| *** carl_baldwin has quit IRC | 00:01 | |
| *** julim has quit IRC | 00:10 | |
| *** mlavalle has quit IRC | 00:13 | |
| *** matsuhashi has joined #openstack-neutron | 00:15 | |
| *** nati_uen_ has quit IRC | 00:40 | |
| *** emagana has quit IRC | 00:44 | |
| *** yamahata_ has joined #openstack-neutron | 00:48 | |
| *** idella4 has joined #openstack-neutron | 00:54 | |
| *** LiJiansheng has joined #openstack-neutron | 00:55 | |
| idella4 | oh anteaya | 00:56 |
|---|---|---|
| idella4 | I tried re-reading our query exchange from last night (my time) but it seems my system corrupted some files | 00:58 |
| idella4 | possibly in the process of the last run of building neutron and running the testsuite | 00:58 |
| idella4 | and it seems the log never got written to the hard drive | 00:59 |
| idella4 | however, note here | 00:59 |
| *** steven-weston has quit IRC | 01:00 | |
| idella4 | <anteaya^M^C> well it can't hurt to add them to your bug report | 01:00 |
| *** matsuhashi has quit IRC | 01:01 | |
| *** matsuhashi has joined #openstack-neutron | 01:02 | |
| idella4 | I got to add here, it achieves nothing bloating a bug report with more samples already declared invalid | 01:02 |
| *** matsuhashi has quit IRC | 01:02 | |
| *** matsuhashi has joined #openstack-neutron | 01:02 | |
| markmcclain | how are you running these tests? | 01:03 |
| markmcclain | on a machine with tox are you invoking the test runner directly? | 01:04 |
| idella4 | dorectly | 01:04 |
| idella4 | tox is toxic | 01:04 |
| idella4 | directly | 01:04 |
| idella4 | quite simply testr init; \ testr run | 01:05 |
| markmcclain | tox is hardly toxic | 01:05 |
| anteaya | hey idella4 | 01:05 |
| idella4 | it is in gentoo but never mind | 01:05 |
| idella4 | hi | 01:05 |
| idella4 | I have been thru that lots already | 01:06 |
| *** morganfainberg has joined #openstack-neutron | 01:06 | |
| idella4 | anteaya: here knows all about it | 01:06 |
| anteaya | well markmcclain is the PTL of the project | 01:06 |
| idella4 | hi markmcclain | 01:06 |
| anteaya | so yes he would be aware of what is happening in neutron | 01:06 |
| markmcclain | idella4: hi | 01:06 |
| idella4 | I'm a gentoo dev | 01:06 |
| idella4 | attempting to add the testsuite to neutron's ebuild | 01:07 |
| idella4 | day 5 at this point | 01:07 |
| idella4 | making new records for degree of difficulty | 01:07 |
| morganfainberg | anteaya, *waves* | 01:07 |
| anteaya | hey morganfainberg | 01:07 |
| morganfainberg | anteaya, lurking here now as well. | 01:07 |
| anteaya | morganfainberg: welcome | 01:08 |
| anteaya | and thank you | 01:08 |
| anteaya | :D | 01:08 |
| markmcclain | idella4: some I'm assuming a lot of this is settling on the right set up deps | 01:08 |
| markmcclain | because you're not using pip to install python deps? | 01:08 |
| idella4 | well nota at all | 01:09 |
| idella4 | all the deps are system installed and they're all fine | 01:10 |
| markmcclain | ls | 01:11 |
| markmcclain | oops :) | 01:11 |
| *** gongysh has quit IRC | 01:11 | |
| idella4 | that's a fundamental long ago sorted | 01:11 |
| idella4 | @ https://bugs.launchpad.net/neutron/+bug/1251657 | 01:12 |
| * markmcclain reads | 01:13 | |
| idella4 | ok | 01:14 |
| markmcclain | so testr is running the tests out of order | 01:16 |
| idella4 | is it? | 01:16 |
| markmcclain | I thought that we had fixed the bad mock in the nicira tests | 01:17 |
| idella4 | no idea | 01:17 |
| idella4 | eeer hm | 01:17 |
| anteaya | yes | 01:17 |
| markmcclain | yeah in parallel mode it should distribute work | 01:17 |
| anteaya | testr runs the tests in any order | 01:17 |
| markmcclain | and optimize the runs over time | 01:17 |
| idella4 | ok you have a tanundrum | 01:17 |
| anteaya | nosetests runs in a specific order, so the move from nose to testr can expose some setup/teardown issues | 01:18 |
