| *** cweid_ has quit IRC | 00:33 | |
| *** ashleighfarnham has quit IRC | 00:53 | |
| *** cheneydc has joined #openstack-trove | 00:57 | |
| *** fabian2 has joined #openstack-trove | 01:09 | |
| *** fabian2 has quit IRC | 01:09 | |
| *** haomaiwang has joined #openstack-trove | 01:12 | |
| *** tellesnobrega_af is now known as tellesnobrega | 01:15 | |
| *** haomaiwang has quit IRC | 01:27 | |
| *** csoukup_ has joined #openstack-trove | 01:55 | |
| *** tellesnobrega is now known as tellesnobrega_af | 02:17 | |
| *** haomaiwang has joined #openstack-trove | 02:23 | |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-trove | 03:01 | |
| *** links has joined #openstack-trove | 03:06 | |
| *** ThunderEmperor has joined #openstack-trove | 03:07 | |
| *** mingdang1 has joined #openstack-trove | 03:38 | |
| *** maishsk has quit IRC | 03:43 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/trove: Updated from global requirements https://review.openstack.org/274662 | 03:44 |
|---|---|---|
| *** haomaiwang has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-trove | 04:01 | |
| *** ewilson has joined #openstack-trove | 04:13 | |
| *** ewilson has quit IRC | 04:52 | |
| *** ewilson has joined #openstack-trove | 04:53 | |
| *** maishsk has joined #openstack-trove | 04:59 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** haomaiwang has joined #openstack-trove | 05:01 | |
| cp16net | pmalik: do you have a trick for getting cassandra to go active? I'm getting an error that the service failed to spawn | 05:06 |
| cp16net | johnma: please look at my comments on this patch and WF it if you agree. thx https://review.openstack.org/#/c/253727/ | 05:07 |
| cp16net | pmalik: i'd like to test some of your casandra patches but having trouble doing so. | 05:08 |
| openstackgerrit | Craig Vyvial proposed openstack/trove: pxc grow/shrink cluster implementation https://review.openstack.org/252090 | 05:16 |
| *** ewilson has quit IRC | 05:25 | |
| *** ThunderEmperor has quit IRC | 05:36 | |
| *** ThunderEmperor has joined #openstack-trove | 05:39 | |
| *** sgotliv has joined #openstack-trove | 05:43 | |
| *** ThunderEmperor has quit IRC | 05:43 | |
| *** maishsk has quit IRC | 05:46 | |
| cp16net | pmalik: hmm looks like i have to use flavor 2 because cassandra is a java memory hog | 05:47 |
| *** haomaiwang has quit IRC | 06:01 | |
| *** haomaiwang has joined #openstack-trove | 06:01 | |
| *** ThunderEmperor has joined #openstack-trove | 06:15 | |
| *** sgotliv has quit IRC | 06:19 | |
| *** ThunderEmperor has quit IRC | 06:46 | |
| *** haomaiwang has quit IRC | 07:01 | |
| *** haomaiwang has joined #openstack-trove | 07:01 | |
| *** haomaiwang has quit IRC | 08:01 | |
| *** haomaiwa_ has joined #openstack-trove | 08:01 | |
| *** sgotliv has joined #openstack-trove | 08:21 | |
| *** akanksha_ has joined #openstack-trove | 08:25 | |
| *** haomaiwa_ has quit IRC | 09:01 | |
| *** haomaiwa_ has joined #openstack-trove | 09:01 | |
| *** chlong has quit IRC | 09:20 | |
| *** haomaiwa_ has quit IRC | 10:01 | |
| *** haomaiwang has joined #openstack-trove | 10:01 | |
| *** cheneydc has quit IRC | 10:02 | |
| *** ThunderEmperor has joined #openstack-trove | 10:23 | |
| *** mingdang1 has quit IRC | 10:54 | |
| *** haomaiwang has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-trove | 11:01 | |
| *** mingdang1 has joined #openstack-trove | 11:39 | |
| *** pbelamge has joined #openstack-trove | 11:51 | |
| *** _amrith_ is now known as amrith | 11:53 | |
| *** haomaiwang has quit IRC | 12:01 | |
| *** haomaiwa_ has joined #openstack-trove | 12:01 | |
| *** links has quit IRC | 12:11 | |
| *** haomaiw__ has joined #openstack-trove | 12:32 | |
| *** haomaiwa_ has quit IRC | 12:34 | |
| *** amrith is now known as _amrith_ | 12:37 | |
