:43 |
| thansen | ok, last line of questions (I think) | 05:43 |
| SlickNik | (RedHatPackagerMixin and DebianPackagerMixin are the two implementations) | 05:44 |
| thansen | yeah | 05:44 |
| thansen | sooo, the volume that gets attached | 05:44 |
| SlickNik | thansen: just curious — where are you writing this down? If you don't mind me asking. :) | 05:44 |
| thansen | on my openstack notes doc in gedit :) | 05:45 |
| SlickNik | I'm wondering if you can make a blog post about it or something — there might be some good info there that might be useful for other folks as well. Just thinking out loud... | 05:45 |
| SlickNik | Or a wiki page. | 05:46 |
| SlickNik | Didn't mean to distract you from your question. | 05:46 |
| thansen | if someone changes the volume size of a trove instance, does trove actually expand the volume or what happens exactly | 05:46 |
| thansen | SlickNik: yeah, I've been thinking about that | 05:46 |
| SlickNik | So trove calls into cinder and basically — 1. detaches the volume from the instance 2. Calls a cinder API to resize the volume 3. Reattaches the volume and "grows" the FS on it. | 05:48 |
| *** vigneshvar has joined #openstack-trove | 05:49 | |
| thansen | SlickNik: ok, this *must* happen via the trove api right? not trying to directly manipulate the volume itself | 05:50 |
| SlickNik | Yes, you make a trove API call to do this, and Trove takes care of making the internal cinder calls to achieve the rest. | 05:50 |
| thansen | right | 05:50 |
| SlickNik | So you're not calling into Cinder or working with the volume directly at any point. | 05:51 |
| thansen | does it handle shutting the db down and restarting it as part of that process then? | 05:51 |
| thansen | ie: steps 0 and 4? | 05:51 |
| SlickNik | Yes, I believe it does to quiesce the data. | 05:52 |
| SlickNik | Otherwise bad things can happen :) | 05:52 |
| *** rkourtz_ has joined #openstack-trove | 05:53 | |
| thansen | right, just making sure | 05:53 |
| thansen | a 'shrink' is available as well? | 05:53 |
| *** rkourtz has quit IRC | 05:54 | |
| SlickNik | you can use the same resize action to "shrink" — I believe you can do it as long as the DB size on disk is less than the volume size you're trying to resize to. | 05:58 |
| *** sbfox has joined #openstack-trove | 05:59 | |
| thansen | SlickNik: ok, makes sense | 05:59 |
| thansen | this is probably a stupid question (near 0 experience with cloud/openstack) but what happens if more compute power is required | 06:00 |
| SlickNik | resize: https://github.com/openstack/trove/blob/master/trove/taskmanager/models.py#L1394-1407 | 06:00 |
| SlickNik | thansen: not a stupid question. | 06:00 |
| thansen | can the nova instance be 'upgraded' | 06:00 |
| thansen | or is a backup/spin up/restore required | 06:01 |
| SlickNik | Nova supports flavor-resize, and the trove API has a flavor resize to do this | 06:01 |
| thansen | ok | 06:01 |
| thansen | when those happen does nova actually handle expanding the fs as well? | 06:01 |
| SlickNik | So you could resize your instance from a smaller flavor to a larger flavor (or vice versa) | 06:02 |
| *** annashen has quit IRC | 06:02 | |
| thansen | right, but will nova actually resize the fs when appropriate? | 06:02 |
| *** Longgeek has quit IRC | 06:03 | |
| SlickNik | Yes, nova handles resizing the root FS on the new flavor — the user db is not on the root FS, but on the volume though. So to enable the DB to use more space, a volume resize is required. | 06:03 |
| thansen | yeah, understood | 06:04 |
