Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-01-16
10:25:23 mdbooth lyarwood: Reading. Might add REGRESSION to the bz title if it's not getting attention.
10:25:50 lyarwood mdbooth: seems to be POST'd for 7.4, I'll ask for 7.3.z now
10:28:19 openstackgerrit Takashi NATSUME proposed openstack/nova: api-ref: Fix keypair API parameters https://review.openstack.org/420626
10:31:49 mdbooth lyarwood: Brilliant, thanks.
10:34:01 openstackgerrit Spencer Yu proposed openstack/nova: Transform instance.live_migration_rollback_dest notification https://review.openstack.org/420629
10:39:32 takashin bauzas: alex_xu: Are you around? Could you review https://review.openstack.org/#/c/408955/ again and check my comments?
10:39:34 openstackgerrit Béla Vancsics proposed openstack/nova: Reduced the complexity of the _init_instance method (4) https://review.openstack.org/420632
10:44:55 mdbooth lyarwood: Reviewed. Would you like me to add the requested test case and resubmit? It'll only take a couple of minutes.
10:45:09 mdbooth Also the os.stat() question.
10:45:28 lyarwood mdbooth: yup sure
10:45:35 mdbooth Ok, will do.
10:48:42 openstackgerrit zhangyujun proposed openstack/nova: Query parameter validation for hosts API https://review.openstack.org/418880
10:50:12 lyarwood mdbooth: thanks btw, I've answered the os.stat question in the review, https://github.com/openstack/nova/blob/ad1c7ac2b102112280a25927d731edb168f80998/nova/virt/libvirt/volume/quobyte.py#L120 is the reason.
10:51:02 openstackgerrit ChangBo Guo(gcb) proposed openstack/nova: Set access_policy for messaging's dispatcher https://review.openstack.org/408494
10:51:23 lyarwood hmmmmm this might not work actually
10:53:11 lyarwood mdbooth: what do you make of L115 from the link above?
10:53:24 lyarwood mdbooth: I don't think that was even valid with findmnt
10:53:43 mdbooth lyarwood: I was just reading it more carefully myself.
10:53:50 mdbooth Surely it previously returned false?
10:53:57 mdbooth I don't think it's mocked in the test.
10:56:02 mdbooth lyarwood: That depends how quobyte volumes are known to mount
10:56:07 lyarwood mdbooth: I think mounted was False previously in this test so os.stat was nnenver called
10:56:39 mdbooth lyarwood: Right. I suspect it was false because it wasn't mocked, and therefore was never going to work in a test environment
10:57:50 mdbooth I think this test needs it to be false, though.
10:57:58 mdbooth That test is bizarre
10:58:06 mdbooth Still trying to wrap my head round what it's doing
10:59:05 mdbooth It's raising an exception if validate_volume() is called with any arguments?
10:59:20 mdbooth Why?
11:00:04 lyarwood validate_volume is calling command = ['getfattr', "-n", "quobyte.info", mnt_base] so I guess it's just mocking a failure of that
11:00:21 lyarwood so it doesn't really matter what we do in connect_volume as long as the volume is mounted in the end
11:00:26 lyarwood urgh
11:00:32 mdbooth It's a terrible test
11:00:58 mdbooth I wouldn't rathole on it. I'd change the is_mounted mock to return False, because that's the previous behaviour, and remove the os.stat mock.
11:01:10 mdbooth Fixing it isn't what you're doing here.
11:01:16 lyarwood yeah true
11:01:16 mdbooth Would you like me to make that change, too?
11:01:22 lyarwood if you have everything open already
11:01:26 mdbooth Yeah.
11:01:29 lyarwood cool thanks
11:06:34 openstackgerrit Takashi NATSUME proposed openstack/nova-specs: Fix formatting errors and warnings https://review.openstack.org/407364
11:18:01 mdbooth lyarwood: Trying trying to work out what that quobyte test did before in case we're missing something.
11:18:36 mdbooth I wonder if it wasn't doing what the author thought it was doing.
11:19:36 openstackgerrit Chris Dent proposed openstack/nova: [placement] add api-ref for GET /resource_providers https://review.openstack.org/418503
11:19:50 lyarwood mdbooth: meh, it might be an older test that tested something useful in the past that's just been left in tbh
11:20:09 lyarwood mdbooth: asserting that we raise the exception that's thrown by the final call in the method is just odd
11:20:19 lyarwood mdbooth: what else could we possibly do?
