| *** tqtran has quit IRC | 00:23 | |
| *** itlinux has quit IRC | 00:43 | |
| *** mingdang1 has joined #openstack-trove | 00:45 | |
| *** jinxing has joined #openstack-trove | 00:51 | |
| *** cheneydc has joined #openstack-trove | 00:54 | |
| *** itlinux has joined #openstack-trove | 00:54 | |
| *** imandhan has quit IRC | 00:56 | |
| *** haomaiwang has joined #openstack-trove | 01:04 | |
| *** can8dnSix has quit IRC | 01:06 | |
| *** tellesnobrega_af is now known as tellesnobrega | 01:07 | |
| *** haomaiwang has quit IRC | 01:09 | |
| *** itlinux has quit IRC | 01:13 | |
| *** jinxing has quit IRC | 01:23 | |
| *** jinxing has joined #openstack-trove | 01:24 | |
| *** itlinux has joined #openstack-trove | 01:25 | |
| *** haomaiwang has joined #openstack-trove | 01:25 | |
| *** mingdang1 has quit IRC | 01:28 | |
| *** itlinux has quit IRC | 01:28 | |
| *** haomaiwang has quit IRC | 01:30 | |
| *** rmevans has joined #openstack-trove | 01:42 | |
| *** nikhil has quit IRC | 01:49 | |
| *** nikhil has joined #openstack-trove | 01:49 | |
| openstackgerrit | William Criswell proposed openstack/trove: Make timeouts datastore specific https://review.openstack.org/303671 | 01:54 |
|---|---|---|
| *** jinxing has quit IRC | 02:12 | |
| *** itlinux has joined #openstack-trove | 02:20 | |
| *** amrith is now known as _amrith_ | 02:30 | |
| *** jinxing has joined #openstack-trove | 02:31 | |
| *** sgotliv has quit IRC | 02:33 | |
| *** tellesnobrega is now known as tellesnobrega_af | 03:36 | |
| *** davideagnello has quit IRC | 03:45 | |
| *** mingdang1 has joined #openstack-trove | 04:02 | |
| *** ozialiendoze has quit IRC | 04:05 | |
| *** ozialien has joined #openstack-trove | 04:06 | |
| *** ozialien has quit IRC | 04:10 | |
| *** ozialien10 has joined #openstack-trove | 04:11 | |
| *** sgotliv has joined #openstack-trove | 04:11 | |
| *** haomaiwang has joined #openstack-trove | 04:26 | |
| *** ozialien10 has quit IRC | 04:28 | |
| *** haomaiwang has quit IRC | 04:31 | |
| *** haomaiwang has joined #openstack-trove | 04:45 | |
| *** haomaiwang has quit IRC | 04:46 | |
| *** haomaiwang has joined #openstack-trove | 04:46 | |
| *** rmevans has quit IRC | 04:52 | |
| *** jinxing has quit IRC | 05:00 | |
| *** rcernin has joined #openstack-trove | 05:03 | |
| *** jinxing has joined #openstack-trove | 05:08 | |
| *** haomaiwa_ has joined #openstack-trove | 05:13 | |
| *** haomaiwang has quit IRC | 05:14 | |
| *** mingdang1 has quit IRC | 05:18 | |
| *** amit213 has quit IRC | 05:27 | |
| *** amit213 has joined #openstack-trove | 05:27 | |
| *** haomaiwa_ has quit IRC | 05:34 | |
| *** imandhan has joined #openstack-trove | 05:49 | |
| *** david-lyle has quit IRC | 06:16 | |
| *** sgotliv has quit IRC | 06:28 | |
| openstackgerrit | Andreas Jaeger proposed openstack/trove-dashboard: Fix periodic translation job https://review.openstack.org/309275 | 06:30 |
| *** mingdang1 has joined #openstack-trove | 06:38 | |
| *** itlinux has quit IRC | 06:42 | |
| *** tesseract has joined #openstack-trove | 06:43 | |
| *** tesseract is now known as Guest83016 | 06:43 | |
| *** Guest83016 has quit IRC | 06:48 | |
| *** sgotliv has joined #openstack-trove | 06:49 | |
| *** haomaiwang has joined #openstack-trove | 06:53 | |
| *** haomaiwang has quit IRC | 06:57 | |
| *** tesseract- has joined #openstack-trove | 07:02 | |
| *** mingdang1 has quit IRC | 07:21 | |
| *** imandhan has quit IRC | 07:24 | |
| -openstackstatus- NOTICE: Logs are failing to be uploaded causing jobs to be marked as UNSTABLE. We are working on repairing the log filesystem and will update when ready. Please do not recheck before then. | 07:31 | |
| *** ChanServ changes topic to "Logs are failing to be uploaded causing jobs to be marked as UNSTABLE. We are working on repairing the log filesystem and will update when ready. Please do not recheck before then." | 07:31 | |
| *** mingdang1 has joined #openstack-trove | 08:21 | |
| *** sgotliv has quit IRC | 08:22 | |
| *** mingdang1 has quit IRC | 08:26 | |
| *** haomaiwang has joined #openstack-trove | 08:26 | |
| *** haomaiwang has quit IRC | 08:30 | |
| *** haomaiwang has joined #openstack-trove | 08:55 | |
| *** haomaiwa_ has joined #openstack-trove | 08:57 | |
| *** haomaiwang has quit IRC | 09:00 | |
| *** haomaiwang has joined #openstack-trove | 09:00 | |
| *** haomaiwa_ has quit IRC | 09:02 | |
| *** haomaiwang has quit IRC | 09:07 | |
| *** haomaiwang has joined #openstack-trove | 09:08 | |
| *** haomaiwang has quit IRC | 09:13 | |
| *** openstackstatus has quit IRC | 09:24 | |
| *** openstackstatus has joined #openstack-trove | 09:26 | |
| *** ChanServ sets mode: +v openstackstatus | 09:26 | |
| *** david-lyle has joined #openstack-trove | 09:28 | |
| *** haomaiwang has joined #openstack-trove | 09:38 | |
| *** tosky has joined #openstack-trove | 09:40 | |
| *** haomaiwang has quit IRC | 09:43 | |
| *** haomaiwang has joined #openstack-trove | 09:53 | |
| *** haomaiwang has quit IRC | 09:58 | |
| *** v12aml has quit IRC | 09:59 | |
| *** cheneydc has quit IRC | 09:59 | |
| *** v12aml has joined #openstack-trove | 10:05 | |
| *** haomaiwang has joined #openstack-trove | 10:07 | |
| *** haomaiwang has quit IRC | 10:11 | |
| *** krotscheck_dcm is now known as krotscheck | 10:19 | |
