Thursday, 2018-02-22

« Wednesday, 2018-02-21 Index Friday, 2018-02-23 »
*** dougbtv__ has joined #openstack-kuryr00:04
*** dougbtv_ has quit IRC00:07
*** yamamoto has joined #openstack-kuryr00:08
*** dougbtv_ has joined #openstack-kuryr00:13
*** yamamoto has quit IRC00:14
*** dougbtv__ has quit IRC00:15
*** dmellado has joined #openstack-kuryr00:33
*** maysamacedos has quit IRC00:50
*** maysamacedos has joined #openstack-kuryr01:02
maysamacedoscelebdor: ping01:07
*** maysamacedos has quit IRC01:10
*** yamamoto has joined #openstack-kuryr01:10
*** celebdor has quit IRC01:11
*** caowei has joined #openstack-kuryr01:12
*** dougbtv_ has quit IRC01:12
*** dougbtv_ has joined #openstack-kuryr01:15
*** yamamoto has quit IRC01:16
*** maysamacedos has joined #openstack-kuryr01:27
*** celebdor has joined #openstack-kuryr01:27
*** kiennt26 has joined #openstack-kuryr01:31
*** celebdor has quit IRC01:39
*** salv-orlando has joined #openstack-kuryr01:41
*** salv-orlando has quit IRC01:46
*** jlabarre has quit IRC01:47
*** c00281451_ has joined #openstack-kuryr01:48
*** c00281451__ has quit IRC01:52
*** c00281451_ has quit IRC01:58
*** yamamoto has joined #openstack-kuryr02:12
*** hongbin has joined #openstack-kuryr02:13
*** yamamoto has quit IRC02:18
*** maysamacedos has quit IRC02:29
*** maysamacedos has joined #openstack-kuryr02:47
*** yamamoto has joined #openstack-kuryr03:14
*** yamamoto has quit IRC03:20
*** janonymous has joined #openstack-kuryr03:37
*** yamamoto has joined #openstack-kuryr03:37
*** salv-orlando has joined #openstack-kuryr03:43
*** salv-orlando has quit IRC03:49
maysamacedosHi everyone, did anyone already had the following error while configuring the overcloud VM in the nested env04:03
maysamacedosovercloud-vm.novalocal etcd[3295]: listen tcp <UNDERCLOUD_IP>:2379 :bind: cannot assign requested address04:05
maysamacedosThe etcd service is active in the undercloud env04:07
maysamacedosand it is accessible in the overcloud04:07
maysamacedosbut cannot be started in the overcloud04:11
maysamacedosI would appreciate any thoughts on it :)04:11
*** maysamacedos has quit IRC04:19
*** salv-orlando has joined #openstack-kuryr04:45
*** salv-orlando has quit IRC04:49
*** hongbin has quit IRC05:02
*** janki has joined #openstack-kuryr05:27
*** salv-orlando has joined #openstack-kuryr05:45
*** salv-orlando has quit IRC05:50
*** yboaron has joined #openstack-kuryr06:07
openstackgerritMerged openstack/kuryr-kubernetes master: Run OpenShift 3.7.1 in DevStack plugin  https://review.openstack.org/52337006:12
*** salv-orlando has joined #openstack-kuryr06:19
*** gcheresh_ has joined #openstack-kuryr06:32
*** gcheresh has joined #openstack-kuryr06:37
*** gcheresh_ has quit IRC06:37
openstackgerritMerged openstack/kuryr-kubernetes master: Refactor Kuryr-Kubernetes gates to Zuul-V3  https://review.openstack.org/54495906:39
*** salv-orlando has quit IRC06:51
*** salv-orlando has joined #openstack-kuryr06:51
*** salv-orlando has quit IRC06:56
*** juriarte has joined #openstack-kuryr07:10
*** juriarte has quit IRC07:15
openstackgerritEunsoo Park proposed openstack/kuryr-kubernetes master: Watcher restarts watching fixed resources  https://review.openstack.org/54689507:19