| thansen | that was more of a generic nova question | 06:04 |
| thansen | SlickNik: ok, any luck any of this will work with nova docker? | 06:07 |
| thansen | :) | 06:07 |
| SlickNik | thansen: Honestly, no idea. | 06:09 |
| SlickNik | I haven't tried it out — perhaps someone else might have and would have a better idea. | 06:10 |
| SlickNik | I'd be curious to know too. | 06:10 |
| thansen | I see that viable for performance bound workloads | 06:10 |
| SlickNik | I'm not sure if the nova-docker driver supports operations like resize. | 06:10 |
| *** timbyr_ has quit IRC | 06:10 | |
| thansen | yeah | 06:10 |
| SlickNik | Although I don't see why the volume resize pieces wouldn't just work. | 06:11 |
| thansen | so I just tried to do a resize on my rdo stack and got this.. | 06:11 |
| thansen | http://pastebin.com/53zvdpNZ | 06:11 |
| thansen | just a generic nova resize that is | 06:11 |
| SlickNik | So the nova-scheduler failed to find a compute instance for the resized VM. | 06:12 |
| thansen | uh | 06:12 |
| SlickNik | nova-scheduler.log probably has some more info about why that happened, perhaps. | 06:13 |
| thansen | pretty sure it's missing an oslo server | 06:14 |
| thansen | kicking that off right from the horizon interface | 06:14 |
| thansen | so the compute instance is certainly there | 06:14 |
| thansen | that's more of a question for the rdo folks though | 06:15 |
| SlickNik | Yeah, you'd probably want to dig into nova-scheduler logs to see why the scheduling failed. | 06:16 |
| thansen | yeah, I'll deal with that tomorrow or something | 06:18 |
| SlickNik | But could be due to a variety of reasons (no capacity, not supported by the hypervisor being used, some other filter not matched) | 06:18 |
| *** Longgeek has joined #openstack-trove | 06:18 | |
| thansen | SlickNik: thanks so much for your time, very appreciated | 06:18 |
| SlickNik | thansen: You're welcome. | 06:18 |
| *** timbyr_ has joined #openstack-trove | 06:19 | |
| SlickNik | bbiab | 06:19 |
| thansen | SlickNik: yeah, very informative..still have a few things to delve into but I feel like I'm starting to grasp trove well enough | 06:20 |
| *** Longgeek has quit IRC | 06:22 | |
| *** eghobo has joined #openstack-trove | 06:24 | |
| *** rkourtz_ has quit IRC | 06:24 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove-specs: Add support for new Vertica datastore in Trove https://review.openstack.org/151126 | 06:36 |
| *** eghobo has quit IRC | 06:44 | |
| *** sushilkm_ has quit IRC | 06:46 | |
| *** johnma has quit IRC | 06:49 | |
| *** vigneshvar has quit IRC | 06:57 | |
| *** vigneshvar has joined #openstack-trove | 06:57 | |
| *** annashen has joined #openstack-trove | 06:58 | |
| *** annashen has quit IRC | 07:07 | |
| *** coolsvap|afk is now known as coolsvap | 07:10 | |
| *** ashleighfarnham has joined #openstack-trove | 07:32 | |
| *** ashleighfarnham has quit IRC | 07:37 | |
| *** romainh has joined #openstack-trove | 07:54 | |
| *** romainh has left #openstack-trove | 07:54 | |
| *** vigneshvar has quit IRC | 07:55 | |
| *** timbyr_ has quit IRC | 07:58 | |
| *** f13o has quit IRC | 07:58 | |
| *** timbyr_ has joined #openstack-trove | 08:15 | |
| *** f13o has joined #openstack-trove | 08:16 | |
| *** pboros has joined #openstack-trove | 08:27 | |
| *** amrith has quit IRC | 08:32 | |
| *** amrith has joined #openstack-trove | 08:32 | |
| *** chlong has quit IRC | 08:36 | |
| *** amrith is now known as _amrith_ | 08:43 | |
| *** _amrith_ is now known as amrith | 08:43 | |