| anteaya | if there are any | 01:18 |
| markmcclain | and we've found a few | 01:18 |
| markmcclain | I thought we had mostly found them all | 01:18 |
| idella4 | with the --parallel option the suite pulls some mem alloc errors | 01:18 |
| idella4 | I assume simply because it's a monster | 01:19 |
| markmcclain | py2.6 or 2.7? | 01:19 |
| idella4 | so I chucked the use of the --parallel option | 01:19 |
| idella4 | 2.7 | 01:19 |
| idella4 | not using py2.6 on it at all | 01:19 |
| markmcclain | ok.. 2.7 is a bit more memory sane | 01:19 |
| idella4 | hmm yep | 01:19 |
| markmcclain | 2.6 unit tests can bloat really quickly | 01:20 |
| idella4 | yep, well 2.6 isn't an issue here since we're not using it | 01:20 |
| markmcclain | when you mentioned mem problems I had to ask | 01:20 |
| idella4 | sure; chucking the --parallel option leaves 'me' with these 2 fails | 01:21 |
| idella4 | and they appear erroneous | 01:21 |
| markmcclain | what level of concurrency are you running with parallel | 01:21 |
| markmcclain | have you tried with a smaller value than the autodetected? | 01:21 |
| idella4 | ok you've gone over my head here slightly but easily 'sorted' | 01:22 |
| idella4 | I have CFLAGS="-march=athlon64 -pipe -fomit-frame-pointer -O2" | 01:23 |
| idella4 | MAKEOPTS="-j5" | 01:23 |
| idella4 | I think the -j5 is answering your query here | 01:23 |
| idella4 | the parallel does what the --parallel daos | 01:23 |
| markmcclain | yeah well testr has it's own method to determine the number of parallel test processes to run | 01:23 |
| idella4 | these settings say what the system is doing | 01:24 |
| idella4 | well it seems to me to run them vi nosetests | 01:24 |
| idella4 | will SOLVE EVERYTHING | 01:24 |
| *** jsergent has quit IRC | 01:25 | |
| idella4 | When I asked your counterpart to consider that he retorted with | 01:25 |
| markmcclain | hmm… interesting that nose is being invoked | 01:25 |
| markmcclain | we remove nose usage >12 mos ago | 01:25 |
| idella4 | we use tox, so use tox, | 01:25 |
| *** networkstatic is now known as networkstatic_zZ | 01:25 | |
| idella4 | hmm well a test runner is but a runner | 01:25 |
| markmcclain | tox is unrelated to nose vs testr | 01:26 |
| idella4 | I am sure testr can be kicked into shape | 01:26 |
| idella4 | well ye it is | 01:26 |
| markmcclain | tox just automates the invocation of the chosen runner | 01:26 |
| idella4 | I get the gist of that now, more so anyways | 01:26 |
| *** nati_ueno has joined #openstack-neutron | 01:26 | |
| markmcclain | and manages the building of virtualenvs for testing | 01:26 |
| idella4 | yep, well | 01:26 |
| idella4 | the whole making of the virtualenvs thing simply confounds gentoo's approach to the core | 01:27 |
| markmcclain | the virtualenvs are only for isolated testing | 01:27 |
| idella4 | the use of tox is discouraged in the pytho-gentoo team, though not banned | 01:27 |
| markmcclain | but you're trying to validate the system install | 01:28 |
| markmcclain | which is a different need | 01:28 |
| idella4 | well I don;t think so | 01:28 |
| idella4 | in an ebuild you have a src_install() { | 01:28 |
| idella4 | along with the various pre src phase | 01:28 |
| idella4 | that culminate in an install | 01:29 |
| idella4 | the running of the test phase is kind of a case in its own right | 01:29 |
| idella4 | I have indeed another example of this in cinder I think it is | 01:29 |
| markmcclain | ok. and src_test runs before src_install? | 01:29 |
| *** armax has left #openstack-neutron | 01:29 | |
| idella4 | yep | 01:29 |
| idella4 | you got it | 01:29 |
| idella4 | and why tox is 'toxic' to gentoo stems from gentoo's style or policy | 01:30 |
| idella4 | in running an ebuild | 01:30 |
| idella4 | beyond downl;oading of the package and any patches or whatever | 01:31 |