11:20:31 mdbooth Yeah, I wonder if the exception was raise by something else
11:20:38 mdbooth Checking now, shouldn't take too long
11:20:53 lyarwood kk
11:20:56 lyarwood thanks again
11:25:27 mdbooth lyarwood: The previous code returned True, and didn't handle the os.stat() error in the test
11:25:54 mdbooth It returned true because of fake_execute
11:26:01 lyarwood ahhhh
11:26:37 mdbooth The os.stat() error is handled by the surrounding code
11:27:29 mdbooth Yup, the os.stat() mock can just be removed
11:27:39 mdbooth It remains a terrible test
11:30:05 mdbooth lyarwood: You never did add anything to openstack-dev-hacks, btw
11:31:11 lyarwood mdbooth: yeah sorry keep meaning to
11:31:25 lyarwood mdbooth: I'll try to get around to it this morning
11:31:43 mdbooth Don't both spending time cleaning it up, btw, just chuck it in there
11:31:56 mdbooth We can cleanup later if required
11:32:10 lyarwood yeah I wanted to remove the grep foo | grep bar crap etc ;)
11:32:49 mdbooth Meh, just drop it in there ;)
11:32:59 openstackgerrit Matthew Booth proposed openstack/nova: libvirt: workaround findmnt behaviour change https://review.openstack.org/420146
11:46:23 lyarwood mdbooth: pushed, also dumped my current crop of git aliases in there, not sure it's that useful so feel free to rm
11:47:36 markus_z Would someone please review a tempest test for l-m: https://review.openstack.org/#/c/346815/22 ?
11:51:38 markus_z tonyb: Would you have a look at the newton backports for bug 1455252 please? https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/newton+topic:%22bug+1455252%22
11:51:38 openstack bug 1455252 in OpenStack Compute (nova) "enabling serial console breaks live migration" [High,Fix released] https://launchpad.net/bugs/1455252 - Assigned to Markus Zoeller (markus_z) (mzoeller)
11:52:18 lyarwood markus_z: I can take a look this afternoon
11:52:42 markus_z lyarwood: Thanks, that would be awesome.
11:57:33 mdbooth lyarwood: Thanks!
11:59:06 sahid johnthetubaguy: hello
11:59:25 sahid i remember that you were kind of interested about the feature emulator threads
11:59:58 sahid and unfortunatly even if the code has been pushed 2 months ago I did not had a lot of reviews
12:00:14 sahid would you be interested on helping jay pipes to review it ?
12:13:44 openstackgerrit Chris Dent proposed openstack/nova: Add rudimentary CORS support to placement API https://review.openstack.org/392891
12:57:38 lyarwood sfinucan: would you mind taking a look at https://review.openstack.org/#/c/420146/ if you have a chance this afternoon
13:21:36 openstackgerrit Mikhail Feoktistov proposed openstack/nova: libvirt: ephemeral disk support for virtuozzo containers https://review.openstack.org/312488
13:25:06 openstackgerrit Mikhail Feoktistov proposed openstack/nova: libvirt: ephemeral disk support for virtuozzo containers https://review.openstack.org/312488
13:26:27 dims sdague : bauzas : Got some time for a review that enable *all* functional tests in py35 (except 1) https://review.openstack.org/#/c/419968/ ? please let me know
13:27:23 dims cdent : the lone hold out is the snowman/gabbi test (unicode_get_a_raw_snowman_unicode.test_request) fails as the snowman char is not ascii, so never reaches the server side
13:29:03 dikonoor alex_xu: sdague: Hi..I have some queries around quotas..who would be the right person to contact ?
13:30:34 cdent dims: odd. i wonder which part of the stack is messing with that, because python3 tests in gabbi itself are pretty okay with that sort of thing, can you link me to a failing trace?
13:31:28 dims one sec, digging up
13:32:43 dims cdent : http://logs.openstack.org/68/419968/9/check/gate-nova-tox-db-functional-py35-ubuntu-xenial/dd20ee5/testr_results.html.gz
13:32:53 cdent thanks
13:33:14 sdague dims: lgtm, maybe once this is all done some of those conditional utf8 encode functions could get an oslo safe_encode wrapper or the like
13:33:26 sdague just to get the if PY2 flags out of the code
13:33:53 cdent dims: that's bizarre, I need to compare with a similar test in gabbi itself
13:34:02 dims cdent : ++
13:40:07 cdent dims: actually that test is bogus on python3 and the error is essentially saying "yo, you've made a bad url"
13:40:43 cdent gabbi is not responsible for url encoding and while python2 is okay with that badly formed url, python3 is not
13:41:07 cdent dims: so I would say: okay to remove that one test as it is bogus
13:47:58 dims cdent : ++
13:55:07 openstackgerrit sahid proposed openstack/nova: numa: add numa constraints for emulator threads policy https://review.openstack.org/385023
13:55:10 openstackgerrit sahid proposed openstack/nova: libvirt: return a CPU overhead if isolate emulator threads requested https://review.openstack.org/385364
13:55:10 openstackgerrit sahid proposed openstack/nova: objects: add new field cpus_reserved in NUMACell https://review.openstack.org/400874
13:55:11 openstackgerrit sahid proposed openstack/nova: numa: fit instance topology with cpus reserved onto cores https://review.openstack.org/400875
13:55:11 openstackgerrit sahid proposed openstack/nova: virt: update overhead to take into account vCPUs https://review.openstack.org/385024
13:55:14 openstackgerrit sahid proposed openstack/nova: libvirt: make emulator threads to run on the reserved pCPU https://review.openstack.org/400878

Earlier   Later