| *** boblebauce has joined #openstack-trove | 08:54 | |
| *** amrith is now known as _amrith_ | 09:02 | |
| *** _amrith_ is now known as amrith | 09:02 | |
| *** vigneshvar has joined #openstack-trove | 09:09 | |
| *** isviridov_away is now known as isviridov | 09:23 | |
| *** Longgeek has joined #openstack-trove | 09:23 | |
| *** erkules_ is now known as erkules | 09:24 | |
| *** sbfox has quit IRC | 09:25 | |
| *** Longgeek has quit IRC | 09:27 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/trove: Updated from global requirements https://review.openstack.org/149985 | 10:07 |
| *** chlong has joined #openstack-trove | 10:13 | |
| *** chlong has quit IRC | 10:27 | |
| *** haomaiwang has quit IRC | 10:34 | |
| *** chlong has joined #openstack-trove | 10:44 | |
| *** boblebauce has quit IRC | 10:46 | |
| *** boblebauce has joined #openstack-trove | 10:52 | |
| *** boblebauce has quit IRC | 11:17 | |
| *** chlong has quit IRC | 11:39 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove-specs: Add support for new Vertica datastore in Trove https://review.openstack.org/151126 | 11:49 |
| *** atomic77 has joined #openstack-trove | 12:03 | |
| *** chlong has joined #openstack-trove | 12:05 | |
| *** radez_g0n3 is now known as radez | 12:08 | |
| openstackgerrit | Denis M. proposed openstack/trove: Fix cassandra.yaml permissions after write_config operation https://review.openstack.org/151198 | 12:11 |
| denis_makogon | SlickNik, amrith guys, could you please review critical patches that are blocking adding Cassandra int-tests to the gate? here they are https://review.openstack.org/#/c/107100/ and recent https://review.openstack.org/151198 | 12:12 |
| amrith | denis_makogon, I looked at your comment re: permissions. | 12:26 |
| amrith | It looks like (from the logging in thebug) that there is a permission issue. I'm unsure where you got the output of ls, was it just done on a system with the problem and you modified the access bits on the "what it should be"? or did you get that on some other machine where cassandra was working. | 12:27 |
| amrith | ? | 12:27 |
| *** IanGovett has joined #openstack-trove | 12:32 | |
| *** tomblank has quit IRC | 12:43 | |
| *** atomic77 has quit IRC | 12:51 | |
| *** ankit_ag has joined #openstack-trove | 12:52 | |
| denis_makogon | amrith, i do have running cluster from previous code, and i was able to ssh into in and saw that required permissions are 644, but not 600 | 12:52 |
| amrith | interesting that the timestamps are all identical. | 12:55 |
| amrith | Like Jan 29 for cassandra.yaml | 12:55 |
| *** adam_g is now known as adam_g_out | 12:57 | |
| amrith | never mind. I'll look more at the change. | 12:57 |
| *** ashleighfarnham has joined #openstack-trove | 12:59 | |
| denis_makogon | amrith, this are two different nodes - one on two equal devstack instances | 13:00 |
| denis_makogon | so, anyway, Cassandra can't start if permissions are incorrectly changed | 13:02 |
| amrith | one without the permission change and one with it? Today is Jan 29 ;) | 13:02 |
| denis_makogon | yes, it is | 13:02 |
| amrith | if the one with 644 has cassandra.yaml dated Jan 29, that is most peculiar. | 13:02 |
| *** ashleighfarnham has quit IRC | 13:04 | |
| *** atomic77 has joined #openstack-trove | 13:04 | |
| *** johnma has joined #openstack-trove | 13:04 | |
| denis_makogon | so, in any case, cassandra expects to have different permissions rather than they are now | 13:19 |
| *** radez is now known as radez_g0n3 | 13:19 | |
