Wednesday, 2016-08-10

*** apearson has joined #openstack-powervm01:35
*** svenkat has joined #openstack-powervm01:45
*** tblakes has quit IRC02:04
*** kriskend has joined #openstack-powervm02:07
*** kriskend_ has joined #openstack-powervm02:07
*** arunman has joined #openstack-powervm02:10
*** seroyer has joined #openstack-powervm02:34
*** arunman has quit IRC02:45
*** kriskend has quit IRC03:04
*** kriskend_ has quit IRC03:04
*** esberglu has joined #openstack-powervm03:11
*** esberglu has quit IRC03:11
*** tjakobs has joined #openstack-powervm03:20
*** seroyer has quit IRC03:28
*** svenkat has quit IRC03:29
*** apearson has quit IRC04:42
*** apearson has joined #openstack-powervm04:47
*** tjakobs has quit IRC04:50
*** apearson has quit IRC04:59
-openstackstatus- NOTICE: zuul is being restarted to reload configuration. Jobs should be re-enqueued but if you're missing anything (and it's not on please issue a recheck in 30min.05:25
*** kotra03 has joined #openstack-powervm05:28
*** arunman has joined #openstack-powervm05:59
*** openstackgerrit has quit IRC08:18
*** openstackgerrit has joined #openstack-powervm08:18
openstackgerritArun Mani proposed openstack/nova-powervm: Deploy of a VM occasionally fails due to invalid keystone token
*** smatzek has joined #openstack-powervm11:01
*** svenkat has joined #openstack-powervm11:23
*** kotra03 has quit IRC11:58
*** arunman has quit IRC12:17
*** burgerk has joined #openstack-powervm12:19
*** edmondsw has joined #openstack-powervm13:00
*** tblakes has joined #openstack-powervm13:08
*** mdrabe has joined #openstack-powervm13:09
*** tblakes has quit IRC13:12
*** tblakes has joined #openstack-powervm13:13
*** kylek3h has joined #openstack-powervm13:17
*** tblakes has quit IRC13:25
*** arunman has joined #openstack-powervm13:31
*** apearson has joined #openstack-powervm13:37
*** burgerk has quit IRC13:40
*** esberglu has joined #openstack-powervm13:42
*** kotra03 has joined #openstack-powervm13:44
*** thorst_ has joined #openstack-powervm13:53
arunmanthorst_:, will it get +2'd today?  :-)14:03
arunmanI've fixed the comments from efried and updated the patchset14:04
*** dwayne has joined #openstack-powervm14:07
*** catintheroof has joined #openstack-powervm14:09
*** seroyer has joined #openstack-powervm14:09
*** tjakobs has joined #openstack-powervm14:11
thorst_arunman: looks like there was one more comment from efried?14:12
thorst_docstring for method?14:12
thorst_o wait, I was looking at wrong patch level14:12
thorst_did you re-cherry pick back to stable mitaka?14:13
*** tblakes has joined #openstack-powervm14:13
*** tblakes_ has joined #openstack-powervm14:15
*** tblakes has quit IRC14:17
*** tblakes_ is now known as tblakes14:17
*** lmtaylor has joined #openstack-powervm14:18
*** efried has joined #openstack-powervm14:22
*** burgerk has joined #openstack-powervm14:22
*** kriskend has joined #openstack-powervm14:25
arunmanthorst_: Not yet. I'm waiting for a +2 in master before I can do a re-cheery pick to mitaka. But, yeah.. can do that soon enough14:27
thorst_go ahead and do that now14:27
thorst_it'll just update14:27
*** kotra03 has quit IRC14:30
*** mdrabe has quit IRC14:31
*** mdrabe has joined #openstack-powervm14:47
arunmanthorst_: Done14:50
arunmanefried: I got your comments incorporated in Hope it's OK now14:52
efriedarunman, roger that, will look in 514:52
efriedadreznec, thorst_: do we not have sonar checking enabled for *-powervm?14:55
adreznecThere's no external sonar, that's not part of the openstack CI infrastructure14:55
efriedarunman, done14:55
adreznecIf we wanted that efried we'd have to set it up ourselves14:55
efriedadreznec, okay.14:55
efriedI'm not going to push for it.  Just need to decide how rigorous we should be on sonar-ish rules in reviews.14:56
adreznecefried: Down with sonar, up with complexity!14:56
*** lmtaylor2 has joined #openstack-powervm15:08
*** lmtaylor has quit IRC15:11
*** miltonm has joined #openstack-powervm15:19
*** smatzek has quit IRC15:32
*** kotra03 has joined #openstack-powervm15:33
*** kotra03 has quit IRC15:35
thorst_esberglu: just verifying that you are redeploying CI now?16:11
esbergluthorst_: I’m working on a stacking issue that will break the deploy16:11
thorst_o boy!16:12
thorst_good luck  :-)16:12
esbergluthorst: I have a workaround that completes the stack but I’m trying to find a proper fix. I could deploy with the workaround16:13
efriedkriskend/seroyer, was it y'all who were advocating for SR-IOV physical port MTU, port switch mode, and flow control?  REST changes (3727) and pypowervm changes (3732) are in the pipe.16:14