| openstackgerrit | Flavio Percoco proposed openstack/trove-specs: Extend Trove to allow for other compute backends https://review.openstack.org/307883 | 10:22 |
| *** haomaiwang has joined #openstack-trove | 10:24 | |
| *** haomaiwang has quit IRC | 10:28 | |
| openstackgerrit | Flavio Percoco proposed openstack/trove-specs: Extend Trove to allow for other compute backends https://review.openstack.org/307883 | 10:32 |
| *** _amrith_ is now known as amrith | 10:32 | |
| *** haomaiwang has joined #openstack-trove | 10:41 | |
| *** haomaiwa_ has joined #openstack-trove | 10:45 | |
| *** haomaiwang has quit IRC | 10:45 | |
| *** haomaiwa_ has quit IRC | 10:49 | |
| -openstackstatus- NOTICE: Log server has been repaired and jobs are stable again. If necessary please recheck changes that have 'UNSTABLE' results. | 10:54 | |
| *** jinxing has quit IRC | 11:05 | |
| *** arist1 has quit IRC | 11:08 | |
| *** arist has joined #openstack-trove | 11:08 | |
| *** haomaiwang has joined #openstack-trove | 11:14 | |
| *** haomaiwang has quit IRC | 11:18 | |
| *** miqui has joined #openstack-trove | 11:55 | |
| *** haomaiwang has joined #openstack-trove | 11:59 | |
| *** haomaiwang has quit IRC | 11:59 | |
| *** haomaiwang has joined #openstack-trove | 12:00 | |
| *** AJaeger has joined #openstack-trove | 12:02 | |
| AJaeger | trove-dashboard team, your import from the translation server fails due to broken tox venv environment. Please review and merge https://review.openstack.org/309275 | 12:03 |
| *** flwang has quit IRC | 12:07 | |
| *** mingdang1 has joined #openstack-trove | 12:09 | |
| *** pmackinn has joined #openstack-trove | 12:23 | |
| *** tellesnobrega_af is now known as tellesnobrega | 12:23 | |
| *** amrith is now known as _amrith_ | 12:33 | |
| *** flwang has joined #openstack-trove | 12:35 | |
| *** haomaiwang has quit IRC | 12:59 | |
| *** mingdang1 has quit IRC | 13:02 | |
| *** mingdang1 has joined #openstack-trove | 13:13 | |
| *** _amrith_ is now known as amrith | 13:29 | |
| *** nikhil_ has joined #openstack-trove | 13:40 | |
| *** nikhil_ is now known as Guest5015 | 13:40 | |
| *** Guest5015 is now known as nikhil_k | 13:40 | |
| *** tellesnobrega is now known as tellesnobrega_af | 13:43 | |
| openstackgerrit | Merged openstack/trove-dashboard: Fix periodic translation job https://review.openstack.org/309275 | 13:44 |
| openstackgerrit | amrith proposed openstack/trove: Reuse Cassandra connections https://review.openstack.org/304316 | 13:53 |
| *** AJaeger has left #openstack-trove | 13:55 | |
| *** shakamunyi has joined #openstack-trove | 14:03 | |
| openstackgerrit | amrith proposed openstack/trove: Cleaned guestagent config for data_registry_ext https://review.openstack.org/284434 | 14:07 |
| *** shakamunyi has quit IRC | 14:15 | |
| openstackgerrit | amrith proposed openstack/python-troveclient: configuration-* cli allow name of configuration group entered https://review.openstack.org/233998 | 14:18 |
| *** shakamunyi has joined #openstack-trove | 14:29 | |
| *** mingdang1 has quit IRC | 14:43 | |
| *** itlinux has joined #openstack-trove | 14:57 | |
| *** itlinux has quit IRC | 14:59 | |
| *** haomaiwang has joined #openstack-trove | 15:01 | |
| -openstackstatus- NOTICE: Log server was repaired as of 10:50 UTC and jobs have been stable since. If necessary, please recheck changes that have 'UNSTABLE' results. | 15:02 | |
| *** nikhil has quit IRC | 15:08 | |
| *** tesseract- has quit IRC | 15:09 | |
| *** stewie925 has joined #openstack-trove | 15:15 | |
| *** openstack has joined #openstack-trove | 15:25 | |
| *** haomaiwang has quit IRC | 15:34 | |
| *** barra204 has joined #openstack-trove | 15:37 | |
| pmackinn | anyone know why we don't fork mongod in the overrides? http://bit.ly/1Sozidc | 15:45 |
| pmackinn | pmalik, ? ^^ | 15:47 |
| amrith | pmackinn, problem on some distributions? | 15:56 |
| amrith | that is typically the case, you let something like upstart or systemd manage the daemonization | 15:59 |
| amrith | you need that so the service control managers can get the right process PID | 16:00 |
| amrith | my recollection from way back | 16:01 |
| pmackinn | amrith, it actually blocks systemd | 16:05 |
| *** davideagnello has joined #openstack-trove | 16:06 | |
| pmackinn | mongod 3 will write out its pid and as long as that agrees with the init.d script, bob is your uncle | 16:07 |
| amrith | what vesion of mongodb? | 16:07 |
| pmackinn | 3+ | 16:07 |
| amrith | >2.6.3? | 16:07 |
| pmackinn | what's in trove-int | 16:07 |
| pmackinn | and master really | 16:08 |
| amrith | interesting, I just a training with master (on Monday) and used MongoDB ... | 16:08 |
| amrith | wonder how that worked | 16:09 |
| pmackinn | training on...? | 16:09 |
| amrith | mysql and mongo with trove | 16:09 |
| *** itisha has joined #openstack-trove | 16:09 | |
| pmackinn | go on | 16:09 |
| amrith | perconalive 2016, santa clar | 16:09 |
| pmackinn | what distro? | 16:09 |
| amrith | master/ubuntu | 16:09 |
| * pmackinn drops the mic | 16:09 | |
| pmackinn | trusty i assume? | 16:10 |
| amrith | I'm very trusty | 16:10 |
| amrith | what's trusty now? 14.04? | 16:10 |
| pmackinn | yes | 16:10 |
| amrith | yes, 14.04 lts | 16:10 |