| -openstackstatus- NOTICE: Infra running with lower capacity now, due to a temporary problem affecting one of our nodepool providers. Please expect some delays in your jobs. Apologies for any inconvenience caused. | 12:41 | |
| *** haomaiw__ has quit IRC | 13:01 | |
| *** haomaiwa_ has joined #openstack-trove | 13:01 | |
| *** openstackgerrit has quit IRC | 13:02 | |
| *** openstackgerrit has joined #openstack-trove | 13:02 | |
| *** haomaiwa_ has quit IRC | 13:16 | |
| *** tosky has joined #openstack-trove | 13:17 | |
| *** pmackinn has joined #openstack-trove | 13:22 | |
| *** haomaiwang has joined #openstack-trove | 13:36 | |
| *** ThunderEmperor has quit IRC | 13:46 | |
| *** haomaiwang has quit IRC | 14:01 | |
| *** haomaiwa_ has joined #openstack-trove | 14:01 | |
| *** ewilson has joined #openstack-trove | 14:07 | |
| *** mingdang1 has quit IRC | 14:08 | |
| *** _amrith_ is now known as amrith | 14:09 | |
| *** csoukup_ has quit IRC | 14:12 | |
| *** ewilson has quit IRC | 14:13 | |
| *** sabeen3 has joined #openstack-trove | 14:28 | |
| openstackgerrit | Merged openstack/trove: Unittests use trove_testtools https://review.openstack.org/274474 | 14:28 |
| *** akanksha_ has quit IRC | 14:47 | |
| *** ewilson has joined #openstack-trove | 14:47 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/trove: Updated from global requirements https://review.openstack.org/274662 | 14:53 |
| *** haomaiwa_ has quit IRC | 15:01 | |
| *** haomaiwang has joined #openstack-trove | 15:01 | |
| *** csoukup_ has joined #openstack-trove | 15:10 | |
| openstackgerrit | Peter Stachowski proposed openstack/trove-specs: Adds MariaDB Galera cluster spec https://review.openstack.org/257594 | 15:11 |
| openstackgerrit | Petr Malik proposed openstack/trove-specs: Implement cassandra cluster provisioning https://review.openstack.org/266139 | 15:13 |
| *** pbelamge has quit IRC | 15:33 | |
| *** pmackinn_ has joined #openstack-trove | 15:40 | |
| *** pmackinn has quit IRC | 15:43 | |
| *** haomaiwang has quit IRC | 16:01 | |
| *** haomaiwang has joined #openstack-trove | 16:01 | |
| openstackgerrit | Duk Loi proposed openstack/trove: Add support for root-disable https://review.openstack.org/196025 | 16:27 |
| openstackgerrit | Peter Stachowski proposed openstack/trove: Add root-actions int-tests https://review.openstack.org/265957 | 16:32 |
| *** ashleighfarnham has joined #openstack-trove | 16:41 | |
| *** mugsie has quit IRC | 16:46 | |
| *** mugsie has joined #openstack-trove | 16:46 | |
| *** haomaiwang has quit IRC | 17:01 | |
| *** 7F1AA87I3 has joined #openstack-trove | 17:01 | |
| *** itlinux has quit IRC | 17:02 | |
| openstackgerrit | Merged openstack/trove-integration: Use 3 nodes for Vertica install and tests https://review.openstack.org/253727 | 17:10 |
| *** tosky has quit IRC | 17:12 | |
| *** sgotliv has quit IRC | 17:14 | |
| openstackgerrit | Craig Vyvial proposed openstack/python-troveclient: support OS_ENDPOINT_TYPE environment variable https://review.openstack.org/275828 | 17:16 |
| *** pbelamge has joined #openstack-trove | 17:20 | |
| *** tellesnobrega_af is now known as tellesnobrega | 17:22 | |
| openstackgerrit | venkatamahesh proposed openstack/trove: Fix spell typo https://review.openstack.org/275839 | 17:25 |
| *** pbelamge has quit IRC | 17:27 | |
| openstackgerrit | Craig Vyvial proposed openstack/python-troveclient: support OS_ENDPOINT_TYPE environment variable https://review.openstack.org/275828 | 17:30 |
| *** pbelamge has joined #openstack-trove | 17:38 | |
| *** ashleighfarnham has quit IRC | 17:40 | |
| cp16net | trove meeting starting | 18:00 |
| *** pmackinn_ is now known as pmackinn | 18:00 | |
| *** 7F1AA87I3 has quit IRC | 18:01 | |
| *** haomaiwang has joined #openstack-trove | 18:01 | |
| pmalik | cp16net Cassandra requires at least some 2GB of RAM... | 18:02 |
| cp16net | until it uses all that and the guest cant connect | 18:03 |