| idella4 | NO CONNECT TO THE INTERNET | 01:31 |
| markmcclain | yeah.. tox typically needs to d/l stuff so I get that policy | 01:31 |
| idella4 | and along comes tox and promptly downloads everything all over again in a virtual env | 01:31 |
| idella4 | so syre it's a gentoo thingy | 01:31 |
| *** networkstatic_zZ is now known as networkstatic | 01:32 | |
| idella4 | you're follwoing this very well markmcclain | 01:32 |
| idella4 | while the re-use of nose is somewhat antagonistic to your abandinning of it | 01:32 |
| idella4 | I get that | 01:32 |
| idella4 | testr ought be ab le to be corrected | 01:33 |
| idella4 | only | 01:33 |
| idella4 | so much to get out | 01:33 |
| markmcclain | I actually still use nose from time to time.. old habits are hard to break :) | 01:33 |
| idella4 | we have 2 ebuilds of neutron-2013.2 | 01:33 |
| idella4 | sure | 01:33 |
| idella4 | I am currently re-running the neutron-2013.9999 ebuild | 01:34 |
| idella4 | which translated means it's utilising the live repo ar launchpad or git or wherever it is | 01:34 |
| idella4 | thos mght finally come up clean | 01:35 |
| idella4 | this mght finally come up clean because | 01:35 |
| idella4 | it like;ly has that fix that you cited many lines ago | 01:35 |
| idella4 | the neutron-2013.2.ebuild is using the tarball os the havana release | 01:35 |
| idella4 | you follow? | 01:36 |
| markmcclain | yes | 01:36 |
| idella4 | the neutron-2013.2.ebuild yielded those 2 errors | 01:36 |
| idella4 | and really 2 little errors out of a suite of over 12ooo tests | 01:36 |
| idella4 | top job | 01:36 |
| markmcclain | yeah.. but still 2 too many | 01:37 |
| idella4 | running without the --parallel option takes it seems around 45 minutes | 01:37 |
| markmcclain | yeah.. that's expected for the havana release | 01:37 |
| markmcclain | we're working to reduce some of hte testing overhead for icehouse to speed things up | 01:38 |
| idella4 | before I added the patch from a bug I've not yet outline | 01:38 |
| idella4 | it's reduced from 2 hours 4 min. | 01:38 |
| idella4 | not until you told me some 10 or so minutes ago did I know of this other fix | 01:39 |
| idella4 | of the ordering of the tests by testr | 01:40 |
| markmcclain | wow didn't know they were pushing 2hrs for some people | 01:40 |
| idella4 | and I'd say the 2 errors are a nonsense too | 01:40 |
| idella4 | well, that's what it took on my gentoo system here | 01:40 |
| idella4 | and I gather much of that was awaiting stupid timeouts | 01:41 |
| idella4 | my suggestion is to simply carve up the tests into some groups | 01:42 |
| markmcclain | we've been heading direction, but it will take some time | 01:42 |
| idella4 | run the groups in a series, they all get done | 01:42 |
| idella4 | ah good | 01:42 |
| markmcclain | a lot of the vendor driver tests are in subdirs of neutron/tests/unit/ | 01:43 |
| markmcclain | the *.py files in that dir are tests against core code | 01:43 |
| idella4 | so I have mostly been running the 2013.2 neutron ebuild on the assumption it's pretty much the equiv of what's in the live repo havana tarball | 01:44 |
| idella4 | this neutron has kind of caught me out here | 01:44 |
| idella4 | it's been patched and fixed at least one test related bug, the json serialization issue | 01:45 |
| markmcclain | this script: http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/sys-cluster/neutron/neutron-2013.2.ebuild?view=markup ? | 01:45 |
| idella4 | well time to twiddle the thunmb until this run completes | 01:45 |
| idella4 | let's look | 01:46 |
| idella4 | yes | 01:46 |
| idella4 | bnefore I started adding the test stuff | 01:46 |
| idella4 | here | 01:47 |
| idella4 | @ http://codepad.org/xMPOHaH1 | 01:48 |
| idella4 | from testuser@archtester ~ $ wgetpaste -s codepad ~/cvsPortage/gentoo-x86/sys-cluster/neutron/neutron-2013.2-r1.ebuild | 01:48 |