| pmackinn | sysvinit | 16:10 |
| amrith | i dunno, I'm marketing | 16:11 |
| pmackinn | lol | 16:11 |
| amrith | that's all geek stuff | 16:11 |
| amrith | I run trove create | 16:11 |
| pmackinn | sure it is | 16:11 |
| amrith | you should check out trove, it's this dbaas thing | 16:11 |
| amrith | and it takes the complexity out of running a plurality of databases in your private cloud. | 16:11 |
| pmackinn | Ubuntu DBaaS | 16:11 |
| amrith | supports a variety of databases and operating systems | 16:11 |
| pmackinn | as long as those OS are ubuntu | 16:12 |
| amrith | or windows | 16:13 |
| amrith | or aix | 16:13 |
| amrith | or vms | 16:13 |
| amrith | are there even other os's? | 16:13 |
| amrith | seriously thoug | 16:13 |
| amrith | thought | 16:13 |
| amrith | though | 16:13 |
| * pmackinn must have missed the windows branch upstream | 16:13 | |
| amrith | there are a bunch of mongod bugs about failure to launch | 16:14 |
| amrith | oh, it is python. you should check that out, it is os agnostic programming language that delivers maximum portability | 16:14 |
| pmackinn | i look forward to slapping you in person next week | 16:14 |
| amrith | sssssh, that's secret | 16:15 |
| amrith | I'll bring all the 'gear' | 16:15 |
| pmackinn | just got weird | 16:15 |
| amrith | serves you right :) | 16:15 |
| amrith | so back to serious | 16:15 |
| pmackinn | go | 16:15 |
| amrith | there are a bunch of mongodb bugs about systemd issues | 16:15 |
| amrith | on start | 16:15 |
| amrith | looking at https://jira.mongodb.org/browse/SERVER-7285 | 16:15 |
| amrith | for example | 16:16 |
| amrith | this one is closed in 3.3.2 | 16:16 |
| amrith | for fedora and rhel | 16:16 |
| pmackinn | the vendor doesn't make proper systemd scripts, relies on translation of init.d relics | 16:16 |
| amrith | could this be what you are encountering? | 16:17 |
| openstackgerrit | Merged openstack/trove: Update the root scenario tests https://review.openstack.org/283754 | 16:17 |
| amrith | or https://jira.mongodb.org/browse/SERVER-16219 | 16:18 |
| amrith | which is open | 16:18 |
| amrith | this is neat :) | 16:19 |
| amrith | https://review.openstack.org/#/c/137230/ | 16:19 |
| *** tosky has quit IRC | 16:19 | |
| *** rcernin has quit IRC | 16:19 | |
| amrith | anyway, safe travels. see ya monday | 16:19 |
| pmackinn | processManagement.fork: true -> just works (tm) | 16:21 |
| *** nikhil has joined #openstack-trove | 16:40 | |
| openstackgerrit | Petr Malik proposed openstack/trove: Create default datastore options https://review.openstack.org/308126 | 16:44 |
| *** imandhan has joined #openstack-trove | 16:49 | |
| amrith | yo pmackinn ... | 16:52 |
| *** haomaiwang has joined #openstack-trove | 16:53 | |
| *** haomaiwa_ has joined #openstack-trove | 16:56 | |
| *** haomaiwang has quit IRC | 16:58 | |
| *** haomaiwa_ has quit IRC | 17:01 | |
| *** pmackinn has quit IRC | 17:11 | |
| *** pmackinn has joined #openstack-trove | 17:12 | |
| *** barclaac has quit IRC | 17:12 | |
| amrith | flaper87, tellesnobrega_af you there? | 17:18 |
| *** barclaac has joined #openstack-trove | 17:20 | |
| stewie925 | hello all - I am working on a bug fix for https://bugs.launchpad.net/trove/+bug/1541202 | 17:22 |
| openstack | Launchpad bug 1541202 in OpenStack DBaaS (Trove) "config parameter was changes from cluster_member_count to min_cluster_member_count" [Low,In progress] - Assigned to STEW TY (stewie925) | 17:22 |
| stewie925 | and I tried to run tox -e py27, pep8 - but I am getting a few FAILs with a common denominator: oslo_config.cfg.NoSuchOptError: no such option in group vertica: cluster_member_count | 17:24 |
| *** johnma has quit IRC | 17:25 | |
| stewie925 | i did change the existing vertica cfg.int_opt of "cluster_member_count" to "min_cluster_member_count" | 17:26 |
| *** tellesnobrega_af is now known as tellesnobrega | 17:27 | |
| stewie925 | and I added a deprecated_name='cluster_member_count' so it can still be referenced in the same cfg.int_opt | 17:28 |
| *** kfox1111 has joined #openstack-trove | 17:31 | |
| kfox1111 | trove question. Just upgraded from kilo to liberty. | 17:31 |
| kfox1111 | have some existing db's. | 17:31 |
| kfox1111 | doing a trove user-list on one returns: TypeError: __init__() got an unexpected keyword argument 'resource_uuid' | 17:32 |
| kfox1111 | do I have to upgrade all the db's guest agent's too? | 17:32 |
| amrith | kfox1111, one second | 17:32 |
| amrith | I've seen that before, it had to do with oslo.context, I think. | 17:33 |
| * amrith runs to find some notes | 17:34 | |
| kfox1111 | thx. :) | 17:34 |
| amrith | kevin, you do have to first update the guests agents | 17:36 |
| amrith | the issue is that there was a change in oslo.context at some point that added this new field | 17:36 |
| kfox1111 | ok. thanks. | 17:37 |
| amrith | and you have therefore got a controller with the newer oslo.context and a guest without | 17:37 |
| dougshelley66 | kfox1111 in liberty trove moved from using context incubator to official oslo.context. | 17:37 |
| dougshelley66 | the incubator context had "instance_uuid". oslo.context has resource_uuid | 17:37 |
| kfox1111 | ah. | 17:37 |