| *** exploreshaifali has joined #openstack-trove | 13:27 | |
| *** boblebauce has joined #openstack-trove | 13:30 | |
| *** bhunter71 has joined #openstack-trove | 13:39 | |
| *** pmackinn has joined #openstack-trove | 13:41 | |
| *** amrith is now known as _amrith_ | 13:56 | |
| *** Barker has joined #openstack-trove | 13:57 | |
| openstackgerrit | Denis M. proposed openstack/trove: Fix Cassandra service status definition https://review.openstack.org/107100 | 14:10 |
| *** Barker has quit IRC | 14:25 | |
| *** exploreshaifali has quit IRC | 14:32 | |
| *** ashleighfarnham has joined #openstack-trove | 14:49 | |
| *** isviridov is now known as isviridov_break | 14:53 | |
| *** ashleighfarnham has quit IRC | 14:54 | |
| *** ankit_ag has quit IRC | 14:56 | |
| *** Barker has joined #openstack-trove | 14:57 | |
| *** Barker has quit IRC | 14:59 | |
| *** tomblank has joined #openstack-trove | 15:00 | |
| *** shayneburgess has joined #openstack-trove | 15:02 | |
| openstackgerrit | Oleksandr Kyrylchuk proposed openstack/trove: Added 'redis' test group to int-tests https://review.openstack.org/151267 | 15:09 |
| *** radez_g0n3 is now known as radez | 15:26 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove-specs: Add Vertica cluster spec https://review.openstack.org/151279 | 15:32 |
| *** Riddhi has joined #openstack-trove | 15:33 | |
| *** Riddhi has quit IRC | 15:39 | |
| *** Riddhi has joined #openstack-trove | 15:40 | |
| *** isviridov_break is now known as isviridov | 15:43 | |
| *** _amrith_ is now known as amrith | 15:43 | |
| amrith | on my guest, the files are owned by cassandra and cassandra runs as the cassandra user. I'm not sure about your system configuration but I'd like to look further before just making a change. | 15:49 |
| *** jcru has joined #openstack-trove | 15:50 | |
| *** shayneburgess has quit IRC | 15:53 | |
| *** annashen has joined #openstack-trove | 15:57 | |
| denis_makogon | amrith, here's what i've been using as source of images, https://github.com/denismakogon/trove-guest-image-elements, and it doesn't seem like an issue of image elements, but more likely guest's fault because temp file permissions are defined as 600 and those permissions are not allowing to read this file to any other users, so 644 gives an ability to read this file to users that are in root group and all others | 15:59 |
| denis_makogon | amrith, stepping aside, could you +2 trivial fix https://review.openstack.org/#/c/151267/ ? | 16:01 |
| denis_makogon | cp16net, yt ? | 16:01 |
| *** mattgriffin has joined #openstack-trove | 16:03 | |
| *** thedodd has joined #openstack-trove | 16:04 | |
| *** tomblank has left #openstack-trove | 16:07 | |
| openstackgerrit | Denis M. proposed openstack/trove: Fix poll_until retriever while creating an instance https://review.openstack.org/97194 | 16:12 |
| *** annashen has quit IRC | 16:13 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove-specs: Add Vertica cluster spec https://review.openstack.org/151279 | 16:13 |
| openstackgerrit | Denis M. proposed openstack/trove: Fix backup state check while restoring an instance https://review.openstack.org/130220 | 16:19 |
| *** Riddhi has quit IRC | 16:29 | |
| *** Riddhi has joined #openstack-trove | 16:29 | |
| *** atomic77 has quit IRC | 16:45 | |
| *** exploreshaifali has joined #openstack-trove | 16:48 | |
| *** atomic77 has joined #openstack-trove | 16:48 | |
| *** david-lyle_afk is now known as david-lyle | 16:49 | |
| *** shayneburgess has joined #openstack-trove | 16:54 | |