*** juriarte has joined #openstack-kuryr07:30
*** pcaruana has joined #openstack-kuryr07:37
*** slaweq has joined #openstack-kuryr07:45
*** slaweq has quit IRC07:48
*** slaweq has joined #openstack-kuryr07:49
*** salv-orlando has joined #openstack-kuryr07:52
*** salv-orlando has quit IRC07:56
*** threestrands_ has joined #openstack-kuryr08:15
*** threestrands_ has quit IRC08:15
*** threestrands_ has joined #openstack-kuryr08:15
*** threestrands has quit IRC08:18
*** Natanbro has joined #openstack-kuryr08:24
openstackgerritMerged openstack/kuryr-tempest-plugin master: Merge create and list pods tests  https://review.openstack.org/54657908:26
*** celebdor has joined #openstack-kuryr08:29
openstackgerritYossi Boaron proposed openstack/kuryr-kubernetes master: [WIP] Services: Rollback openstack resources in case of annotation failure  https://review.openstack.org/54678408:37
openstackgerritMichał Dulko proposed openstack/kuryr-kubernetes master: Add experimental containerized jobs  https://review.openstack.org/53482708:38
openstackgerritYossi Boaron proposed openstack/kuryr-kubernetes master: Services: Rollback openstack resources in case of annotation failure  https://review.openstack.org/54678408:39
*** gcheresh_ has joined #openstack-kuryr08:41
*** salv-orlando has joined #openstack-kuryr08:52
*** salv-orlando has quit IRC08:57
*** salv-orlando has joined #openstack-kuryr08:57
*** kaliya has quit IRC09:06
*** yboaron has quit IRC09:08
*** kaliya has joined #openstack-kuryr09:12
celebdordulek: did you see my night review?09:27
celebdormorning, technically09:27
*** juriarte has quit IRC09:27
dulekcelebdor: 2 AM is night by my standards. ;)09:27
dulekcelebdor: I'm thinking it through.09:28
celebdorok09:28
dulekcelebdor: I also started to wonder about project_id where I should put the SG.09:28
dulekcelebdor: It should be the same as LB, right?09:28
celebdordulek: yes!09:29
celebdorthis is tenant specific09:29
dulekcelebdor: So I'd better add it implicitly on SG creation.09:29
*** garyloug has joined #openstack-kuryr09:30
celebdordulek: "add it implicitly"09:30
celebdordo you mean explicitly?09:30
dulekcelebdor: Ah, yes.09:31
celebdoror I just don't understand due to sleeping 3h09:31
celebdorok09:31
dulekcelebdor: Is there any better reference for python-neutronclient that reading it's code?09:33
celebdordulek: yes, of course09:34
celebdorReading nova code09:34
celebdorxD09:34
celebdorit is very sad09:34
celebdorI actually would prefer to drop python-neutronclient09:34
celebdoras it is easier to use requests directly09:35
celebdorsince the API reference documentation is good and exists09:35
celebdorin contrast to python-neutronclient09:35
dulekcelebdor: API reference you say?09:35
irenabcelebdor, planning revolutions?09:35
dulekcelebdor: https://developer.openstack.org/api-ref/network/v2/#security-groups-security-groups09:35
dulekcelebdor: Where do I put project_id on POST? Inside of security_group object or outside?09:36
*** kiennt26 has quit IRC09:36
dulekcelebdor: It's different in the table than in the example.09:36
*** threestrands_ has quit IRC09:37
dulekI'm starting to see errors like this on the Octavia gates: http://logs.openstack.org/27/534827/10/check/kuryr-kubernetes-tempest-daemon-octavia/cb782c2/controller/logs/devstacklog.txt.gz#_2018-02-22_09_03_32_79409:43
dulekThey're unrelated to Kuryr and hopefully transient, so rechecking should help.09:43
*** juriarte has joined #openstack-kuryr09:45
celebdorirenab: always09:49
celebdorI'm growing tired of OSt lib dependencies09:49
celebdorhowever misguided that may be09:49
celebdorbut maybe I'm just grumpy today09:49
celebdordulek: outside09:49
celebdorthe sg object09:49
dulekcelebdor: So same goes with name? :>09:50
celebdordulek: ok, ok09:55
celebdorpoint taken09:55
celebdorthe api reference may suck a bit09:55
