Tuesday, 2016-02-09

*** armax has quit IRC00:19
*** armax_ has joined #openstack-neutron-ovn00:19
*** allan_h has joined #openstack-neutron-ovn00:55
*** allan_h has quit IRC01:12
Sam-I-Amwhy am i still here01:58
russellbbecause you love it01:58
russellband it's cold outside01:59
*** yamamoto has joined #openstack-neutron-ovn01:59
Sam-I-Amit was nice here today. i got outside for a bit.01:59
Sam-I-Amnow i'm back to troubleshooting what appears to be a lab network problem that has set back most of this afternoons progress on anything01:59
Sam-I-Amgit is broken01:59
Sam-I-Amand nothing uses that, luckily01:59
Sam-I-Amchanged git_base to use http and now tox blows up02:00
Sam-I-Amwas also looking at more mtu business in create_network02:01
russellbyaks everywhere02:05
Sam-I-Amyep nothing is working with vagrant anymore02:06
Sam-I-Amso my lab is dead02:06
Sam-I-Amon that note, it becomes tomorrow's problem02:07
Sam-I-Amhave a good night. dont work too hard.02:09
russellbwe'll see about that02:11
russellbi can't seem to feel caught up these days02:11
*** yamamoto has quit IRC02:24
*** gangil has quit IRC02:29
*** s3wong has quit IRC02:44
*** armax_ is now known as armax02:54
*** gangil has joined #openstack-neutron-ovn03:09
*** gangil has joined #openstack-neutron-ovn03:09
*** yamamoto has joined #openstack-neutron-ovn03:12
openstackgerritlitong01 proposed openstack/networking-ovn: Vagrant: make provider network creation configurable  https://review.openstack.org/27647504:16
*** armax_ has joined #openstack-neutron-ovn04:18
*** armax has quit IRC04:21
*** armax_ is now known as armax04:21
*** brad_behle has quit IRC04:31
*** salv-orl_ has joined #openstack-neutron-ovn04:41
*** salv-orlando has quit IRC04:43
*** allan_h has joined #openstack-neutron-ovn04:44
*** allan_h has quit IRC05:28
*** jckasper has quit IRC05:40
*** chandrav has joined #openstack-neutron-ovn06:11
*** chandrav has quit IRC06:12
*** numans has joined #openstack-neutron-ovn06:19
*** armax has quit IRC06:49
*** jckasper has joined #openstack-neutron-ovn07:41
*** jckasper has quit IRC07:45
*** gangil has quit IRC08:14
*** salv-orlando has joined #openstack-neutron-ovn10:40
*** salv-orl_ has quit IRC10:43
*** yamamoto has quit IRC11:05
openstackgerritNuman Siddique proposed openstack/networking-ovn: Refactor ovn_nb_sync.OvnNbSynchronizer  https://review.openstack.org/27779211:29
openstackgerritNuman Siddique proposed openstack/networking-ovn: Refactor ovn_nb_sync.OvnNbSynchronizer  https://review.openstack.org/27779211:37
*** jckasper has joined #openstack-neutron-ovn11:43
*** jckasper has quit IRC11:47
*** yamamoto has joined #openstack-neutron-ovn11:51
openstackgerritNuman Siddique proposed openstack/networking-ovn: Neutron ovn northbound db sync tool  https://review.openstack.org/27780512:03
openstackgerritNuman Siddique proposed openstack/networking-ovn: Neutron ovn northbound db sync tool  https://review.openstack.org/27780512:06
*** chandrav has joined #openstack-neutron-ovn12:09
*** rtheis has joined #openstack-neutron-ovn12:23
*** openstackgerrit_ has joined #openstack-neutron-ovn12:39
*** yamamoto has quit IRC12:46
*** numans has quit IRC13:02
*** yamamoto has joined #openstack-neutron-ovn13:03
*** yamamoto has quit IRC13:16
*** brad_behle has joined #openstack-neutron-ovn13:22
*** yamamoto has joined #openstack-neutron-ovn13:23
russellbrtheis: you around?  there seems to be a failing unit test related to AZs13:28
rtheisrussellb: yes I am13:29
russellbi first saw the failure here https://review.openstack.org/#/c/277792/213:29
*** numans has joined #openstack-neutron-ovn13:29
russellbbut i can reproduce locally without any patches applied13:29
rtheisrussellb: does it fail consistently?13:30
* russellb checks13:30
russellbyes, fails every time for me13:31
*** regXboi has joined #openstack-neutron-ovn13:31
russellbi had to get a fresh tox env to reproduce it though (tox -r -e py27, or just rm -rf .tox)13:31
russellbwhich makes it sound related to a fairly recent change in neutron13:32
rtheisrussellb: yeah, I'll take a look13:32
russellbok thank you!13:32
*** numans has quit IRC13:34
regXboirussellb: morning - I *think* I've figured out what's going on with column tracking, but if I'm correct, it looks like there is an off by one error in the code that loops through tracked rows13:38