| kfox1111 | hmm.. so I'm going to have to prepare new images with the right bits in, and then go fix up all the old ones. | 17:38 |
| kfox1111 | amrith: this is what I was suggesting with using seperate docker containers for db and guest agent. This sort of thing would be easier in that case. :) | 17:39 |
| amrith | kfox1111, yes, we're talking about that; not specifically about docker containers but for vms and bare metal as well | 17:39 |
| kfox1111 | you could just docker pull the new gest agent container and restart it. | 17:39 |
| amrith | there's a session called 'superconductor' that would have this discussion | 17:40 |
| amrith | and I think it would be particularly useful for containers | 17:40 |
| kfox1111 | I think the containers could really help with vm's and bare metal too. | 17:40 |
| amrith | since many db vendors provide docker images. | 17:40 |
| kfox1111 | I've been using kubernete's 'kube' agent in stand alone mode. | 17:40 |
| kfox1111 | you just drop a yaml file in /etc/kubernetes/manifests, and it ensures the containers are running the versions/config you specify. | 17:40 |
| kfox1111 | would be really easy to do containers without the overhead of a whole COE. while just needing nova. | 17:41 |
| kfox1111 | upgrading the guest agent then on it would simply be: sed -i 's/kilo/liberty/' /etc/kubernetes/manifests | 17:42 |
| *** tqtran has joined #openstack-trove | 17:43 | |
| kfox1111 | tqtran: hi. :) | 17:43 |
| tqtran | kfox1111: hey kevin | 17:43 |
| *** haomaiwang has joined #openstack-trove | 17:45 | |
| *** sgotliv has joined #openstack-trove | 17:46 | |
| *** haomaiwang has quit IRC | 17:49 | |
| kfox1111 | hmm... so, most of these db's are Ubuntu 15.04 at the moment. and I can't blow them out... | 17:52 |
| kfox1111 | sudo add-apt-repository cloud-archive:liberty fails... | 17:54 |
| kfox1111 | can I just sed -ri 's/instance_uuid/resource_uuid/g' /usr/lib/python2.7/dist-packages/trove/ ? | 17:56 |
| kfox1111 | hehe. yeah. looks like that works. | 17:59 |
| *** haomaiwang has joined #openstack-trove | 17:59 | |
| kfox1111 | so, I'll just throw that workaround into the template for now, and then work on getting an updated distro image going. | 17:59 |
| *** imandhan has quit IRC | 18:02 | |
| kfox1111 | is there any liberty mysql clustering instructions anywhere? | 18:04 |
| *** haomaiwang has quit IRC | 18:04 | |
| *** rmevans has joined #openstack-trove | 18:04 | |
| dougshelley66 | kfox1111 i'm assuming your guests still have the old trove/openstack/common/context.py in which case i believe the sed fix you describe should work | 18:05 |
| *** shakamunyi has quit IRC | 18:05 | |
| *** barra204 has quit IRC | 18:05 | |
| kfox1111 | yeah. | 18:06 |
| *** shakamunyi has joined #openstack-trove | 18:06 | |
| kfox1111 | works ok at least for user-list | 18:06 |
| dougshelley66 | kfox1111 i believe the "mysql" clustering in liberty was the implementation of percona xtradb cluster | 18:06 |
| dougshelley66 | using the "pxc" manager | 18:06 |
| *** rmevans has quit IRC | 18:07 | |
| *** barra204 has joined #openstack-trove | 18:07 | |
| amrith | yes, that's correct. | 18:07 |
| amrith | I believe that you need a pxc guest | 18:07 |
| amrith | other than that, I think it is largely like mongo except that instances are similar, unlike mongodb where they are specialized. | 18:08 |
| kfox1111 | is there any examples of how to set that up? | 18:13 |
| kfox1111 | or do I just register an image and see what breaks? | 18:14 |
| kfox1111 | I had to do quite a complicated cloudinit script to get just a basic mysql to work with kilo. | 18:14 |
| *** SlickNik has quit IRC | 18:16 | |
| *** SlickNik has joined #openstack-trove | 18:17 | |
| *** vipuls has joined #openstack-trove | 18:18 | |
| *** vipul has quit IRC | 18:18 | |
| stewie925 | hello I have a trove question - am fixing a bug to change config parameter but when I ran I am getting a NoSuchOptError | 18:31 |
| stewie925 | the error detailsa are in http://paste.openstack.org/show/495185/ | 18:31 |
| *** haomaiwang has joined #openstack-trove | 18:35 | |
| stewie925 | i changed a config parameter in vertica from 'cluster_member_count' to 'min_cluster_member_count', but kept the 'cluster_member_count' by setting it as a deprecated_name so it can be backward referenced | 18:36 |
| amrith | stewie925, did you push your change upstream? | 18:36 |
| stewie925 | hi amrith yes I did | 18:37 |
| amrith | https://review.openstack.org/#/c/308693/ ? | 18:38 |
| stewie925 | yes that is correct | 18:38 |
| amrith | stewie925, so here's what happened. | 18:39 |
| amrith | first, in your change, you changed trove/common/cfg.py and there are a number of issues with the formatting. | 18:40 |
| amrith | what I suggest you do is run tox on your system before you push the change up, it'll be much easier to debug there. | 18:40 |
| amrith | so you could run | 18:40 |
| amrith | tox -e pep8 | 18:40 |
| amrith | and it'll tell you that the line you added has indentation and formatting issues. spaces and tabs, things like that | 18:40 |
| amrith | then there's the other half of this ... | 18:40 |
| amrith | you changed a paramter name but the old parameter is used all over the code, in many places | 18:41 |
| *** haomaiwang has quit IRC | 18:41 | |