dulekAPI docs suck as well. Just a bit less than python-neutronclient. ;)09:55
celebdordulek: I'll tell you my ways09:55
celebdorwhen I find something that looks fishy09:55
celebdorI use the cli like so09:55
celebdoropenstack -vvv security group create and my params09:56
celebdorand then I see the culr request it sends09:56
celebdorsince the openstack cli doens't usually fail09:56
celebdoryou get the real deal09:56
celebdorand yes. It is very sad this needs to be done09:56
celebdorbut I don't know a better way09:56
dulekSure.09:56
*** yboaron has joined #openstack-kuryr10:01
*** yamamoto has quit IRC10:06
*** juriarte has quit IRC10:09
*** juriarte has joined #openstack-kuryr10:18
irenabdmellado, celebdor can we start our chat with 10 mins delay?10:20
celebdorI suppose10:30
dmelladocelebdor: you can join now if you want10:30
celebdorokey dokey10:31
openstackgerritGenadi Chereshnya proposed openstack/kuryr-kubernetes master: Creating gate for testing ports pool feature  https://review.openstack.org/54694210:44
*** gcheresh_ has quit IRC10:50
*** salv-orl_ has joined #openstack-kuryr11:00
*** juriarte has quit IRC11:01
*** salv-orlando has quit IRC11:03
*** juriarte has joined #openstack-kuryr11:03
openstackgerritMerged openstack/kuryr-kubernetes master: TrivialFix: Make timeout a parameter in wait_for  https://review.openstack.org/54614911:05
*** yamamoto has joined #openstack-kuryr11:06
*** pcaruana has quit IRC11:07
*** yamamoto has quit IRC11:13
*** garyloug has quit IRC11:13
*** garyloug has joined #openstack-kuryr11:14
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: Translate k8s policy to SG  https://review.openstack.org/52691611:20
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: Create network policy handler and driver  https://review.openstack.org/53065511:20
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: [WIP] DisposableWatch for k8s client  https://review.openstack.org/54344811:20
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: [WIP] Annotate pods that selected by k8s label-selector  https://review.openstack.org/54566311:20
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: [WIP] Add policy security-groups driver  https://review.openstack.org/54587211:20
*** celebdor has quit IRC11:25
dulekKinda lame question but how do I create a SG with None as --protocol?11:30
*** maysamacedos has joined #openstack-kuryr11:32
*** gcheresh_ has joined #openstack-kuryr11:33
irenabdulek: https://bugs.launchpad.net/python-openstackclient/+bug/171254011:35
openstackLaunchpad bug 1712242 in python-openstackclient "duplicate for #1712540 security group rule create doesn't allow protocol-less rules" [Undecided,Fix released] - Assigned to Daniel Speichert (dasp)11:35
dulekirenab: Aww, I had too old version. Thanks!11:35
openstackgerritYossi Boaron proposed openstack/kuryr-kubernetes master: Services: Rollback openstack resources in case of annotation failure  https://review.openstack.org/54678411:39
*** celebdor has joined #openstack-kuryr11:41
openstackgerritGenadi Chereshnya proposed openstack/kuryr-kubernetes master: Creating gate for testing ports pool feature  https://review.openstack.org/54694211:53
*** gcheresh_ has quit IRC11:55
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: Create network policy handler and driver  https://review.openstack.org/53065512:03
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: [WIP] DisposableWatch for k8s client  https://review.openstack.org/54344812:03