russellbit could be, i'm not sure that code is used anywhere yet13:39
regXboiI've dropped in instrumentation to check how much it improves13:39
russellbthough there is test coverage13:39
russellbsee tests/test-ovsdb.c and search for TRACKED13:40
regXboiyes, I did13:40
regXboibut here's why I say there is an off-by-one13:40
regXboiI dropped in instrumentation in lflow.c to check the number of flows looped through13:40
regXboiand after an initial stack, that instrumentation says 1213:40
regXboibut, when I put in the code to loop through tracked rows, that instrumentation says 13 *AND* I see occasional WARN messages about not being able to process filters13:41
regXboiwhich leads me to think there is an off-by-one13:41
regXboiI'm trying to finish getting the seqno checking right and then I'll chase this off-by-one some more13:42
*** jckasper has joined #openstack-neutron-ovn13:42
*** jckasper_ has joined #openstack-neutron-ovn13:43
rtheisrussellb: I think the failure is caused by https://github.com/openstack/neutron/commit/62c0c0771fb1e46b275ab9694ae94cbada6b922c13:44
*** jckasper has quit IRC13:47
rtheisanyone seeing "/opt/stack/networking-ovn/devstack/plugin.sh: line 86: /opt/stack/neutron/devstack/lib/ovs: No such file or directory" when they try to stack.sh with OVN?13:48
regXboirtheis: no, but I don't think I'm quite tip of the tree anymore13:48
russellbrtheis: you need to update neutron for that error13:48
russellbrtheis: we're now using some devstack code in the neutron repo13:49
russellbrtheis: and "the test just got added" would explain it :)13:49
rtheisrussellb: this is a fresh stack...no neutron yet13:49
russellbwell... we may have broken things then13:50
rtheisLooks like I don't even have /opt/stack/neutron13:50
regXboimoar yelling?13:51
Sam-I-Amrussellb: yeah i saw that last night13:51
russellbrtheis: we must assume neutron is already there, which is true for the gate env13:51
* regXboi is guessing he won't update code basis today13:51
Sam-I-Amwhat changed that would break this?13:51
regXboiSam-I-Am: did you just ask what changed to break devstack?13:52
russellbthat one.13:52
russellbeasy fix, one sec ...13:53
Sam-I-Amohhhh yeah13:53
Sam-I-Amthat would do it13:53
* Sam-I-Am looks at blame13:54
*** openstackgerrit_ has quit IRC13:56
Sam-I-Amare we just debugging why this patch doesnt work, or reverting it?14:01
russellbi know the problem14:02
russellbi'll post a patch in a sec14:02
russellbtrying to work out how to get it to merge14:02
russellbsince there's also a new unit test added in neutron that's failing14:02
Sam-I-Amthere inlies the fun problem14:02
Sam-I-Amyeah i saw that too14:02
Sam-I-Amthis sounds like a good reason to go back to bed14:02
openstackgerritRichard Theis proposed openstack/networking-ovn: WIP: Add extension fields on network create  https://review.openstack.org/27784414:04
openstackgerritRussell Bryant proposed openstack/networking-ovn: Disable failing new unit test.  https://review.openstack.org/27784514:05
openstackgerritRussell Bryant proposed openstack/networking-ovn: devstack: Source ovs devstack lib later.  https://review.openstack.org/27784614:05
*** numans has joined #openstack-neutron-ovn14:12
regXboirussellb: it's not an off-by-one - it's a problem with processing a delete twice14:18
*** numans has quit IRC14:20
*** numans has joined #openstack-neutron-ovn14:22
regXboior maybe I should say, not releasing a row once it is deleted14:22
regXboiwell... now *that's* annoying14:29
regXboithe logical datapath check is getting in my way14:30
*** yamamoto has quit IRC14:32
*** yamamoto has joined #openstack-neutron-ovn14:33
*** yamamoto has quit IRC14:37
*** tongli has joined #openstack-neutron-ovn14:40
regXboirussellb: what would you say if I said I wanted to take out the logical datapath check in lflow_run in favor of seqno tracking?14:56
openstackgerritRichard Theis proposed openstack/networking-ovn: WIP: Add extension fields on network and port create  https://review.openstack.org/27784415:03
*** yamamoto has joined #openstack-neutron-ovn15:07
*** shettyg has joined #openstack-neutron-ovn15:11
*** palexster has quit IRC15:12
*** palexster has joined #openstack-neutron-ovn15:14
Sam-I-Amrussellb: huh @ fail on 277845 - secgroup ext not enabled?15:17