| amrith | so you will have to go and change all of those. | 18:41 |
| stewie925 | oh I see - thank you amrith - let me try running the tox test you just suggested | 18:41 |
| stewie925 | oh | 18:41 |
| stewie925 | many thanks! | 18:41 |
| amrith | I can see about a dozen places | 18:41 |
| amrith | then there's a sample configuration file that you should update | 18:41 |
| amrith | and then the tests | 18:41 |
| *** barra204 has quit IRC | 18:41 | |
| *** shakamunyi has quit IRC | 18:41 | |
| *** shakamunyi has joined #openstack-trove | 18:42 | |
| *** barra204 has joined #openstack-trove | 18:42 | |
| amrith | and before you push up changes, run tox on your local machine | 18:42 |
| amrith | like | 18:42 |
| amrith | tox -e pep8 | 18:42 |
| amrith | tox -e py27 | 18:42 |
| amrith | tox -e py34 | 18:42 |
| amrith | or whatever tests you want to run | 18:42 |
| amrith | or just run | 18:43 |
| amrith | tox | 18:43 |
| amrith | and it'll do them all | 18:43 |
| stewie925 | appreciate your valiable input - am running the pep8 test as we speak | 18:43 |
| amrith | great | 18:43 |
| *** imandhan has joined #openstack-trove | 19:04 | |
| *** itisha has quit IRC | 19:09 | |
| *** haomaiwa_ has joined #openstack-trove | 19:13 | |
| *** haomaiwa_ has quit IRC | 19:22 | |
| openstackgerrit | Petr Malik proposed openstack/trove: Make 'default_password_length' datastore-specific https://review.openstack.org/307939 | 19:33 |
| openstackgerrit | Petr Malik proposed openstack/trove: Make 'default_password_length' datastore-specific https://review.openstack.org/307939 | 19:34 |
| *** twm2016 has joined #openstack-trove | 19:42 | |
| kfox1111 | been a long time since I played with adding datastores... is there an example out there? | 19:50 |
| *** amrith is now known as _amrith_ | 19:53 | |
| stewie925 | hello amrith | 19:54 |
| stewie925 | my pep8 test is now a success (made some indentation changes), but i have the sam issue(s) with the py27 test | 19:55 |
| stewie925 | there's a bunch of fails, but they are all related to the 'oslo_config.cfg.NoSuchOptError: no such option in group vertica: cluster_member_count' error | 19:56 |
| *** itisha has joined #openstack-trove | 20:12 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove: Add variables for quota configuration https://review.openstack.org/274872 | 20:13 |
| *** tellesnobrega is now known as tellesnobrega_af | 20:14 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove: Refactor scenario tests to facilitate multi-group https://review.openstack.org/309190 | 20:18 |
| *** tellesnobrega_af is now known as tellesnobrega | 20:38 | |
| openstackgerrit | Petr Malik proposed openstack/trove: Add db-models and RootController for Postgres https://review.openstack.org/278044 | 20:44 |
| openstackgerrit | Petr Malik proposed openstack/trove: Cleanup guestagent models https://review.openstack.org/278074 | 20:44 |
| *** rmevans has joined #openstack-trove | 20:46 | |
| kfox1111 | so, do you create percona as a datastore version or as its own datastore? | 20:57 |
| *** haomaiwang has joined #openstack-trove | 21:00 | |
| openstackgerrit | Petr Malik proposed openstack/trove: Postgresql Incremental Backup and Restore https://review.openstack.org/259167 | 21:02 |
| openstackgerrit | Petr Malik proposed openstack/trove: Postgresql Streaming Replication https://review.openstack.org/288712 | 21:02 |
| *** haomaiwang has quit IRC | 21:04 | |
| *** davideagnello has quit IRC | 21:06 | |
| *** imandhan has quit IRC | 21:10 | |
| *** _amrith_ is now known as amrith | 21:19 | |
| stewie925 | hi guys - would there be a reason why py27 test fails whereas py34 test is successful? | 21:20 |
| *** haomaiwang has joined #openstack-trove | 21:29 | |
| *** sgotliv has quit IRC | 21:30 | |
| *** haomaiwang has quit IRC | 21:34 | |
| kfox1111 | ok, so, I created a percona datastore type and added a version. but it does not show up under create new cluster in horizon. | 21:35 |
| kfox1111 | any ideas? | 21:35 |
| amrith | percona or pxc kfox1111 | 21:47 |
| amrith | stewie925, because py34 has way fewer tests than py27 | 21:47 |
| *** stewie925_ has joined #openstack-trove | 21:47 | |
| kfox1111 | I did that I think. | 21:47 |
| amrith | kfox1111, you said percona. I'm saying percona and pxc are different. | 21:48 |
| *** stewie925 has quit IRC | 21:48 | |
| amrith | if you did pxc, then, ... lets see | 21:48 |
| kfox1111 | oh. | 21:48 |
| kfox1111 | the data store needs to be named 'pxc'? | 21:48 |
| amrith | yes | 21:48 |
| kfox1111 | ah. gotcha. | 21:48 |
| kfox1111 | is the plugin pxc then too? | 21:49 |
| amrith | let me check | 21:49 |
| amrith | so for pxc, you have a config option min_cluster_member_count which should be set to 2 | 21:49 |
| kfox1111 | ok. so.... | 21:50 |
| amrith | then, how do you register the datastore .... | 21:50 |
| * amrith looking | 21:50 | |
| kfox1111 | trove-manage datastore_update 'pxc' '' | 21:50 |
| kfox1111 | trove-manage datastore_version_update pxc 'pxc-5.5' pxc 1c7163d9-bfc2-4c3c-9e60-0155c09d9d00 '' 1 | 21:50 |
| amrith | so I'm seeing 5.6 | 21:51 |
| kfox1111 | I saw a ticket that said 5.6 wasn't supported in liberty. | 21:51 |
| amrith | oh, liberty | 21:51 |
| amrith | one second | 21:51 |