| *** rwsu-afk is now known as rwsu | 16:55 | |
| *** Riddhi has quit IRC | 16:56 | |
| *** pboros has quit IRC | 16:58 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove-specs: Add Vertica cluster spec https://review.openstack.org/151279 | 16:59 |
| *** pmalik has joined #openstack-trove | 17:07 | |
| *** coolsvap is now known as coolsvap|afk | 17:10 | |
| *** boblebauce has quit IRC | 17:16 | |
| *** isviridov is now known as isviridov_away | 17:17 | |
| openstackgerrit | Sushil Kumar proposed openstack/trove-specs: Add Vertica cluster spec https://review.openstack.org/151279 | 17:28 |
| *** annashen has joined #openstack-trove | 17:52 | |
| *** mattgriffin has quit IRC | 17:53 | |
| *** eghobo has joined #openstack-trove | 17:53 | |
| *** sbfox has joined #openstack-trove | 17:53 | |
| *** pmalik has quit IRC | 17:55 | |
| *** radez is now known as radez_g0n3 | 18:00 | |
| *** mattf is now known as _mattf | 18:04 | |
| *** pmalik has joined #openstack-trove | 18:05 | |
| *** f13o has quit IRC | 18:06 | |
| *** timbyr_ has quit IRC | 18:06 | |
| *** f13o has joined #openstack-trove | 18:08 | |
| *** timbyr_ has joined #openstack-trove | 18:08 | |
| *** eghobo has quit IRC | 18:21 | |
| *** coolsvap|afk is now known as coolsvap | 18:22 | |
| *** thedodd has quit IRC | 18:28 | |
| *** ashleighfarnham has joined #openstack-trove | 18:32 | |
| *** exploreshaifali has quit IRC | 18:34 | |
| *** eghobo has joined #openstack-trove | 18:44 | |
| *** atomic77 has quit IRC | 18:47 | |
| *** openstackgerrit has quit IRC | 18:50 | |
| *** openstackgerrit has joined #openstack-trove | 18:51 | |
| *** Riddhi has joined #openstack-trove | 18:54 | |
| *** boblebauce has joined #openstack-trove | 18:55 | |
| *** coolsvap is now known as coolsvap|afk | 18:56 | |
| *** shayneburgess1 has joined #openstack-trove | 18:59 | |
| *** shayneburgess has quit IRC | 18:59 | |
| *** tomblank has joined #openstack-trove | 19:00 | |
| *** shayneburgess has joined #openstack-trove | 19:03 | |
| *** eghobo has quit IRC | 19:04 | |
| *** shayneburgess1 has quit IRC | 19:05 | |
| *** boblebauce has quit IRC | 19:05 | |
| *** eghobo has joined #openstack-trove | 19:07 | |
| *** annashen has quit IRC | 19:07 | |
| *** eghobo has quit IRC | 19:09 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove-integration: kick-start doesn't warn on stale image file https://review.openstack.org/151348 | 19:09 |
| *** mattgriffin has joined #openstack-trove | 19:11 | |
| *** annashen has joined #openstack-trove | 19:20 | |
| *** tomblank has left #openstack-trove | 19:23 | |
| jodah | is the cluster resource not documented yet or am i missing it? | 19:37 |
| *** jwang__ has joined #openstack-trove | 19:51 | |
| *** ashleigh_ has joined #openstack-trove | 19:56 | |
| *** johnma_ has joined #openstack-trove | 19:56 | |
| *** annashen has quit IRC | 19:56 | |
| *** johnma has quit IRC | 19:56 | |
| *** jwang_ has quit IRC | 19:56 | |
| *** johnma_ is now known as johnma | 19:56 | |
| *** ashleighfarnham has quit IRC | 19:57 | |
| *** Riddhi has quit IRC | 19:57 | |
| *** sbfox has quit IRC | 19:58 | |
| *** thedodd has joined #openstack-trove | 20:00 | |
| *** annashen has joined #openstack-trove | 20:01 | |
| *** sbfox has joined #openstack-trove | 20:01 | |
| *** ashleigh_ has quit IRC | 20:02 | |
| *** openstackgerrit has quit IRC | 20:04 | |
| *** openstackgerrit has joined #openstack-trove | 20:04 | |
| *** shayneburgess1 has joined #openstack-trove | 20:13 | |
| *** vigneshvar has quit IRC | 20:14 | |