Sam-I-Amer, wait... a bit further down15:17
russellbsome ipv6 test fail, no idea15:18
*** chandrav has quit IRC15:18
Sam-I-Ammight be worth a recheck15:19
russellbwe've got a db issue in the q-svc log that we still need to fix15:19
*** numans has quit IRC15:20
Sam-I-Amwonder whats up with that15:23
*** numans has joined #openstack-neutron-ovn15:23
*** palexster has quit IRC15:24
regXboithe tracking code gets rows LIFO, which means the sequence numbers aren't monotonically increasing15:32
*** palexster has joined #openstack-neutron-ovn15:39
regXboiso that snippet of code gets more complex now15:41
*** numans has quit IRC15:43
*** yamamoto has quit IRC15:48
Sam-I-Amrussellb: 277846 passed but has a variety of interesting tbacks15:57
* russellb nods15:57
russellbthe db tracebacks are fairly consistently there i think15:57
Sam-I-Amthe equivalent neutron/ml2 jobs dont seem to have them15:58
* russellb nods15:59
*** armax has joined #openstack-neutron-ovn16:08
*** palexster has quit IRC16:34
*** salv-orl_ has joined #openstack-neutron-ovn16:41
regXboiha ha... ok, this works16:44
*** salv-orlando has quit IRC16:44
regXboinow to run a test to see how well it does and then update the patch set16:44
* regXboi dances a slow jig16:44
*** c_z has joined #openstack-neutron-ovn17:12
*** allan_h has joined #openstack-neutron-ovn17:20
*** c_z has quit IRC17:23
-openstackstatus- NOTICE: Gerrit is restarting now, to alleviate current performance impact and WebUI errors.17:24
*** c_z has joined #openstack-neutron-ovn17:26
*** palexster has joined #openstack-neutron-ovn17:31
rtheisTesting out provider networks by following http://docs.openstack.org/developer/networking-ovn/testing.html ... however I didn't "observe that ovn-controller automatically created patch ports between br-int and br-provider".  Is the documentation still accurate?17:32
russellbrtheis: it is not accurate, thanks17:35
Sam-I-Amisnt that the ovs-vsctl step?17:36
openstackgerritRussell Bryant proposed openstack/networking-ovn: testing: Remove comment about looking at patch ports.  https://review.openstack.org/27798517:37
russellbrtheis: ^17:37
russellbSam-I-Am: the patch ports now aren't created until later, and it's more of an implementation detail anyway..17:37
*** c_z has quit IRC17:40
*** palexster has quit IRC17:46
*** palexster has joined #openstack-neutron-ovn18:04
russellbrtheis: patch ports will appear once you start creating neutron ports, if you want to look18:07
russellbbut it doesn't really matter18:07
rtheisrussellb: thanks, I'll take a look18:07
russellbwell, it matters a lot, but i'm not sure it's important to look at18:07
russellbdepends how much you care about diving into how OVN is setting up what you asked for18:08
*** palexster has quit IRC18:14
rtheisindeed they appear after creating a subnet and port for the provider network.18:16
*** gangil has joined #openstack-neutron-ovn18:20
*** gangil has joined #openstack-neutron-ovn18:20
openstackgerritRichard Theis proposed openstack/networking-ovn: Add extension fields on network and port create  https://review.openstack.org/27784418:20
openstackgerritRussell Bryant proposed openstack/networking-ovn: devstack: Source ovs devstack lib later.  https://review.openstack.org/27784618:24
*** palexster has joined #openstack-neutron-ovn18:27
*** aginwala has joined #openstack-neutron-ovn18:32
*** aginwala has quit IRC18:45
*** c_z has joined #openstack-neutron-ovn18:46
*** arosen12 has joined #openstack-neutron-ovn18:48
*** aginwala has joined #openstack-neutron-ovn18:49
*** gangil has quit IRC19:02
*** palexster has quit IRC19:03
*** gangil has joined #openstack-neutron-ovn19:11
*** gangil has joined #openstack-neutron-ovn19:11
*** jckasper_ has quit IRC19:13
*** jckasper has joined #openstack-neutron-ovn19:14
*** jckasper has quit IRC19:14
*** jckasper has joined #openstack-neutron-ovn19:14
*** Mic22 has quit IRC19:23
*** c_z has quit IRC19:32
*** aginwala has quit IRC19:49
*** aginwala has joined #openstack-neutron-ovn19:55
*** jckasper has quit IRC20:19
*** jckasper has joined #openstack-neutron-ovn20:19
*** jckasper has quit IRC20:23
*** salv-orl_ has quit IRC20:36
*** aginwala has quit IRC20:40
*** aginwala has joined #openstack-neutron-ovn20:43
*** jckasper has joined #openstack-neutron-ovn20:44
*** salv-orlando has joined #openstack-neutron-ovn20:53
*** aginwala has quit IRC20:57
tongli@Sam-I-Am, matt, is the gate back up now?21:02