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: [WIP] Annotate pods that selected by k8s label-selector  https://review.openstack.org/54566312:04
*** yamamoto has joined #openstack-kuryr12:09
*** pcaruana has joined #openstack-kuryr12:09
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: [WIP] Annotate pods that selected by k8s label-selector  https://review.openstack.org/54566312:12
*** yamamoto has quit IRC12:14
openstackgerritMichał Dulko proposed openstack/kuryr-kubernetes master: Services: Set SGs for N-S with haproxy provider  https://review.openstack.org/54677712:15
*** gcheresh has quit IRC12:16
openstackgerritEyal Leshem proposed openstack/kuryr-kubernetes master: [WIP] Add policy security-groups driver  https://review.openstack.org/54587212:21
*** janki has quit IRC12:27
*** yamamoto has joined #openstack-kuryr12:27
*** garyloug has quit IRC12:33
*** gcheresh_ has joined #openstack-kuryr12:45
*** garyloug has joined #openstack-kuryr12:53
*** gcheresh has joined #openstack-kuryr12:59
*** jlabarre has joined #openstack-kuryr13:01
openstackgerritDanil Golov proposed openstack/kuryr-kubernetes master: Allow passing multiple VIFs to CNI  https://review.openstack.org/47101213:01
openstackgerritDanil Golov proposed openstack/kuryr-kubernetes master: Add SR-IOV capabilities to VIF handler  https://review.openstack.org/51228013:01
openstackgerritDanil Golov proposed openstack/kuryr-kubernetes master: Add SR-IOV binding driver to CNI  https://review.openstack.org/51228113:01
openstackgerritDanil Golov proposed openstack/kuryr-kubernetes master: Add multi-vif driver to kuryr-k8s controller  https://review.openstack.org/52459013:01
openstackgerritDanil Golov proposed openstack/kuryr-kubernetes master: Allow requesting additional subnets via annotation  https://review.openstack.org/52459113:01
openstackgerritDanil Golov proposed openstack/kuryr-kubernetes master: Allow setting specific ports for SRIOV handler  https://review.openstack.org/52459213:01
*** salv-orl_ has quit IRC13:10
openstackgerritYossi Boaron proposed openstack/kuryr-kubernetes master: Services: Rollback openstack resources in case of annotation failure  https://review.openstack.org/54678413:11
*** salv-orlando has joined #openstack-kuryr13:11
*** salv-orlando has quit IRC13:15
*** salv-orlando has joined #openstack-kuryr13:15
openstackgerritDaniel Mellado proposed openstack/kuryr-kubernetes master: Creating gate for testing ports pool feature  https://review.openstack.org/54694213:26
*** yamamoto has quit IRC13:34
*** yamamoto has joined #openstack-kuryr13:44
*** yamamoto has quit IRC13:44
dulekdmellado: How about switching daemon-side VIF choice with Kuryr plugin discussion?13:51
dulekdmellado: I think the latter doesn't need 1,5 hour.13:51
dulek1.513:51
dmelladodulek: ok from my side, just switch it13:52
dulekDone.13:53
*** janonymous has quit IRC13:56
openstackgerritMichał Dulko proposed openstack/kuryr-kubernetes master: Services: Set SGs for N-S with haproxy provider  https://review.openstack.org/54677714:01
openstackgerritMichał Dulko proposed openstack/kuryr-kubernetes master: Services: Set SGs for N-S with haproxy provider  https://review.openstack.org/54677714:14
celebdordulek: sounds good to me!14:14
dulekcelebdor: Hm?14:14
dmelladocelebdor: how about slides14:17
dmellado?14:17
dmelladoxD14:17
dulekcelebdor: Ah, that. :P14:17
dulekcelebdor, dmellado: Can we figure out what we want to merge before release request?14:17
dulekPlease take a look on that: https://etherpad.openstack.org/p/kuryr-queens-release14:18
