| *** saneax is now known as saneax-_-|AFK | 00:05 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Remove excess printing from stats test https://review.openstack.org/417200 | 00:06 |
|---|---|---|
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Verify nodes and requests are not leaked https://review.openstack.org/417136 | 00:06 |
| *** saneax-_-|AFK is now known as saneax | 00:13 | |
| openstackgerrit | Merged openstack-infra/nodepool: Register launcher name with ZooKeeper https://review.openstack.org/416335 | 00:21 |
| *** saneax is now known as saneax-_-|AFK | 00:48 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Remove excess printing from stats test https://review.openstack.org/417200 | 00:55 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Verify nodes and requests are not leaked https://review.openstack.org/417136 | 00:56 |
| *** saneax-_-|AFK is now known as saneax | 00:56 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Verify nodes and requests are not leaked https://review.openstack.org/417136 | 01:11 |
| *** saneax is now known as saneax-_-|AFK | 01:13 | |
| *** Cibo has quit IRC | 01:13 | |
| *** tflink has quit IRC | 01:13 | |
| *** greghaynes has quit IRC | 01:13 | |
| *** _ari_ has quit IRC | 01:13 | |
| *** cinerama has quit IRC | 01:13 | |
| *** jamielennox has quit IRC | 01:13 | |
| *** cinerama has joined #zuul | 01:14 | |
| *** _ari_ has joined #zuul | 01:14 | |
| *** tflink has joined #zuul | 01:14 | |
| *** greghaynes has joined #zuul | 01:14 | |
| *** Cibo has joined #zuul | 01:16 | |
| *** jamielennox has joined #zuul | 01:18 | |
| *** saneax-_-|AFK is now known as saneax | 01:21 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Verify nodes and requests are not leaked https://review.openstack.org/417136 | 01:24 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Verify nodes and requests are not leaked https://review.openstack.org/417136 | 01:28 |
| jeblair | that stack is green now | 01:34 |
| jeblair | clarkb: when you have a moment, can you take a look at whether the tearDown() stuff in https://review.openstack.org/417136 is sane? | 01:34 |
| jeblair | also, if anyone needs me to split up that last patch, i can. i think it's comprehensible, but it's a little random as it fixes a lot of related bugs at once. | 01:35 |
| jeblair | i think the main thing left to do in the series is handle failed allocations | 01:35 |
| *** saneax is now known as saneax-_-|AFK | 01:47 | |
| *** saneax-_-|AFK is now known as saneax | 01:55 | |
| *** saneax is now known as saneax-_-|AFK | 02:41 | |
| *** saneax-_-|AFK is now known as saneax | 03:40 | |
| *** Cibo_ has joined #zuul | 07:02 | |
| *** EmilienM has quit IRC | 08:24 | |
| *** greghaynes has quit IRC | 08:31 | |
| *** harlowja has quit IRC | 08:31 | |
| *** jamielennox has quit IRC | 08:31 | |
| *** phschwartz has quit IRC | 08:31 | |
| *** openstackgerrit has quit IRC | 08:31 | |
| *** jkt has quit IRC | 08:31 | |
| *** adam_g has quit IRC | 08:31 | |
| *** Cibo has quit IRC | 08:31 | |
| *** _ari_ has quit IRC | 08:31 | |
| *** jasondotstar has quit IRC | 08:31 | |
| *** rcarrillocruz has quit IRC | 08:31 | |
| *** TheJulia has quit IRC | 08:31 | |
| *** fungi has quit IRC | 08:31 | |
| *** nibalizer has quit IRC | 08:31 | |
| *** yolanda has quit IRC | 08:31 | |
| *** gothicmindfood has quit IRC | 08:31 | |
| *** SotK has quit IRC | 08:31 | |
| *** bstinson has quit IRC | 08:31 | |
| *** rmoe has quit IRC | 08:31 | |
| *** Cibo_ has quit IRC | 08:31 | |
| *** clarkb has quit IRC | 08:31 | |
| *** gundalow has quit IRC | 08:31 | |
| *** Shrews has quit IRC | 08:31 | |
| *** tristanC has quit IRC | 08:31 | |
| *** nigelb has quit IRC | 08:31 | |
| *** jeblair has quit IRC | 08:31 | |
| *** cinerama has quit IRC | 08:31 | |
| *** tflink has quit IRC | 08:31 | |
| *** timrc has quit IRC | 08:31 | |
| *** Zara has quit IRC | 08:31 | |
| *** dmsimard has quit IRC | 08:31 | |
| *** zaro has quit IRC | 08:35 | |
| *** jasondotstar has joined #zuul | 08:37 | |