| kfox1111 | and min_cluster_member_count is under [pxe] in trove.conf? or one of the others? | 21:52 |
| amrith | trove-manage datastore_update pxc | 21:52 |
| amrith | that would be [pxc] | 21:52 |
| amrith | trove-manage datastore_version_update pxc 'pxc-version' pxc <image-id> '' 1 | 21:52 |
| amrith | trove-manage datastore_update pxc 'pxc-version' | 21:53 |
| amrith | that should be it | 21:53 |
| amrith | bbiab | 21:55 |
| kfox1111 | ok. thanks for the help. :) | 21:56 |
| kfox1111 | So, then: | 21:57 |
| kfox1111 | trove cluster-create test-db-cluster pxc pxc-5.5 --instance flavor=m1.small,volume=1 --instance flavor=m1.small,volume=1 --instance flavor=m1.small,volume=1 | 21:57 |
| kfox1111 | fails on quota... | 21:57 |
| kfox1111 | though instance count in nova looks fine. | 21:57 |
| stewie925_ | hi guys, wondering if anyone can shed more light on 'deprecate_name'? | 22:04 |
| *** rmevans has quit IRC | 22:05 | |
| stewie925_ | i mean 'deprecated_name' | 22:05 |
| stewie925_ | what i understand is deprecated_name contains the name of the old config parameter | 22:08 |
| amrith | kfox1111, sorry about the 2 | 22:08 |
| amrith | i think changing min_cluster_count to 2 is something I had just to get around the size of the cluster and my laptop | 22:08 |
| amrith | so leaving it at the default is fine | 22:08 |
| kfox1111 | oh, its not liking m1.small. it wants '2'. | 22:08 |
| amrith | now, it doesn't launch ... | 22:08 |
| kfox1111 | ah. ok. I was thinking that might be the case. | 22:09 |
| amrith | so the default pxc size is 3 (quorum). you can't do quorum consensus with 2 | 22:09 |
| amrith | what do you mean it wants 2? | 22:09 |
| kfox1111 | sorry. flavor=2. | 22:10 |
| kfox1111 | its wanting flavor id's, not flavors. | 22:10 |
| amrith | the help says it should take either | 22:10 |
| amrith | mumble mumble | 22:10 |
| kfox1111 | yeah. I see it fail in nova-api's logs. :/ | 22:10 |
| kfox1111 | but it stopped complaining when I put the flavor id 2 in it. | 22:11 |
| amrith | are the nova instances launched? | 22:11 |
| kfox1111 | no. same error. its complaing about quota. | 22:11 |
| amrith | quota on instances? | 22:11 |
| kfox1111 | but it was when it was a nova flavor id error... | 22:11 |
| kfox1111 | yeah. | 22:11 |
| amrith | nova is complaining? | 22:12 |
| amrith | that's out of trove's control then | 22:12 |
| kfox1111 | no. nothing in the logs. | 22:12 |
| amrith | who is complaining about quota? | 22:12 |
| kfox1111 | trove cluster-create test-db-cluster pxc pxc-5.5 --instance flavor=2,volume=1 --instance flavor=2,volume=1 --instance flavor=2,volume=1 | 22:12 |
| kfox1111 | ERROR: Quota exceeded for resources: ['instances']. (HTTP 413) | 22:12 |
| twm2016 | stewie925_: Try asking your question in #openstack-oslo | 22:12 |
| kfox1111 | if I go into the dashboard though, say launch nova instance, select the same flavor, and hit 3 instances, it says there is pleanty of quota. | 22:13 |
| stewie925_ | thanks Trevor | 22:13 |
| kfox1111 | I also have quota for 20 cinder volumes and currently have 8 allocated. | 22:13 |
| kfox1111 | so I'm not sure which quota its talking about. | 22:13 |
| amrith | stewie925_, https://review.openstack.org/#/c/307939/ may give you some indication about deprecation | 22:13 |
| kfox1111 | it said it was a quota problem too when flavor was m1.small, so I think it may be saying quota when it may not be. | 22:14 |
| amrith | anything in trove API or taskmanager logs? | 22:14 |
| kfox1111 | I"m going to crank up debugging on trove and see if it says anything more. | 22:14 |
| amrith | well quota is an error | 22:14 |
| amrith | so if you see nothing in the trove logs as it is, I'd look in Nova's logs | 22:14 |
| amrith | or cinders | 22:14 |
| amrith | or neutrons | 22:14 |
| amrith | wonder why the log doesn't say who ran out of quota | 22:14 |
| amrith | trove must at least bitch about the resource creation failing or timing out | 22:15 |
| amrith | if it was nova or something else | 22:15 |
| kfox1111 | nothing in nova cinder or neutron. :/ | 22:15 |
| kfox1111 | let me crank up some debugging.... | 22:15 |
| kfox1111 | ah. for some reason the trove logs are only going to journald... at least I'm seeing them now... | 22:16 |
| kfox1111 | weird... very unclear.. let me get a pastebin going... | 22:17 |
| kfox1111 | http://pastebin.com/vK8inxsH | 22:18 |
| kfox1111 | its in trove/quota/quota.py | 22:19 |
| amrith | looking | 22:20 |
| amrith | so you have enough quota with nova and not enough with Trove? | 22:21 |
| kfox1111 | seems like it? | 22:21 |
| amrith | that's weird | 22:22 |
| amrith | looking at what it's up to ... | 22:22 |
| kfox1111 | I only see logs in nova's api at the time.... | 22:22 |
| amrith | so I suspect it is launching one instance or two instances and failing on the third | 22:23 |
| amrith | so how about a trove cluster-list | 22:23 |
| kfox1111 | 4 calls to /v2/dd1e905f59564454b1d89792888d12df/flavors/2 HTTP/1.1" status: 200 | 22:23 |
| kfox1111 | nothing else. | 22:23 |
| kfox1111 | I don't see any instance creates. | 22:24 |
| amrith | ok | 22:24 |
| amrith | trove list | 22:24 |
| amrith | how many do you have? | 22:24 |
| kfox1111 | and a nova flavor show does the same api call. nova flavor-show 2 works. | 22:25 |