| *** exploreshaifali has joined #openstack-trove | 20:14 | |
| *** shayneburgess has quit IRC | 20:15 | |
| SlickNik | jodah: It's not in the API docs yet like you mention — I think the docs were overlooked when the feature merged so they still need to be added. | 20:15 |
| SlickNik | jodah: Let me quickly open a LP bug for it so we can track it. | 20:16 |
| SlickNik | right now I think the best docs we have for it are in the cluster specs. | 20:16 |
| SlickNik | But we need to fix that. | 20:16 |
| jodah | SlickNik: somewhat related, would be nice to get the latest dox on openstack's site | 20:17 |
| jodah | not sure what's involved in that | 20:18 |
| jodah | Where does python-troveclient read its public (api) url from? | 20:19 |
| *** annashen has quit IRC | 20:20 | |
| SlickNik | From the keystone-catalog once authentication happens. | 20:22 |
| jodah | is there an override option? | 20:22 |
| SlickNik | ( you can explicitly specify a bypass URL to override that) | 20:22 |
| jodah | --endpoint-type? | 20:22 |
| *** fifieldt has quit IRC | 20:24 | |
| SlickNik | —bypass-url | 20:24 |
| *** rkourtz has joined #openstack-trove | 20:24 | |
| *** fifieldt has joined #openstack-trove | 20:25 | |
| jodah | that bypasses the keystone server or the trove API? | 20:25 |
| jodah | seeing something like this | 20:25 |
| jodah | ubuntu@ubuntu:~/devstack$ trove --bypass-url http://10.0.1.7:8779 list | 20:25 |
| jodah | ERROR: AuthSystemNotFound: '' | 20:25 |
| SlickNik | actually not sure, let me take a quick look to confirm. | 20:26 |
| jodah | looking to just point to a diff API | 20:26 |
| *** pmalik has quit IRC | 20:28 | |
| SlickNik | jodah: As I can recall, and from a quick inspection of the code, bypass-url should override the trove endpoint-url from the catalog. I'm not entirely sure what the ERROR you're running into is. | 20:31 |
| SlickNik | the auth-url you set defines the keystone endpoint to hit. | 20:31 |
| jodah | hmm... ok | 20:32 |
| jodah | what library satisfies this dependency? ImportError: No module named xmlrpc_client | 20:32 |
| jodah | not seeing something like that in python-troveclient/requirements.txt | 20:33 |
| jodah | chasing down the transitive dependencies is a pain | 20:35 |
| SlickNik | Haven't seen that one before — what requires it, or are you seeing it in a traceback? | 20:37 |
| jodah | it's six | 20:37 |
| SlickNik | it looks like it's part of the standard lib in Python3 | 20:38 |
| SlickNik | so probably under six.moves | 20:38 |
| SlickNik | yeah | 20:38 |
| jodah | but of course i'm rockin 2.7.6 python :) | 20:38 |
| jodah | ok - that'll fix it | 20:38 |
| jodah | is it the thing to do - to have diff virtualenvs for each tool you run? (i've seen this suggested, but seems kinda over the top)? | 20:39 |
| SlickNik | Well personally, I have one virtualenv for the openstack python tools I run. | 20:41 |
| SlickNik | tools + clients. | 20:41 |
| jodah | one for each separate client? | 20:41 |
| SlickNik | No just one that I run all the clients from. | 20:41 |
| jodah | ok - that's reasonable | 20:42 |
| *** pmalik has joined #openstack-trove | 20:42 | |
| jodah | where do you store your venv files for openstack? | 20:44 |
| *** ashleighfarnham has joined #openstack-trove | 20:45 | |
| *** rkourtz has quit IRC | 20:45 | |
| *** Riddhi has joined #openstack-trove | 20:46 | |
| SlickNik | In my home folder under venvs, but you can put them anywhere. (~/venvs/OpenStack/) | 20:46 |
| jodah | cool, tnx | 20:47 |