| *** _ari_ has joined #zuul | 08:37 | |
| *** gothicmindfood has joined #zuul | 08:37 | |
| *** EmilienM has joined #zuul | 08:37 | |
| *** Cibo_ has joined #zuul | 08:37 | |
| *** jamielennox has joined #zuul | 08:37 | |
| *** greghaynes has joined #zuul | 08:37 | |
| *** tflink has joined #zuul | 08:37 | |
| *** cinerama has joined #zuul | 08:37 | |
| *** harlowja has joined #zuul | 08:37 | |
| *** openstackgerrit has joined #zuul | 08:37 | |
| *** Shrews has joined #zuul | 08:37 | |
| *** nigelb has joined #zuul | 08:37 | |
| *** phschwartz has joined #zuul | 08:37 | |
| *** timrc has joined #zuul | 08:37 | |
| *** nibalizer has joined #zuul | 08:37 | |
| *** yolanda has joined #zuul | 08:37 | |
| *** jkt has joined #zuul | 08:37 | |
| *** clarkb has joined #zuul | 08:37 | |
| *** adam_g has joined #zuul | 08:37 | |
| *** Zara has joined #zuul | 08:37 | |
| *** SotK has joined #zuul | 08:37 | |
| *** tristanC has joined #zuul | 08:37 | |
| *** dmsimard has joined #zuul | 08:37 | |
| *** TheJulia has joined #zuul | 08:37 | |
| *** fungi has joined #zuul | 08:37 | |
| *** rmoe has joined #zuul | 08:37 | |
| *** jeblair has joined #zuul | 08:37 | |
| *** gundalow has joined #zuul | 08:37 | |
| *** rcarrillocruz has joined #zuul | 08:38 | |
| *** _ari_ has quit IRC | 08:38 | |
| *** _ari_ has joined #zuul | 08:38 | |
| *** EmilienM has quit IRC | 08:39 | |
| *** EmilienM has joined #zuul | 08:39 | |
| *** bstinson has joined #zuul | 08:39 | |
| *** Cibo has joined #zuul | 08:39 | |
| *** TheJulia has quit IRC | 08:40 | |
| *** zaro has joined #zuul | 08:47 | |
| *** TheJulia has joined #zuul | 08:58 | |
| *** hashar has joined #zuul | 08:59 | |
| *** bhavik1 has joined #zuul | 11:19 | |
| *** jamielennox is now known as jamielennox|away | 12:08 | |
| *** saneax is now known as saneax-_-|AFK | 13:18 | |
| *** bhavik1 has quit IRC | 17:33 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Use constants for state names https://review.openstack.org/417470 | 18:16 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Handle nodepool allocation failure https://review.openstack.org/417471 | 18:16 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: Set valid states per data model class https://review.openstack.org/417482 | 18:51 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: Set valid states per data model class https://review.openstack.org/417482 | 18:55 |
| clarkb | jeblair: my biggest concern with that change is addCleanup() is a stack so you get nice reverse order of addition cleanups. tearDown() is called as soon as unittest believes the test case to be done, so as long as that ordering is fine I think y ou are ok | 18:57 |
| clarkb | jeblair: the other major difference is cleanups are called even when setup fails and teardowns are not. Which may be desireable in this case if I am reading it correctly | 18:58 |
| clarkb | and double checkign assertFinal state it is good and does only seem to assert not change state so we don't need it to run when setup fails to clean anything up | 19:00 |
| clarkb | the webapp one is the one I would be concerned about. Because it is affecting state we may actually want/need it to be a cleanup | 19:01 |
| jeblair | clarkb: yeah, we could make an argument that it's better to do our assertions in a cleanup, after some things have shut down, but actually figuring out which things are okay to shut down is kind of gnarley (it's okay to cleanup the fake nodepool. it's not okay to cleanup the zookeeper chroot test fixture). so i figure since all the tests are written to assume a stable null state at the end, doing it in the teardown should be okay, and less ... | 19:01 |
| jeblair | ... likely to get us into trouble or otherwise be confusing | 19:01 |
| clarkb | or maybe have a trailing belts and suspenders cleanup for webapp if necessary | 19:02 |
| clarkb | for the setUp() failed case and no teardown is called | 19:03 |