thorst_esberglu: where is the said 'workaround'?16:15
kriskendefried: yeah16:15
esbergluthorst_: Reverting a change in devstack16:16
thorst_esberglu: ruh roh16:16
thorst_can you paste it here (adreznec)16:16
efriedesberglu I still failed to stack having reverted that change16:16
efried2016-08-10 16:16:21.649 | ++lib/keystone:create_keystone_accounts:383  openstack project show admin -f value -c id16:17
efried2016-08-10 16:16:23.824 | Discovering versions from the identity service failed when creating the password plugin. Attempting to determine version from URL.16:17
efried2016-08-10 16:16:23.824 | Could not determine a suitable URL for the plugin16:17
efried2016-08-10 16:16:23.858 | +lib/keystone:create_keystone_accounts:383  admin_project=16:17
efried2016-08-10 16:16:23.862 | +lib/keystone:create_keystone_accounts:1   exit_trap16:17
esbergluHmm. Worked for me.16:17
efriedwho tf knows if it's the same problem.16:17
esbergluefried: I got that but when I was using master devstack but stable/mitaka for other stuff16:18
thorst_esberglu: so the devstack failed for mitaka, but not master?16:18
efriedesberglu everything's master in my env.16:19
efriedthorst_, fyi, the delta in question (supposedly) is
esbergluefried: Idk then. I’ve seen that error pop up from other stuff in the past too16:20
thorst_so basically, they're creating initial networks no matter what.16:21
adreznecthorst_: Yes16:21
thorst_and they assume we will always have l3 now...16:21
thorst_which we don't16:21
efriedI thought we did.16:22
adreznecNo, there's still an if neutron-l3 check there for the l3 process startup... but I think some plugins support it for l216:23
openstackLaunchpad bug 1604768 in devstack "lib/neutron doesn't create initial networks unless neutron-l3 is enabled" [Undecided,Fix released] - Assigned to YAMAMOTO Takashi (yamamoto)16:23
adreznecThat's the bug they were 'fixing'16:23
thorst_which seems like not a proper bug...16:23
adreznecIn a patch comment when asked about non-l3 plugin support, committer replied 'services/l3 has its own checks for q-l3/neutron-l3 and extensions.'16:24
thorst_efried / esberglu: Can you guys post to pastebin where we fail?16:24
adreznecI wonder if we're failing that check with our plugin somehow16:24
adreznecOr rather, accidentally passing that check16:25
thorst_adreznec: define 'plugin' here?  A devstack plugin or a neutron plugin?16:25
thorst_two very different plugins16:25
thorst_we don't have our own neutron plugin...16:25
adreznecwell, plugin/agent16:26
adreznecMaybe it assumes that all ML2 agents support l3?16:26
adreznecWhich would be a bad assumption16:26
thorst_but the comment in the review says that they support where we don't16:26
thorst_I think we need to be telling it to make a private network16:26
thorst_which is why I want the full call stack16:26
thorst_we have to get is_provider_network to return false16:27
thorst_and get is_networking_extension_supported to return false.16:27
thorst_efried / esberglu: that all make sense?16:28
efriedthorst_, nope, not a bit.16:28
thorst_make sure Q_USE_PROVIDER_NETWORKING = FALSE16:28
thorst_also look at what's going on in devstack's is_networking_extension_supported function.  We need to make sure that when SEA is enabled that returns False when 'Router' is passed in16:29
thorst_which is in l3 in the devstack project, line 37616:39
esbergluthorst_: Sorry got pulled away helping a guy with a novalink install. Here’s the trace16:42
esberglu2016-08-10 16:42:06.842 | [Call Trace]16:42
esberglu2016-08-10 16:42:06.842 | ./
esberglu2016-08-10 16:42:06.842 | /opt/stack/devstack/lib/neutron:558:start_neutron_new16:42
esberglu2016-08-10 16:42:06.842 | /opt/stack/devstack/lib/neutron:440:create_neutron_initial_network16:42
esberglu2016-08-10 16:42:06.842 | /opt/stack/devstack/lib/neutron_plugins/services/l3:145:die_if_not_set16:42
esberglu2016-08-10 16:42:06.842 | /opt/stack/devstack/functions-common:211:die16:43
esberglu2016-08-10 16:42:06.844 | [ERROR] /opt/stack/devstack/functions-common:145 Failure retrieving project_id for demo16:43
thorst_why aren't we creating the project 'demo'16:43
esbergluAnd that’s with Q_USE_PROVIDER_NETWORKING = False16:43
thorst_esberglu: do we override OS_PROJECT_NAME?16:45
thorst_is the partial devstack'd system still available?16:46
esbergluthorst_: I was hitting on neo1416:48
esbergluThe control node stacked fine, not the compute nodes16:48
thorst_on just a compute node...16:48
thorst_now THAT is interesting.16:48