| *** imandhan has joined #openstack-trove | 18:07 | |
| *** bhaskarduvvuri has joined #openstack-trove | 18:14 | |
| *** sabeen3 has quit IRC | 18:29 | |
| *** bhaskard has joined #openstack-trove | 18:33 | |
| *** bhaskarduvvuri has quit IRC | 18:33 | |
| *** bhaskard is now known as bhaskarduvvuri | 18:33 | |
| amrith | pmackinn, yt? | 18:46 |
| *** csoukup_ has quit IRC | 18:47 | |
| *** ashleighfarnham has joined #openstack-trove | 18:48 | |
| *** ewilson has quit IRC | 18:58 | |
| *** haomaiwang has quit IRC | 19:01 | |
| *** haomaiwang has joined #openstack-trove | 19:01 | |
| *** csoukup_ has joined #openstack-trove | 19:07 | |
| *** sabeen1 has joined #openstack-trove | 19:19 | |
| *** sabeen2 has joined #openstack-trove | 19:21 | |
| cp16net | pmalik: i think there will be a need for setting some ENV variables for cassandra running in the jvm to not take all the memories. we can log a bug on this | 19:22 |
| cp16net | pmalik: i ran a cassandra instance last night to test your changes and it worked. but then this morning i look at the instance and its dead because OSError: [Errno 12] Cannot allocate memory | 19:23 |
| *** sabeen1 has quit IRC | 19:24 | |
| cp16net | https://bugs.launchpad.net/trove/+bug/1541569 | 19:28 |
| openstack | Launchpad bug 1541569 in Trove "cassandra instance died after ~12 hours running" [Low,New] | 19:28 |
| amrith | cp16net, the minimum memory recommended for cassandra is 4gb | 19:36 |
| amrith | and I believe that the default jvm is setup that way | 19:37 |
| amrith | which would mean that over time, you could run into this. | 19:37 |
| cp16net | yikes ok | 19:37 |
| amrith | cp16net, we also need to talk about python-troveclient <-> trove ... | 19:39 |
| amrith | lmk when you have some time | 19:39 |
| cp16net | i guess similar to mongo in that regard where it needs 5gb disk space at least unless setting small-files | 19:39 |
| amrith | it will make your skin crawl | 19:39 |
| * amrith mumbles something about small-files being disabled by default | 19:39 | |
| cp16net | yea | 19:39 |
| cp16net | what about troveclient? | 19:39 |
| amrith | so here's the skinny | 19:40 |
| amrith | there are three actors involved | 19:40 |
| amrith | test-requirements.txt and requirements.txt in the trove repository | 19:40 |
| amrith | python-troveclient in pypi | 19:40 |
| amrith | and upper-constraints.txt and global-requirements.txt in the requirements repository | 19:41 |
| amrith | when we run tests, some tests require the python-troveclient, others don't | 19:41 |
| amrith | the ones that do must get the right one | 19:41 |
| amrith | the ones that don't still use test-requirements so they should get something valid. | 19:41 |
| amrith | the tooling in infra is not setup to use tarballs. | 19:42 |
| cp16net | what does that mean? | 19:42 |
| amrith | the way in which they would like us to do things is push frequent releases of python-troveclient to pypi | 19:42 |
| amrith | and then always reference python-troveclient by version. | 19:42 |
| cp16net | oh | 19:42 |
| amrith | now, enter the super-hero (abramley). | 19:42 |
| amrith | he observed that in kilo and liberty, we are using pip v7 | 19:42 |
| amrith | in mitaka we're using pip v8. | 19:43 |
| amrith | pip v8 has this new check that says that if you pin a module with a tar file, you can't also constrain it with a version. | 19:43 |
| amrith | that's the failure we're seeing. | 19:43 |
| amrith | we fixed it in master (now mitaka) by changing a couple of things | 19:43 |
| amrith | let me find you the link (one second) | 19:44 |
| amrith | https://review.openstack.org/#/c/274345/3/test-requirements.txt | 19:44 |
| amrith | now, we thought we were being very clever when we did this ... https://review.openstack.org/#/c/274345/3/test-requirements.txt and this https://review.openstack.org/#/c/252584/ | 19:45 |