| jeblair | clarkb: yeah, though the webapp test was written to create the state in setup, so undoing that in teardown seems reasonable. if the setup fails there, i don't think it's actually important to clean it up | 19:03 |
| clarkb | oh /me reads setUp more closely | 19:03 |
| clarkb | jeblair: I think the concern would be if change A succeeds but B somehow fails then we wouldn't cleanup A | 19:03 |
| clarkb | but that is all just in memory state looks like so probably fine | 19:04 |
| jeblair | clarkb: yeah, but that's just leaving a change in the queue, which is yeah, that | 19:04 |
| jeblair | clarkb: basically what happens any time a test fails | 19:04 |
| clarkb | yup so ya I think that teardown change should be fine | 19:07 |
| *** Shuo has joined #zuul | 19:30 | |
| Shuo | Where can I see the total number of average commits per day (or hour/week are fine as well) on the dashboard? Also where can I see the total number of tests launched through the queue of zuul? | 19:33 |
| clarkb | Shuo: Gerrit is probably the best source of data for commit info | 19:35 |
| clarkb | Shuo: it has a REST api that you can query to get that sort of info | 19:35 |
| clarkb | Shuo: zuul records job launches with graphite if you configure it with a graphite server. We do that and you can see some of the resulting data at http://status.openstack.org/zuul | 19:36 |
| clarkb | (graphs at bottom of the page) | 19:36 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: Set valid states per data model class https://review.openstack.org/417482 | 19:37 |
| Shuo | clarkb: thanks a lot, I am using the stats from the graph at the bottom. Too lazy to query the API endpoint :-) | 19:40 |
| Shuo | clarkb: there is an "in use" and a 'building' in the Test Nodes graph, what does it mean for a node that is in use but are not building? | 19:41 |
| clarkb | Shuo: those are nodepool states. Building means it is in the process of being nova booted and having its intial config applied. In use means that a job is running on it. Ready means it has finished building but has not yet had a job run on it. Delete means it is queued for deletion | 19:42 |
| Shuo | clarkb: thanks, I guess my interest is to get the ballpark number of how many machines are executing the testing work, which I should be using the number of 'in-use' (roughly speaking) | 19:44 |
| clarkb | Shuo: yes | 19:45 |
| Shuo | clarkb: the other two numbers that I am interested in are probably 'patchset created' (how active developers your working), 'change merged' (how many effective work has been accpted to the product code) | 19:47 |
| clarkb | Shuo: that data is also captured in graphite and graphed on that status page | 19:48 |
| Shuo | those two numbers look roughly ~100/hour, ~10/hour respectively. | 19:48 |
| clarkb | in the Gerrit events per hour graph | 19:48 |
| clarkb | and if you want finer detail on that you can always query gerrit directly | 19:48 |
| Shuo | those two numbers look roughly ~100/hour, ~10/hour respectively in the graph of "Gerrit Event" | 19:48 |
| clarkb | yup | 19:49 |
| Shuo | yeah, ballpark number is sufficient to me at this moment | 19:49 |
| Shuo | clarkb: thanks. | 19:49 |
| Shuo | clarkb: is there the original graphite dashboard URL (is it publicly available)? the graph embeded in ZUUL has been tailored to a specific time scale. | 19:51 |
| clarkb | Shuo: yes graphite.openstack.org | 19:52 |
| clarkb | or view image url and edit from there is what I tend to do | 19:52 |
| Shuo | thanks! | 19:54 |
| timrc_ | mordred, nibalizer: This spoke to me :) https://xkcd.com/1782/ | 20:10 |
| nibalizer | yesssss | 20:10 |
| nibalizer | they heard that twitter fight in your mentions last week | 20:10 |
| timrc_ | nibalizer: lolol | 20:10 |
| dmsimard | xkcd, always so damn relevant | 21:10 |
| timrc_ | dmsimard: ++ | 22:42 |
| *** Shuo has quit IRC | 23:10 | |
| *** saneax-_-|AFK is now known as saneax | 23:22 | |
| *** hashar has quit IRC | 23:38 | |
| *** jamielennox|away is now known as jamielennox | 23:39 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!