| kfox1111 | 3. but nothing related to the cluster I'm trying to create. | 22:25 |
| kfox1111 | is there a trove quota itself? or does it just use the other projects quota's? | 22:26 |
| amrith | no, trove has a quota | 22:26 |
| amrith | its own quotas | 22:26 |
| amrith | i.e. you have a nova quota | 22:26 |
| amrith | and a trove instance quota | 22:26 |
| amrith | similarly a cinder quota | 22:26 |
| amrith | and a trove volume quota | 22:26 |
| kfox1111 | ah. where are those set? | 22:26 |
| amrith | and I can never remeber where we store it :) | 22:26 |
| *** imandhan has joined #openstack-trove | 22:27 | |
| kfox1111 | trove help | grep quota and trove-manage -h | grep quota are blank. | 22:27 |
| kfox1111 | so is grep -ri quota /etc/trove/ | 22:27 |
| amrith | yes, there are management apis for it | 22:30 |
| kfox1111 | the quotas table is blank. there are entries in quota_usages; | 22:31 |
| amrith | and do the entries in quota_usages match your existing instances | 22:31 |
| kfox1111 | I haven't counted closely, but it looks like it. | 22:32 |
| kfox1111 | so it may just not be allowing any new things to be launched. | 22:32 |
| kfox1111 | I'm guessing db quota might have been added in liberty? | 22:33 |
| kfox1111 | so they pre date having any quota records? | 22:33 |
| amrith | so the default | 22:33 |
| amrith | for instances | 22:33 |
| amrith | comes from your config | 22:33 |
| kfox1111 | where would the default be? | 22:33 |
| amrith | for max_instances_per_user | 22:33 |
| kfox1111 | ok. | 22:33 |
| amrith | the default is 5 | 22:33 |
| amrith | which isn't enough | 22:33 |
| kfox1111 | I don't have anything set there I think. | 22:33 |
| kfox1111 | ah. ok. | 22:33 |
| amrith | that's right | 22:33 |
| amrith | the default is 5 | 22:33 |
| kfox1111 | so I need to up the default. | 22:33 |
| kfox1111 | know what the key is? | 22:33 |
| amrith | yes, sending it to you in the paste | 22:33 |
| kfox1111 | ok. thanks. | 22:34 |
| amrith | can't do that | 22:34 |
| amrith | see https://gist.github.com/amrith/4da5fd06b3149f799ede45f588e6585a | 22:34 |
| amrith | you want to change this in your trove taskmanager config file | 22:34 |
| amrith | and restart the taskmanager | 22:34 |
| kfox1111 | ah. ok. thanks. | 22:35 |
| amrith | one second kevin | 22:35 |
| amrith | let me confirm that it isn't trove API config file | 22:35 |
| amrith | ok, to be sure put it in trove.conf as well | 22:36 |
| amrith | I think it should be just one | 22:36 |
| amrith | but I don't want to waste your time restarting and trying | 22:36 |
| amrith | so put it in both please | 22:36 |
| amrith | trove.conf and trove-taskmanager.conf | 22:36 |
| amrith | and restart both services | 22:36 |
| amrith | I'll confirm which it is in paralle | 22:36 |
| amrith | parallel | 22:36 |
| kfox1111 | there we go... the max backups one is newer then liberty. | 22:38 |
| kfox1111 | ok. got taskman restarted... testing... | 22:39 |
| *** davideagnello has joined #openstack-trove | 22:39 | |
| kfox1111 | no, crashed again... NoSuchOptError: no such option: max_volumes_per_user | 22:39 |
| kfox1111 | removing that one too... | 22:39 |
| amrith | wait | 22:39 |
| amrith | let me check the section | 22:39 |
| amrith | it is global | 22:40 |
| kfox1111 | Default section? | 22:40 |
| amrith | did you put at end of file? | 22:40 |
| amrith | yes | 22:40 |
| amrith | common options | 22:40 |
| kfox1111 | yeah, I put it at the begining under the default section. | 22:40 |
| kfox1111 | it didn't like the volume options, but took the rest. | 22:40 |
| kfox1111 | but still complaining. | 22:40 |
| amrith | all 3 are in liberty | 22:41 |
| kfox1111 | the quota error is showing in the trove-api logs. should I just move it to trove.conf? | 22:42 |
| amrith | yes plz | 22:42 |
| amrith | i coulve sworn it would be tm | 22:42 |
| kfox1111 | oh. I had comma's on the end of the two that it was complaining about. | 22:43 |
| amrith | and it is same parser for all 3 | 22:43 |
| kfox1111 | there we go. | 22:43 |
| kfox1111 | in trove.conf, restarted api and taskman. then it at least took it. returned a cluster id. | 22:43 |
| amrith | whats nova say? | 22:44 |
| kfox1111 | 3 vm's launched. | 22:44 |
| amrith | good | 22:44 |
| kfox1111 | so, while this is going... | 22:45 |
| kfox1111 | its working now on the cli. | 22:45 |
| kfox1111 | in horizon though, I cant select a datastore. no entries. | 22:45 |
| amrith | lmc | 22:45 |
| amrith | oh liberty ... | 22:46 |
| kfox1111 | sorry to keep throwing that monkey rench at you. we missed mitaka on our last upgrade window by like 3 days. :/ | 22:47 |
| kfox1111 | now we gota wait 6 months till our next window. :/ | 22:47 |
| amrith | 2m bb | 22:48 |
| kfox1111 | k | 22:48 |
| amrith | ok, am back | 22:52 |
| *** twm2016 has quit IRC | 22:53 | |
| amrith | sorry, was informed that 'i can't have any pudding if I don't ...' | 22:53 |
| amrith | did it work kevin? | 22:54 |
| kfox1111 | lol. sorry about the pudding. :) | 22:55 |
| kfox1111 | no. I think somethings wrong with the guest agent now. | 22:55 |
| kfox1111 | I don't see connections in list_connections | 22:55 |
| dougshelley66 | trove list --include-clustered | 22:58 |