| amrith | here's why I said abramley is a super-hero | 19:45 |
| abramley | yes - do tell | 19:45 |
| amrith | he observed that when mitaka becomes stable and we move to newton as master, we'd have to put a line with a tar file in test-requirements.txt | 19:45 |
| amrith | and that would break. | 19:45 |
| amrith | so, I went asking around and in a conversation with dhellmann I learned a couple of things. | 19:46 |
| amrith | the preferred approach is that we release python-troveclient often and only constrain with versions. | 19:46 |
| amrith | now, that means that our changes for kilo and liberty are not good | 19:46 |
| amrith | we should really pin to specific pypi versions. | 19:46 |
| amrith | so that's item #1 | 19:46 |
| amrith | I don't like the release often approach | 19:46 |
| amrith | because that would mean that we'd have a python-troveclient released to pypi which had to go out ahead of the corresponding controller functionality | 19:47 |
| amrith | this is harder than ... always have master pin to a tar file | 19:47 |
| amrith | and when a release goes stable, tag it to a version | 19:47 |
| amrith | that means in kilo and liberty, we'd match test-requirements.txt to the correct pypi version and make global-requirements.txt match that (in those releases). | 19:48 |
| amrith | in master we'd have our tar file. | 19:48 |
| amrith | when master becomes mitaka, we'd update global-requirements.txt and test-requirements.txt to have the right version | 19:48 |
| amrith | in stable/mitaka | 19:48 |
| amrith | and go on with newton. | 19:48 |
| amrith | when we get to stable/newton, we do the same thing and move along with master being octopus or whatever it is called. | 19:49 |
| amrith | how's that for a plan? | 19:49 |
| amrith | abramley, does this all kind-of make sense? | 19:49 |
| cp16net | yeah while i like the idea of pinning the version in stable/* branches i'm not a fan of having to push untested client code | 19:49 |
| amrith | I agree | 19:49 |
| amrith | i believe the approach makes more sense for libraries (example, oslo) | 19:50 |
| cp16net | this has been an issue since the beginning for us | 19:50 |
| amrith | so I'm going to do the following (lmk if you agree) | 19:50 |
| amrith | I'll email the ML and propose that this is the course that trove will be taking | 19:50 |
| abramley | amrith I believe that is the current thinking | 19:50 |
| amrith | explain why we're not going with the release client often approach | 19:50 |
| abramley | do we know that this is the way it is handled in other projects | 19:50 |
| openstackgerrit | Petr Malik proposed openstack/trove: Implement user functions for Cassandra datastore https://review.openstack.org/206739 | 19:50 |
| amrith | and ask the question abramley just asked | 19:51 |
| amrith | how do other projects do it | 19:51 |
| cp16net | ok | 19:51 |
| amrith | then we can decide how we want to go forward. | 19:51 |
| cp16net | i'm for it | 19:51 |
| amrith | ok, thx | 19:52 |
| amrith | abramley, ok? | 19:52 |
| abramley | ok | 19:52 |
| cp16net | amrith: i'm not sure that this is possible but since we have the pyton-troveclient code checked out in the gate use it instead of the tarball or version? | 19:52 |
| amrith | there's an option called LIBS_FROM_GIT | 19:53 |
| openstackgerrit | Petr Malik proposed openstack/trove: Implement Cassandra Configuration Groups https://review.openstack.org/206740 | 19:53 |
| amrith | that is the recommended way of doing it. | 19:53 |
| amrith | we could explore that as well. | 19:53 |
| cp16net | that sounds like the ideal way to move forward on master | 19:53 |
| cp16net | then it doesnt matter what is in *requirements.txt | 19:54 |
| amrith | that may be the case, I know not what LIBS_FROM_GIT does | 19:54 |