dulekirenab: ^14:18
dulekWe have two critical bugs fixed and one half-fixed (for E-W only, N-S patch is in review).14:18
dulekSo 4 patches are to be decided what do we want with them.14:20
dulekAll are listed in the etherpad.14:20
dulekSo what's critical enough to get merged?14:21
dulekI know that without https://review.openstack.org/#/c/545944/ regular StatefulSets are rather broken with kuryr-daemon. So I'd say we need this.14:22
dulek(we've discovered StatefulSet issue today, luckily patch was created previously)14:22
dmelladodulek: sorry, in meetins all day14:33
dmelladohmmm I'd consider that critical14:33
dmelladoand the N-S too14:33
dmelladowdyt celebdor ?14:33
dulekOkay, please wait a bit with 545944 then, I'll make 100% sure that it solves StatefulSets issue.14:35
dmelladodulek: awesome, thanks!14:41
dmelladobtw, which beer do you fancy for Dublin? :P14:42
dmelladoGuiness one?14:42
celebdordmellado: in regard to what?14:42
celebdordulek: did you verify the N-S one?14:43
dmelladocelebdor: so, slides, regarding the slides14:43
dmelladoand secondly, regarding release14:43
dulekdmellado: I prefer IPAs, stouts are not really my taste. ;)14:43
dulekcelebdor: Yup. Though I'd be happier if someone poked it a bit more.14:43
*** yamamoto has joined #openstack-kuryr14:45
celebdorirenab: do you think you could try dulek's patch so we can merge it before release deadline?14:49
*** yamamoto has quit IRC14:51
celebdordulek: regarding https://review.openstack.org/#/c/546777/4/kuryr_kubernetes/controller/drivers/lbaasv2.py14:53
celebdorshouldn't line 437 raise14:53
celebdorI mean, after logging we should raise, shouldn't we?14:54
celebdorah, no14:54
celebdornothing14:54
celebdorI see you return none instead14:54
celebdor:-)14:54
dulekcelebdor: Yup! That's a bad case, but worst case I get no SG or SG rule.14:54
dulekIf I raise I can get orphaned LB and Listener.14:54
*** Natanbro has quit IRC14:56
celebdorok14:56
celebdorirenab: we would also need https://review.openstack.org/#/c/545944/ merged14:58
celebdorso. To merge both on master and on queens: https://review.openstack.org/#/c/545944/ and https://review.openstack.org/#/c/54677714:58
dulekI've verified that https://review.openstack.org/#/c/545944/ solves the StatefulSets issue.15:07
openstackgerritGenadi Chereshnya proposed openstack/kuryr-kubernetes master: Creating gate for testing ports pool feature Partially-Implements: blueprint enhance-upstream-gates  https://review.openstack.org/54694215:07
celebdordulek: with openshift?15:07
celebdordulek: in that case... You should definitely add a closes-bug for that bug as well15:08
dulekcelebdor: K8s. Why OpenShift here? ltomasbo had related but different issue with OpenShift.15:08
celebdordulek: the error was reported with openshfit 3.715:08
dulekcelebdor: We have two different StatefulSet issues actuall…15:08
dulekactually.15:08
dulekcelebdor: You're talking about this one: https://bugs.launchpad.net/kuryr-kubernetes/+bug/174883615:09
openstackLaunchpad bug 1748836 in kuryr-kubernetes "SelfLink-based resource management creates the bugs" [Undecided,In progress] - Assigned to Eunsoo Park (esevan.park)15:09
celebdoryes15:09
celebdorso your patch does not solve that15:09
celebdorI presume15:09
dulekNope.15:09
dulekcelebdor: And with ltomasbo we've noticed that with kuryr-daemon `oc deploy blah` creates blah-deploy container, that has no random component in name. So doing a deployment again fails, because registry in CNI daemon doesn't expect that.15:10