*** jckasper has quit IRC21:06
*** jckasper has joined #openstack-neutron-ovn21:06
Sam-I-Amtongli: no21:07
*** jckasper has quit IRC21:08
*** jckasper has joined #openstack-neutron-ovn21:09
mesterysalv-orlando tongli: I believe based on what rtheis said we need https://review.openstack.org/#/c/277844/ and https://review.openstack.org/#/c/277846/ to merge to fix the gate21:13
mesteryNot salv-orlando but salv-orlando21:13
mesteryNot salv-orlando or salv-orlando but Sam-I-Am21:14
russellbarmax: lol21:14
mesteryStupid autocorrect21:14
mesteryrussellb: armax? :)21:14
russellbpatch to fix is in the gate21:14
Sam-I-Amit'll be fixed soonish i hope21:14
Sam-I-Amin the meantime plenty of other stuff that i've neglected21:14
armaxwhat did I do?21:14
russellb277844 indeed21:14
russellbarmax: mestery may have summoned you by accident.21:14
tongli@Sam-I-Am, thanks.21:15
*** Mic22 has joined #openstack-neutron-ovn21:15
russellb277844 is running gate jobs, should be finished in < 30 mins21:15
armaxok, I need to change nick21:15
armaxthis world is too small for me21:15
mesteryarmax: There's so much love for you in all the channels :)21:16
armaxI kno21:16
armaxdespite the many attempts to get people to hate me21:16
mesteryarmax: There's just too much of you to love ;)21:16
armaxmestery: ditto21:17
Sam-I-Amtongli: your recheck wont work21:17
Sam-I-Amtongli: not until those patches merge and you rebase21:17
tongli@Sam-I-Am, yeah, I thought the gate is up so I did that before I asked. I probably should have waited a bit longer.21:18
tongli@Sam-I-Am, sorry about that.21:18
Sam-I-Amno prob21:18
openstackgerritMerged openstack/networking-ovn: Add extension fields on network and port create  https://review.openstack.org/27784421:19
mesteryrussellb: I'd like to create a job for kuryr+ovn, even if it's just an install job, to ensure things work there. Thoughts?21:21
russellbmestery: sounds great.21:21
russellband i'm fine with an install job to start with21:21
mesteryrussellb: Thinking of putting it into the networking-ovn job list21:21
russellbnothing < install job < actually using the thing21:21
russellbfine with me21:21
Sam-I-Amwould it use conventional l3?21:21
Sam-I-Ami also have ideas for a native l3 job21:22
mesterySam-I-Am: For now, it would just install :)21:22
Sam-I-Ami was trying to use the kuryr vagrant yesterday and its broken21:23
Sam-I-Amat least with vbox21:24
Sam-I-Amthen i lost git access from the lab21:24
mesteryWFM? :)21:25
openstackgerritMerged openstack/networking-ovn: devstack: Source ovs devstack lib later.  https://review.openstack.org/27784621:37
mesteryrussellb Sam-I-Am: https://review.openstack.org/27810121:39
Sam-I-Ami thought i was working on that21:41
mesterySam-I-Am: You were, but I got antsy and had to do something while sitting in all these meetings :)21:42
Sam-I-Amhmm ok21:42
Sam-I-Amalso was thinking it'd go in ovn21:42
mesteryWait, isn't it in ovn?21:42
* regXboi goes to look at the sun21:42
Sam-I-Amerrr... non-voting/experimental21:43
Sam-I-Amone step down :)21:43
*** regXboi has quit IRC21:43
mesterySam-I-Am: This is what happens when I roll patches while in meetings :)21:44
* Sam-I-Am backs away slowly21:45
*** azbiswas has joined #openstack-neutron-ovn21:48
*** aginwala has joined #openstack-neutron-ovn22:33
*** shettyg has quit IRC22:38
*** salv-orl_ has joined #openstack-neutron-ovn22:40
*** salv-orlando has quit IRC22:43
*** rtheis has quit IRC22:46
*** tongli has quit IRC22:56
*** jckasper has quit IRC23:04
*** jckasper has joined #openstack-neutron-ovn23:05
*** aginwala has quit IRC23:14
*** aginwala has joined #openstack-neutron-ovn23:16
*** jckasper has quit IRC23:17
*** jckasper has joined #openstack-neutron-ovn23:17
*** jckasper has quit IRC23:18
*** jckasper has joined #openstack-neutron-ovn23:18
*** yamamoto_ has joined #openstack-neutron-ovn23:21
*** yamamoto_ has quit IRC23:23
*** gangil1 has joined #openstack-neutron-ovn23:33
*** gangil has quit IRC23:33
*** allan_h has quit IRC23:39
*** azbiswas has quit IRC23:44
*** yamamoto_ has joined #openstack-neutron-ovn23:50

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!