| *** annashen has joined #openstack-trove | 20:53 | |
| jodah | man, can't catch a break | 20:53 |
| jodah | (openstack)aurora:venvs jonathan$ trove --bypass-url http://localhost:8779/v1.0/821eb40621c043979d429342b713a63f list | 20:53 |
| jodah | ERROR: AuthSystemNotFound: '' | 20:53 |
| jodah | this is running from a different machine | 20:53 |
| jodah | ....perhaps bypass-url is intended to be a keystone url? | 20:55 |
| jodah | Is anyone familiar with the cause of AuthSystemNotFound when using --bypass-url? | 20:57 |
| *** shayneburgess1 has quit IRC | 21:00 | |
| *** eghobo has joined #openstack-trove | 21:04 | |
| *** eghobo has quit IRC | 21:11 | |
| *** ViswaV has joined #openstack-trove | 21:17 | |
| *** miqui has quit IRC | 21:18 | |
| *** miqui has joined #openstack-trove | 21:21 | |
| *** pmalik has quit IRC | 21:23 | |
| SlickNik | jodah: I'm wondering if this is a regression after the keystone-v3 fix that introduced the AuthSystem concept. | 21:26 |
| *** amrith is now known as _amrith_ | 21:30 | |
| *** radez_g0n3 is now known as radez | 21:32 | |
| jodah | SlickNik possibly. is anyone else developing with trove-api on one machine and devstack on another? | 21:35 |
| SlickNik | Not 100% sure but I think ekotowski was doing that a while ago. | 21:40 |
| SlickNik | jodah: Also can you try https://pypi.python.org/pypi/python-troveclient/1.0.7 and see if you're seeing the same issue? <= The previous version before the v3 changes were merged. | 21:41 |
| jodah | 1.0.7 seems to work. so yea, i'd say regression | 21:43 |
| *** cweid has quit IRC | 21:46 | |
| jodah | want me to file it? | 21:47 |
| SlickNik | yeah that would be sweet. Thanks! | 21:47 |
| jodah | np | 21:47 |
| *** shayneburgess has joined #openstack-trove | 21:51 | |
| *** radez is now known as radez_g0n3 | 22:00 | |
| *** Riddhi has quit IRC | 22:01 | |
| *** Riddhi has joined #openstack-trove | 22:01 | |
| *** eghobo has joined #openstack-trove | 22:03 | |
| *** chlong has quit IRC | 22:08 | |
| *** atomic77 has joined #openstack-trove | 22:21 | |
| *** mattgriffin has quit IRC | 22:25 | |
| *** mattgriffin has joined #openstack-trove | 22:32 | |
| *** ViswaV has quit IRC | 22:32 | |
| *** shayneburgess has quit IRC | 22:37 | |
| *** shayneburgess has joined #openstack-trove | 22:39 | |
| jodah | SlickNik Does the root extension apply to anything in addition to mysql? | 22:40 |
| *** jcru has quit IRC | 22:40 | |
| jodah | I'd like to make it a top level extension (not mysql specific) to support vertica as well | 22:40 |
| jodah | or... do we have technology specific extensions (mysql, vertica) with root impls for each? | 22:41 |
| *** ViswaV has joined #openstack-trove | 22:41 | |
| *** eghobo has quit IRC | 22:42 | |
| *** openstackgerrit has quit IRC | 22:51 | |
| *** sbfox has quit IRC | 22:51 | |
| *** openstackgerrit has joined #openstack-trove | 22:51 | |
| *** IanGovett has quit IRC | 23:09 | |
| *** sgotliv has quit IRC | 23:15 | |
| jodah | Any pointers on what this error message from python-troveclient is referring to? "ERROR: flavor" | 23:15 |
| jodah | the HTTP request seems to have been accepted. not sure what the error represents though (the flavors i passed exist) | 23:17 |
| *** amcrn has joined #openstack-trove | 23:20 | |
| *** mattgriffin has quit IRC | 23:35 | |
| *** exploreshaifali has quit IRC | 23:37 | |
| *** thedodd has quit IRC | 23:41 | |
| *** atomic77 has quit IRC | 23:50 | |
| *** chlong has joined #openstack-trove | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!