Wednesday, 2018-09-05

*** bobh has joined #openstack-mistral01:32
*** apetrich has quit IRC02:09
*** bobh has quit IRC02:15
*** rbrady-afk has quit IRC02:45
*** gkadam has joined #openstack-mistral03:21
*** gkadam has quit IRC04:10
*** gkadam has joined #openstack-mistral04:10
*** bobh has joined #openstack-mistral04:12
*** bobh has quit IRC04:18
*** hardikjasani has joined #openstack-mistral04:41
*** apetrich has joined #openstack-mistral05:46
*** akovi has joined #openstack-mistral05:52
pgaxatted0ugal: I checked and stable/rocky was up-to-date in terms of migrations (rocky is at #29 and I want to backport #30) so it should be fine06:26
pgaxatted0ugal: I just saw rakhmerov's answer on the change, if you need some context I can help. We have a large number of regions so if we enable too many regions for a customer, this bug prevents him to use mistral completely06:36
rakhmerovpgaxatte: I believe we can get approval from the stable branch team if we provide a good justification06:38
pgaxatterakhmerov: I hope so :)06:38
rakhmerov:)06:38
pgaxatteI just realized how fast the catalog grows when you add a region, especially if you use long FQDN for the endpoints06:41
pgaxatterakhmerov: a standard region for us (with neutron,heat-cfn,keystone,nova,mistral,glance,cinderv2,heat,swift,cinder endpoints) takes around 3kB in the catalog and I stripped the url and region name. So you have to add to that the size of each service URL x3 (admin, public and internal endpoints) and the (size of the region name) x (number of service) x 307:13
pgaxattea single region takes ~5kB in the catalog for us07:13
pgaxattea typical client has around 6 regions in his catalog (he can have more or less depending if asked support to have more or has an old project that had less...)07:15
pgaxattethat's about as much details as I can provide :D07:16
pgaxattethe service URL can get pretty as many of them also include the project ID07:17
rakhmerovyes, that's ok07:45
rakhmerovplease write to the mailing list and explain briefly that it's very important for you in production07:46
rakhmerovetc. etc07:46
mcdoker18Hi, are there any ideas about how we should test the ssh action properly? https://review.openstack.org/#/c/599862/07:53
hardikjasanimcdocker18: I've been thinking about it. Apart from mocking (which is of no use), there seems to be no way to have meaningful unit test.07:55
hardikjasaniIsn't run_test.sh run only in functional-tests?07:55
mcdoker18hardikjasani: Yes, you are right :(07:58
*** jaosorior has quit IRC08:03
*** jaosorior has joined #openstack-mistral08:05
rakhmerovhardikjasani: run_test.sh yes, as far as I know. It used to be used at least in functional tests08:17
*** gkadam_ has joined #openstack-mistral08:23
*** gkadam has quit IRC08:27
pgaxatterakhmerov: I'm writing the email to openstack-dev to ask for permission. We need approval from whom exactly to cherry pick the migration?09:22
rakhmerovfrom the stable branch team09:22
rakhmerovI guess that's how they're called09:22
pgaxatteok. I cited your name in the email saying you advised me to ask there for permission, is this ok?09:24
pgaxatterakhmerov ^09:25
rakhmerovpgaxatte: sure, no problem09:25
pgaxatterakhmerov: great, email sent09:27
rakhmerov👍09:27
rakhmerovthumbs up09:27
rakhmerovUnicode: U+1F44D, UTF-8: F0 9F 91 8D09:27
rakhmerovoops... some redundant info was printed )09:27
d0ugallol09:29
d0ugalrakhmerov: I like my emoji to come with additional metadata :-D09:30
rakhmerovyeah09:32
rakhmerovso that people know how to reproduce them )09:32
pgaxatterakhmerov: nice :)09:40
d0ugalpgaxatte: I guess a release note would be a good addition.09:44
d0ugalJust to let operators know that a migration needs to run.09:44