thorst_I wonder if the accidentally broke separation of compute / controller.16:48
*** smatzek has joined #openstack-powervm16:51
thorst_weird...the openstack project list still works there.16:52
thorst_I'm not sure why this is happening.16:52
*** smatzek has quit IRC16:52
*** thorst_ has quit IRC16:58
*** jwcroppe has quit IRC17:16
*** jwcroppe has joined #openstack-powervm17:16
*** thorst_ has joined #openstack-powervm17:32
*** thorst_ has quit IRC17:37
*** thorst_ has joined #openstack-powervm17:40
arunmanthorst_: Need your +2 for stable-mitaka too, , Please add it when you can today :-)17:48
thorst_arunman: Why does it say uploaded patch set 717:48
thorst_that looks odd.17:48
arunmanWell.. I tried to abandon by mistake and then did a restore. That's the reason17:50
thorst_so can you just go to the original change set17:50
thorst_hit the cherry-pick button17:50
*** smatzek has joined #openstack-powervm17:51
thorst_and just type stable/mitaka17:51
thorst_it will upload the latest patch set.17:51
thorst_you don't need to abandon previous17:51
*** smatzek has quit IRC17:51
arunmanI did that, and now it says Updated PatchSet 8.17:52
thorst_I'll w+1 it when we get through the CI17:53
*** arunman has quit IRC18:09
*** thorst_ has quit IRC18:12
*** thorst_ has joined #openstack-powervm18:14
*** catintheroof has quit IRC18:34
openstackgerritMerged openstack/nova-powervm: Added revert to delete LPAR from destination host on failed rebuild
thorst_tblakes: you see the error here?
thorst_are you driving that?19:26
*** tblakes_ has joined #openstack-powervm19:27
tblakes_Yep I'm fixing that now19:27
esbergluthorst_: So setting Q_USE_PROVIDER_NETWORKING ensured is_provider_network returns false19:30
esbergluBut I can’t get is_networking_extension_supported to return false for router.19:30
*** tblakes has quit IRC19:30
esbergluDo you know how the ext-list is generated?19:30
*** tblakes_ is now known as tblakes19:31
thorst_esberglu: let me look19:31
esbergluAll I need to do to get that function to return false is remove router from that list19:32
thorst_esberglu: what was the neo?19:34
thorst_esberglu: I think on the controller node...19:39
thorst_service_plugins =
thorst_I'm wondering if that being set is causing the router to be active.19:39
thorst_esberglu: It looks like in order for that to be set, your controller would have needed the L3 agent to be set to true.  I wonder if you have an inconsistency between controller and compute19:50
thorst_esberglu: lol...I think I found it19:51
thorst_q-l3 ==> neutron-l319:51
esbergluDid the name change or did we just have it wrong?19:51
thorst_q-dhcp ==> neutron-dhcp19:51
thorst_they changed the names19:51
thorst_so all our local.confs need to be updated.19:51
thorst_I *think*19:52
thorst_it kinda looks like they're in transition19:52
thorst_but q-l3 is also used in many places still19:53
adreznecFarewall, quantum!19:53
thorst_so...looks like we have some local.conf work for master branch things19:54
esbergluIf it’s still using q-l3 in some places will it work to just disable both?19:54
thorst_esberglu: yep19:54
adreznecI was wondering when the new neutron-lib stuff would come around to bite us19:54
thorst_disable both19:54
thorst_that'll give you consistency back to mitaka as well19:55
*** thorst_ has quit IRC20:02
*** thorst_ has joined #openstack-powervm20:08
*** thorst_ has quit IRC20:08
*** thorst_ has joined #openstack-powervm20:08
*** svenkat has quit IRC21:10
*** lmtaylor2 has left #openstack-powervm21:10
*** apearson has quit IRC21:19
*** jwcroppe has quit IRC21:28
*** jwcroppe has joined #openstack-powervm21:28
*** jwcroppe has quit IRC21:29
esbergluthorst_: Still no luck after adding the neutron-* stuff21:31
thorst_not super sure to be honest...21:31
esbergluMind taking a look at the conf if you have a minute?21:31
thorst_I'm doing some devstacks myself I will hopefully hit it as well21:32
thorst_send to me in an e-mail, I'll see if I can tonight21:32
*** burgerk has quit IRC21:44
*** esberglu has quit IRC21:48
*** tjakobs has quit IRC21:53
*** kriskend has quit IRC21:56
*** seroyer has quit IRC22:00
*** mdrabe has quit IRC22:14
*** kylek3h has quit IRC22:20
*** jwcroppe has joined #openstack-powervm22:29
*** thorst_ has quit IRC22:32
*** jwcroppe has quit IRC22:34
*** edmondsw has quit IRC22:36
*** tblakes has quit IRC22:39
*** thorst_ has joined #openstack-powervm22:49
*** thorst_ has quit IRC22:54
*** jwcroppe has joined #openstack-powervm23:00

Generated by 2.14.0 by Marius Gedminas - find it at!