dulekcelebdor: Noticing that I've thought of StatefulSets.15:10
dulekcelebdor: So it's a similar issue for upstream K8s too.15:10
dulekcelebdor: Because StatefulSets has repeatable names as well.15:11
dulekcelebdor: It's probably useful to add another bug for that?15:11
celebdordulek: wait a sec15:12
celebdorbut this your latest patch would fix the '-deploy' pods, right?15:12
dulekcelebdor: Yes, it should.15:12
celebdorso then it goes in :-)15:14
celebdorit was already marked critical as well15:15
*** caowei has quit IRC15:15
dulekcelebdor: I'll add a comment describing the problem there.15:15
dulekI mean the StatefulSet issue.15:16
dulekyboaron: Can you take a look if I addressed your comments correctly in https://review.openstack.org/#/c/546777 ?15:18
dulekyboaron: Ah, you've did. :)15:19
yboarondulek, Yep , you got my +1 - nice work !15:19
celebdordulek: thanks15:26
*** slaweq has quit IRC15:30
*** slaweq has joined #openstack-kuryr15:30
*** garyloug has quit IRC15:32
*** slaweq has quit IRC15:35
*** yboaron has quit IRC15:36
*** gcheresh_ has quit IRC15:43
*** hongbin has joined #openstack-kuryr15:46
*** yamamoto has joined #openstack-kuryr15:47
*** yamamoto has quit IRC15:51
*** garyloug has joined #openstack-kuryr15:53
dulekcelebdor: So… What's the plan? I'm not sure if irenab will be online today and we need an RC today.16:01
dulekcelebdor: I've created the release patch already: https://review.openstack.org/#/c/547005, waiting to decide where we're planning to cut off.16:01
ltomasbo dulek I really would like that it includes your bug fix...16:03
dulekltomasbo: I guess we could ask celebdor to ninja-approve this one - it's pretty self-contained. The second though touches more stuff.16:04
*** gcheresh has quit IRC16:04
ltomasboyep16:05
*** gcheresh has joined #openstack-kuryr16:05
dmelladolet me have a look too16:19
dmelladoI might have ninja powers16:19
dmelladodulek: ltomasbo16:20
dulek:D16:20
dmelladoso, dulek will you bump the hash in the end?16:20
dulekdmellado: That's the idea, I've put -W until we decide what can we get in.16:21
dmelladoso, besides the release patch, is your patch for the -deploy already merged?16:21
dulekdmellado: Nope. It's this one: https://review.openstack.org/#/c/545944/16:22
dmelladoirenab: could you please have a look at dulek 's patch? ^^16:24
dmelladodulek: from my side, get that in16:24
dmelladoand once that it's there, update the openstack/releases hash and let me know16:24
dmelladoI'll provide ptl +1 for it16:24
dulekdmellado: Sure, though I think smcginnis will need celebdor's +1, as he's Queens PTL. Anyway I have around 2 h max until I'll need to leave.16:25
dulekSo if irenab won't show up it's up to celebdor.16:25
dmelladodulek: irenab told me that she'd merge it once she gets home16:26
celebdordulek: she said she'd be online16:26
dmelladojust spoke with her16:26
dmelladoso we'd be fine16:26
dulekAwesome. I'll look back here in an hour then. :)16:26
dmelladocelebdor: worst case we'll have you be what you love most, dictator for life and so xD16:27
celebdorxD16:27
dmelladodulek: celebdor I need to go now, but ping me in case you'd need anything from me later16:27
dmelladodulek: did you verify the patch in any case?16:27
dmelladoshould we run any check before merging?16:27
dmelladoI would like to avoid a terrible revert16:27
dmellado(and note that I'm trusting you)16:28
dulekdmellado: Smallest patch is fine.16:29
dulekThe bigger one - I've tried it with a lot of cases, but still I'd be happy if someone else poked it.16:29