pgaxatted0ugal: oh yeah good catch but can I add it to the cherry pick itself?09:45
d0ugalpgaxatte: You would need to add it to master first, and then backport it.09:45
pgaxatteor should I do another change and cherry pick it to stable/rocky after?09:45
pgaxatteok09:46
pgaxattethought so09:46
d0ugalIf you like, you can squash the backport into one commit - but I don't mind much about that.09:46
pgaxatted0ugal: I don't know how to do that09:46
d0ugalpgaxatte: don't worry, just do it as two commits :)09:46
pgaxattealright :)09:47
d0ugalI can show you if you are cuious09:47
d0ugalcurious09:47
pgaxatted0ugal: yeah sure because I don't see if this is more on git side or on gerrit09:56
d0ugalpgaxatte: The git side09:56
pgaxatteso it's like a cherry pick of two commit in a new branch that are squashed togeter?09:57
d0ugalpgaxatte: yup09:57
pgaxatted0ugal: sounds dangerous :D09:57
d0ugalpgaxatte: Once you have both commits, you can use "git rebase HEAD~2" and then change "pick" to "squash" on the second commit09:57
d0ugaloops09:58
d0ugalThat should be "git rebase HEAD~2 -i"09:58
d0ugalwithout "-i" it wont do anything09:58
pgaxatteyeah to have the interactive selection09:58
d0ugalyup09:58
pgaxattebut I still need to create a new change for the release note anyway right ?09:59
d0ugalpgaxatte: but if you are happier doing it as two commits, please do that :)09:59
d0ugalpgaxatte: Yeah, you do09:59
d0ugalpgaxatte: we need two commits in master because the first merged already09:59
*** mcdoker18 has quit IRC10:00
pgaxatted0ugal: ok I thought there would some gerrit magic in all this but no this is entirely a git manoeuvre. I'll stick to the separate commits :D10:01
d0ugal:)10:01
*** zigo has joined #openstack-mistral10:03
*** zigo has quit IRC11:45
*** gkadam_ has quit IRC11:54
openstackgerritAdriano Petrich proposed openstack/mistral master: Remove extra information from std.ssh action  https://review.openstack.org/59936312:01
openstackgerritPierre Gaxatte proposed openstack/mistral master: Add release note for auth_context bugfix  https://review.openstack.org/60001812:14
pgaxatted0ugal, rakhmerov: the release note for the auth_context bugfix: https://review.openstack.org/60001812:15
pgaxattenot sure if it needs an upgrade section or if the migration instructions should be elsewhere12:16
d0ugalpgaxatte: I think that looks good, thanks!12:19
openstackgerritAndras Kovi proposed openstack/mistral master: Remove extra information from std.ssh action  https://review.openstack.org/59936312:40
*** toure|gone is now known as toure13:04
*** hardikjasani has quit IRC13:18
openstackgerritMerged openstack/mistral master: Minor bug of _assert_multiple_items function  https://review.openstack.org/59929613:38
*** akovi has quit IRC13:55
*** d0ugal has quit IRC13:58
*** jaosorior has quit IRC14:06
*** jtomasek has quit IRC14:15
*** d0ugal has joined #openstack-mistral14:57
*** bobh has joined #openstack-mistral14:58
*** jtomasek has joined #openstack-mistral15:14
*** bobh has quit IRC15:38
*** bobh has joined #openstack-mistral15:56
*** jaosorior has joined #openstack-mistral15:57
*** bobh has quit IRC16:19
*** shardy has quit IRC16:28
*** bobh has joined #openstack-mistral16:35
*** thrash is now known as thrash|biab16:35
*** openstackgerrit has quit IRC18:00
*** thrash|biab is now known as thrash18:01
*** d0ugal has quit IRC18:10
*** jtomasek has quit IRC19:04
*** bobh has quit IRC20:44
*** bobh has joined #openstack-mistral20:44
*** bobh has quit IRC21:03
*** bobh has joined #openstack-mistral21:43
*** toure is now known as toure|gone21:53
*** bobh has quit IRC22:06

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!