| markmcclain | thing look right | 01:51 |
| *** wenjianhn has joined #openstack-neutron | 01:51 | |
| idella4 | I've run it a few times now, see I have added the patch to fix the json thingy | 01:54 |
| markmcclain | yes.. I'm seeing | 01:54 |
| markmcclain | also noticed in the report that you had a run that asked for password | 01:54 |
| idella4 | well, if you have one that fixes the ncira thingy, | 01:55 |
| markmcclain | for future ref that occurs when mock hasn't been setup for a system util | 01:55 |
| idella4 | I that came from running by nosetests | 01:55 |
| markmcclain | the test suite shouldn't be invoking system utils | 01:55 |
| markmcclain | so looking I'm looking in the history to find the commit for nicira | 01:56 |
| idella4 | ok well that could mb come from use of nosetests | 01:56 |
| idella4 | ok | 01:56 |
| markmcclain | nose or testr | 01:56 |
| markmcclain | that is a symptom of improper mocking | 01:56 |
| idella4 | initially I was running them by nosetests | 01:56 |
| markmcclain | a password prompt should never pop up | 01:56 |
| idella4 | it wasn't until I lobbed in here and talked with anteaya a little | 01:57 |
| idella4 | by which time I had figure to run testr directly | 01:57 |
| markmcclain | ok.. so maybe the improper mocking of the nicira edge-1 hasn't merged | 01:57 |
| idella4 | testr is kind of new to me and I only have grasped it | 01:58 |
| idella4 | exactly | 01:58 |
| idella4 | I have run the 2013-9999.ebuild once or twice | 01:58 |
| markmcclain | if you skip this directory: | 01:58 |
| markmcclain | neutron/tests/unit/nicira/vshield | 01:58 |
| idella4 | and I've run so many I can't remember what was what | 01:59 |
| markmcclain | you won't invoke any of the tests that try to test the edge in NSX | 01:59 |
| idella4 | well that's what I'd normally do to attempt to isolate it | 01:59 |
| idella4 | so simply mv the folder out of there? | 02:00 |
| idella4 | that's what I did with test_ovs_tunnel.py | 02:00 |
| markmcclain | another trick if you're running testr in a env that where you've already done a testr run is pass —failing | 02:00 |
| idella4 | oh did that | 02:01 |
| markmcclain | k | 02:01 |
| idella4 | and they kept on passing them | 02:01 |
| idella4 | persistently which says to me they're getting muddled in the full run | 02:01 |
| markmcclain | that can happen | 02:01 |
| idella4 | I haven't yet seen a run of testr run --failing fail the said test | 02:02 |
| idella4 | well I'd say it did | 02:02 |
| idella4 | well it explains it in any case | 02:02 |
| idella4 | so simply mv the folder out of there like I did with test_ovs_tunnel.py | 02:03 |
| markmcclain | yeah… that will buy some time to further explore the issue | 02:03 |
| markmcclain | nicira is also a paid support product, so if folks want gentoo+nsx they can also help debug the build issue | 02:04 |
| markmcclain | I know they've helped with other failures on other distros | 02:04 |
| markmcclain | the more difficult trick will be to reduce the bug down to something that reproducible for fixing | 02:05 |
| idella4 | right so with the use of our gentoo USE flag, | 02:06 |
| idella4 | we could add to the ebuild a USE flag called something like ncira or nsx or such | 02:06 |
| markmcclain | yeah | 02:06 |
| idella4 | and make the leaving of the neutron/tests/unit/nicira/vshield conditional on the use of the USE flag | 02:07 |
| idella4 | well that was hard!! | 02:07 |
| idella4 | well @ day 5 and a whisker away from a clean run | 02:08 |
| idella4 | and making a workable test phase for neutron-2013.2* ebuilds | 02:08 |
| idella4 | what a day | 02:08 |
| markmcclain | sorry it was harder than it should have been… it's dinner time for me, but I'll make a note to build the tarball in the morning and make sure that bit rot hasn't already crept in after 30 days | 02:09 |