| amrith | or how one sets it. | 19:54 |
| amrith | want to take a look at it? | 19:54 |
| cp16net | yeah i've seen that variable used for many other repos | 19:54 |
| cp16net | but never looked into the details | 19:54 |
| cp16net | i can but it maybe a while before i can | 19:55 |
| cp16net | got some other stuff on the burner atm | 19:55 |
| abramley | but isn't that already set | 19:55 |
| abramley | in localrc.rc in trove-integration ? | 19:55 |
| * cp16net shrugs | 19:55 | |
| amrith | i don't know | 19:55 |
| abramley | oh but that probably isn't used in the gate | 19:56 |
| abramley | we would need it in the devstack plugin | 19:56 |
| amrith | but I intend to look into it soon (not right now, still have freaking alligators chomping at my feet) | 19:56 |
| *** haomaiwang has quit IRC | 20:01 | |
| *** pbelamge has quit IRC | 20:01 | |
| *** pbelamge has joined #openstack-trove | 20:01 | |
| *** 7JTAAES9V has joined #openstack-trove | 20:01 | |
| cp16net | amrith: i hear ya... :-P | 20:01 |
| *** pmackinn is now known as pmackinn|mtg | 20:05 | |
| *** pbelamge has quit IRC | 20:21 | |
| *** sgotliv has joined #openstack-trove | 20:25 | |
| cp16net | pmalik: i responded to your comment on the pxc-grow-shrink | 20:28 |
| *** sabeen2 has quit IRC | 20:31 | |
| *** sabeen1 has joined #openstack-trove | 20:31 | |
| *** sgotliv has quit IRC | 20:36 | |
| openstackgerrit | Petr Malik proposed openstack/trove: Implement Cassandra Configuration Groups https://review.openstack.org/206740 | 20:49 |
| openstackgerrit | Petr Malik proposed openstack/trove: Add backup & restore for Cassandra https://review.openstack.org/206751 | 20:58 |
| *** 7JTAAES9V has quit IRC | 21:01 | |
| *** haomaiwa_ has joined #openstack-trove | 21:01 | |
| openstackgerrit | Nirav Shah proposed openstack/trove: Use correct depends_on decorator for log tests https://review.openstack.org/275923 | 21:01 |
| *** miqui has joined #openstack-trove | 21:08 | |
| *** amrith is now known as _amrith_ | 21:09 | |
| openstackgerrit | Petr Malik proposed openstack/trove: Add root-actions int-tests https://review.openstack.org/265957 | 21:19 |
| *** bhaskarduvvuri has quit IRC | 21:30 | |
| openstackgerrit | Merged openstack/trove: Updated from global requirements https://review.openstack.org/274662 | 21:30 |
| *** sgotliv has joined #openstack-trove | 21:32 | |
| *** sabeen1 has quit IRC | 21:34 | |
| *** sabeen1 has joined #openstack-trove | 21:34 | |
| *** sabeen3 has joined #openstack-trove | 21:39 | |
| *** sabeen1 has quit IRC | 21:41 | |
| openstackgerrit | Dina Belova proposed openstack/trove: Use OSprofiler options consolidated in lib itself https://review.openstack.org/275942 | 21:51 |
| *** haomaiwa_ has quit IRC | 22:01 | |
| *** haomaiwang has joined #openstack-trove | 22:01 | |
| openstackgerrit | Craig Vyvial proposed openstack/python-troveclient: support OS_ENDPOINT_TYPE environment variable https://review.openstack.org/275828 | 22:11 |
| *** _amrith_ is now known as amrith | 22:12 | |
| openstackgerrit | Craig Vyvial proposed openstack/python-troveclient: support OS_ENDPOINT_TYPE environment variable https://review.openstack.org/275828 | 22:13 |
| *** sgotliv has quit IRC | 22:21 | |
| *** arist has quit IRC | 22:25 | |
| *** arist has joined #openstack-trove | 22:26 | |
| *** farid_ has joined #openstack-trove | 22:27 | |
| *** farid_ has quit IRC | 22:34 | |
| *** arist has left #openstack-trove | 22:47 | |
| *** sabeen3 has quit IRC | 22:58 | |
| *** haomaiwang has quit IRC | 23:01 | |
| *** haomaiwa_ has joined #openstack-trove | 23:01 | |
| *** csoukup_ has quit IRC | 23:07 | |
| *** sabeen has joined #openstack-trove | 23:27 | |
| *** csoukup_ has joined #openstack-trove | 23:31 | |
| *** farid has joined #openstack-trove | 23:32 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!