dmelladoack, irenab ^^ thanks!16:30
dmelladodulek: if we're formal, (and we won't be this time because of the rush)16:30
dmelladowe'd need tests for that16:30
dmelladoplease add that to your backlog16:30
dmelladoxd16:30
dmelladodulek: there you go, Irena said the same thing as I did, but it went in16:33
dmelladocelebdor: dulek, I'll let you handle this, ping me if you need my ack for something16:33
*** slaweq has joined #openstack-kuryr16:40
*** slaweq has quit IRC16:45
*** yamamoto has joined #openstack-kuryr16:48
*** yamamoto has quit IRC16:52
*** juriarte has quit IRC16:59
dulekcelebdor, irenab: Okay, so https://review.openstack.org/#/c/547005 is getting in. I've created stable/queens backport for that: https://review.openstack.org/#/c/547071/17:23
*** janonymous has joined #openstack-kuryr17:27
celebdordulek: we still need the N-S, right?17:44
dulekcelebdor: Would be great to have but it's still not accepted on master.17:45
*** garyloug has quit IRC17:46
openstackgerritMerged openstack/kuryr-kubernetes master: Delete pods from registry in CNI daemon  https://review.openstack.org/54594417:46
*** yamamoto has joined #openstack-kuryr17:48
*** yamamoto has quit IRC17:54
openstackgerritAntoni Segura Puimedon proposed openstack/kuryr-tempest-plugin master: test_container: Add non-layered demo test image  https://review.openstack.org/54708817:59
celebdor:O18:00
celebdorhttps://hub.docker.com/r/kuryr/demo/tags/18:00
celebdordulek: dmellado: ltomasbo: compressed it is actually only 7MiB18:00
openstackgerritAntoni Segura Puimedon proposed openstack/kuryr-tempest-plugin master: test_container: Add non-layered demo test image  https://review.openstack.org/54708818:06
dulekcelebdor: I got to go now. https://review.openstack.org/#/c/547071/ is going through Jenkins, then it'll need to go once again through gate once approved.18:06
celebdorok18:06
dulekcelebdor: Once it merges hash in https://review.openstack.org/#/c/547005 will need to be adjusted to point to merge-commit with that patch on stable/queens.18:07
celebdordulek: I'll still try to get irenab to merge the N-S18:07
dulek(TBH I'm not 100% sure about that merge-commit, maybe it'll not happen)18:07
celebdorbut I need to take off for 1h18:08
dulekcelebdor: Okay, then let me create a backport for you already. ;)18:08
celebdorthanks18:08
dulekcelebdor: smcginnis wrote yesterday evening that he expects releases in 24h, but I guess he can be persuaded to accept a later one. :)18:08
dulekcelebdor: #openstack-release is the channel to ask, I guess. ;)18:09
openstackgerritMichał Dulko proposed openstack/kuryr-kubernetes master: Services: Set SGs for N-S with haproxy provider  https://review.openstack.org/54677718:12
irenabdulek: looking into the patch now18:12
dulekirenab: Awww. I've overwritten it. Just a second…18:13
openstackgerritMichał Dulko proposed openstack/kuryr-kubernetes master: Services: Set SGs for N-S with haproxy provider  https://review.openstack.org/54677718:13
dulekirenab: Okay, that's the one.18:14
irenabok18:14
dulekcelebdor, irenab: And this is the cherry-pick: https://review.openstack.org/#/c/547090/18:15
dulekOkay, gotta go, I'll be able to periodically check IRC on my phone.18:16
*** dulek_ has joined #openstack-kuryr18:21
irenabdulek: ping18:21
irenabwhat happens if service port is changed by the user after service is created? How is it handled?18:22
dulek_lbaas.py should handle this.18:22
dulek_And run ensure_*18:22
dulek_So SG for that port will be added.18:24
dulek_SG rule.18:24