| kfox1111 | oh.. I think I need to slide port 22 into the security group to get in.... | 22:58 |
| amrith | ah, not able to catch up with the speedy rabbit | 22:58 |
| kfox1111 | ah. they are showing up errored in include-clustered | 22:58 |
| amrith | ok, time to get port 22 warmed up | 22:59 |
| amrith | wait a second | 22:59 |
| *** itisha has quit IRC | 22:59 | |
| amrith | what version of oslo.context did you use there :) | 22:59 |
| amrith | on the guest? | 23:00 |
| amrith | did it by some chance get the latest latest latest | 23:00 |
| kfox1111 | I was lazy, and grabbed from http://tarballs.openstack.org/trove/images/ubuntu/ | 23:00 |
| amrith | ok, that has the new oslo.context | 23:00 |
| kfox1111 | which, we should have another conversation about.... we should add those to apps.openstack.org. :) | 23:00 |
| amrith | and you sed -i on the host, no? | 23:00 |
| amrith | :) | 23:00 |
| amrith | let's chat in austin | 23:01 |
| kfox1111 | no. on the vm's. so they should work. | 23:01 |
| kfox1111 | kk. :) | 23:01 |
| amrith | well, these vm's should have the right stuff | 23:01 |
| amrith | so something else broke | 23:01 |
| kfox1111 | I've got a one click installer for glance images in horizon now as part of apps.openstack.org. :) | 23:01 |
| kfox1111 | yeah. I updated the security group. I'm going in. :) | 23:01 |
| kfox1111 | hmmm.. but its not my cloud key. | 23:02 |
| amrith | yes | 23:02 |
| amrith | those images | 23:02 |
| amrith | use a different key | 23:02 |
| amrith | here's what you do | 23:02 |
| amrith | clone openstack/trove-integration | 23:02 |
| amrith | then waltz over to trove-integration/scripts/files/keys | 23:03 |
| kfox1111 | ci-info: | ssh-rsa | 99:ac:87:91:22:12:a1:88:5d:26:9b:63:94:f9:44:0d | - | trove@devstack | | 23:03 |
| amrith | and grab the private key there | 23:03 |
| amrith | yes | 23:03 |
| kfox1111 | oh.. | 23:03 |
| kfox1111 | k. | 23:03 |
| amrith | the private key for that is on trove-integration's source tree | 23:03 |
| kfox1111 | yup. I'm in. | 23:05 |
| kfox1111 | weird... | 23:06 |
| kfox1111 | dpkg -l | grep trove | 23:06 |
| kfox1111 | blank. | 23:06 |
| amrith | that's true | 23:06 |
| amrith | this is trove integration | 23:06 |
| amrith | OH CRAP | 23:06 |
| amrith | these elements are for dev and devstack use | 23:06 |
| amrith | they have no trove | 23:06 |
| amrith | they'll scp them over from the host | 23:06 |
| amrith | at launch time :) | 23:06 |
| amrith | these elements are trove-integration elements generated with redstack | 23:07 |
| kfox1111 | no curl... | 23:07 |
| amrith | so they'll (at runtime) scp the guest code from the host | 23:07 |
| amrith | at 10.0.0.1 | 23:07 |
| amrith | or 10.0.0.2 | 23:07 |
| amrith | so yes, there's no guest there | 23:07 |
| kfox1111 | oh.. ok. | 23:07 |
| amrith | that's why you get no rabbit connection coz there's no guest | 23:07 |
| kfox1111 | maybe I should just drop these images then, and go back to cloudinit for this. | 23:07 |
| kfox1111 | I was hoping it would have all the pxc stuff loaded. but I can figure that out via cloudinit I think. | 23:08 |
| amrith | is that how you do it? | 23:08 |
| amrith | for mysql? | 23:08 |
| kfox1111 | yeah. | 23:11 |
| amrith | we should talk | 23:12 |
| kfox1111 | I'd prefer to do it with heat instead, but last I tried it wasn't entirely stable. | 23:12 |
| amrith | there's a whole presentation about making guest images for trove | 23:12 |
| amrith | presentation/work session ... | 23:12 |
| kfox1111 | yeah. we use dib to build our root images. | 23:12 |
| kfox1111 | but its still kind of manual to sync the images, so the more we have, the more work it is. | 23:13 |
| kfox1111 | we need to automate that pipeline. | 23:13 |
| kfox1111 | we were hoping glance would get multiversion support at some point, but that spawned into glare and then seemingly backburnered. :/ | 23:13 |
| kfox1111 | kind of hoping the whole thing gets rekergiggered around with docker containers. :) | 23:14 |
| amrith | yes, that's what we use too (dib) | 23:14 |
| kfox1111 | then we can use versioin X of the db container, and version Y of the guest agent that matches our cloud, all running on our base images. | 23:14 |
| amrith | we're doing something like that (planning) in the session on 'superconductor'. bad name actually but would be much better for image making | 23:15 |
| kfox1111 | but I'm guessing thats not a 6 month thing. :/ | 23:15 |
| amrith | that is | 23:15 |
| kfox1111 | let me double check my schedule and see if I can make it. | 23:15 |
| kfox1111 | remember the time/date? | 23:15 |
| amrith | even if you can't make the session, i'd like to chat with you | 23:15 |
| amrith | i'll pm you my phone number; I won't be able to IRC from my phone but if you are able, we can coordinate by text/email. | 23:16 |
| kfox1111 | Yeah. that would be good. | 23:16 |
| kfox1111 | yeah. I can text/email. | 23:16 |
| *** stewie925_ has quit IRC | 23:25 | |
| openstackgerrit | Merged openstack/trove-integration: Fix "trove_volume_size" test prop type https://review.openstack.org/288570 | 23:29 |
| *** nikhil has quit IRC | 23:58 | |
| *** nikhil_k is now known as nikhil | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!