irenabtrying to track how the SG rule is modified18:25
*** dulek__ has joined #openstack-kuryr18:26
dulek__Rather a new one is added.18:27
irenabthe old one is not removed, correct?18:27
*** dulek_ has quit IRC18:31
*** openstackgerrit has quit IRC18:33
dulek__irenab: Hm... Correct. Some more elaborated logic in ensure_security_group_rules could help with that.18:37
dulek__Hm, this might be hard to do. We cannot do transactions on Neutron API.18:38
dulek__We might end up screwing up if ensure is run in parallel.18:39
*** slaweq has joined #openstack-kuryr18:39
dulek__Hm, though it probably cant...18:39
irenabI do not think it can18:41
irenabits a bit of security issue to allow both ports but we can deal with it later, just need to add bug to trace it18:42
*** slaweq has quit IRC18:43
irenabthe patch is also lacking additional unit tests, so maybe add it to the same bug?18:43
irenabotherwise lgtm18:44
irenabdulek__: celebdor : running devstack with the patch for dragonflow backend, will plus +2 once verify. Please add bug for the k8s service port update and additional tests.18:45
*** yamamoto has joined #openstack-kuryr18:50
*** yamamoto has quit IRC18:56
dulek__Thanks irenab!18:57
*** yamamoto has joined #openstack-kuryr18:59
*** slaweq has joined #openstack-kuryr19:00
*** gcheresh has quit IRC19:03
*** yamamoto has quit IRC19:04
*** slaweq has quit IRC19:05
dulek__irenab: This backport also needs +2: https://review.openstack.org/#/c/547071/119:11
irenabdulek__: for some reason it didn’t work for DF, but I am not sure it is related to your patch, maybe DF issue, will have to debug. Lets merge it as long as you sure it fixed the native ovs19:14
*** yamamoto has joined #openstack-kuryr19:14
irenabI see the additional SG on VIP port, so it seems ok from the configuration point of view19:15
dulek__Yes, it should have additional. Plus rules are okay?19:16
irenabyes19:16
dulek__:)19:17
dulek__You have the default SG as Kuryr's SG?19:17
dulek__If not you may have issue with the Kuryr SG rules. I fought with remote_group_id all morning.19:18
* dulek__ will need to leave phone for next 30 minutes.19:18
*** yamamoto has quit IRC19:19
*** yamamoto has joined #openstack-kuryr19:29
*** yamamoto has quit IRC19:34
*** janonymous has quit IRC19:36
*** yamamoto has joined #openstack-kuryr19:44
*** yamamoto has quit IRC19:52
*** slaweq has joined #openstack-kuryr20:01
*** slaweq has quit IRC20:06
dulek__irenab: There's one last backport - the deletion of pods in CNI daemon: https://review.openstack.org/#/c/54707120:24
*** slaweq_ has joined #openstack-kuryr20:24
*** slaweq_ has quit IRC20:29
*** openstackgerrit has joined #openstack-kuryr20:37
openstackgerritMerged openstack/kuryr-kubernetes master: Services: Set SGs for N-S with haproxy provider  https://review.openstack.org/54677720:37
*** yamamoto has joined #openstack-kuryr20:48
*** yamamoto has quit IRC20:55
*** slaweq_ has joined #openstack-kuryr21:09
*** slaweq_ has quit IRC21:14
*** dulek__ has quit IRC21:31
*** pcaruana has quit IRC21:37
*** slaweq_ has joined #openstack-kuryr21:46
*** slaweq_ has quit IRC21:50
*** yamamoto has joined #openstack-kuryr21:53
*** yamamoto has quit IRC21:57
*** yamamoto has joined #openstack-kuryr22:01
*** slaweq_ has joined #openstack-kuryr22:18
*** slaweq_ has quit IRC22:23
*** yamamoto has quit IRC23:02
*** slaweq has joined #openstack-kuryr23:33
*** hongbin has quit IRC23:34
*** slaweq has quit IRC23:39
« Wednesday, 2018-02-21 Index Friday, 2018-02-23 »

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