| idella4 | excellent | 02:09 |
| idella4 | I have Tue Nov 19 10:09:42 WST 2013 | 02:09 |
| markmcclain | also 2013.2.1 is due out around Dec 12th | 02:10 |
| idella4 | I'm quite accustomed to being in the day in everyone else's night | 02:10 |
| idella4 | ah righty | 02:10 |
| markmcclain | so it's good to find any build bugs before then too | 02:10 |
| idella4 | just before my birthday | 02:10 |
| idella4 | that too | 02:10 |
| anteaya | awesome | 02:10 |
| anteaya | glad you two had a chance to chat | 02:10 |
| markmcclain | hopefully it will be a good birthday present and you can remove the special build rules | 02:11 |
| idella4 | me, oh you're too kind | 02:11 |
| idella4 | yep | 02:11 |
| markmcclain | idella4: have a good rest of your day | 02:11 |
| idella4 | than ks | 02:11 |
| idella4 | bon appetite, especially if you're in Quebec | 02:11 |
| *** changbl has quit IRC | 02:12 | |
| anteaya | south eastern States | 02:12 |
| anteaya | are you in Atlanta, markmcclain? | 02:12 |
| idella4 | ah right | 02:12 |
| anteaya | I'm in Canada | 02:12 |
| anteaya | and also calling it a night | 02:12 |
| idella4 | not quite as southern as lil' ol' me | 02:12 |
| anteaya | do you have enough direction to entertain you for a while? | 02:12 |
| anteaya | :D | 02:13 |
| idella4 | er yes | 02:13 |
| anteaya | cool | 02:13 |
| idella4 | for a while | 02:13 |
| anteaya | sounds good | 02:13 |
| anteaya | have fun, I will look to talk to you in about 12 hours or so | 02:13 |
| idella4 | roger that Ontario | 02:13 |
| idella4 | 'night | 02:14 |
| anteaya | night | 02:15 |
| *** b3nt_pin has quit IRC | 02:23 | |
| *** amotoki has joined #openstack-neutron | 02:27 | |
| *** ljjjustin has joined #openstack-neutron | 02:33 | |
| *** yamahata_ has quit IRC | 02:44 | |
| *** wenjianhn has quit IRC | 02:47 | |
| *** wenjianhn has joined #openstack-neutron | 02:47 | |
| *** nati_ueno has quit IRC | 02:47 | |
| *** changbl has joined #openstack-neutron | 02:51 | |
| *** amritanshu_RnD has joined #openstack-neutron | 02:54 | |
| *** carl_baldwin has joined #openstack-neutron | 03:01 | |
| *** alagalah_ has quit IRC | 03:04 | |
| *** alagalah has quit IRC | 03:09 | |
| *** salv-orlando has quit IRC | 03:13 | |
| *** coolsvap has joined #openstack-neutron | 03:15 | |
| *** suresh12 has quit IRC | 03:16 | |
| *** suresh12 has joined #openstack-neutron | 03:16 | |
| *** suresh12 has quit IRC | 03:20 | |
| *** alagalah has joined #openstack-neutron | 03:23 | |
| *** matsuhashi has quit IRC | 03:29 | |
| *** matsuhashi has joined #openstack-neutron | 03:30 | |
| *** changbl has quit IRC | 03:31 | |
| *** matsuhashi has quit IRC | 03:35 | |
| *** krast has joined #openstack-neutron | 03:41 | |
| *** krast has quit IRC | 03:41 | |
| *** nati_ueno has joined #openstack-neutron | 03:41 | |
| *** alagalah has quit IRC | 03:45 | |
| *** dkliban has joined #openstack-neutron | 03:47 | |
| *** carl_baldwin has quit IRC | 03:52 | |
| *** nati_ueno has quit IRC | 04:06 | |
| *** nati_ueno has joined #openstack-neutron | 04:06 | |
| *** gdubreui has quit IRC | 04:12 | |
| *** gdubreui has joined #openstack-neutron | 04:12 | |
| *** gdubreui has quit IRC | 04:13 | |
| *** gdubreui has joined #openstack-neutron | 04:13 | |
| idella4 | crrrrikey; Ran 18185 tests in 5986.417s \ FAILED (id=0, failures=1, skips=328)\ | 04:19 |
| idella4 | don't ya just hate that | 04:19 |
| openstackgerrit | Dane LeBlanc proposed a change to openstack/neutron: Cisco nexus plugin fails to untrunk vlan if other hosts using vlan https://review.openstack.org/54612 | 04:20 |
| *** suresh12 has joined #openstack-neutron | 04:27 | |
| *** suresh12 has quit IRC | 04:32 | |
| *** dcahill has quit IRC | 04:41 | |
| *** matsuhashi has joined #openstack-neutron | 04:46 | |
| *** balar has quit IRC | 04:55 | |
| *** nati_uen_ has joined #openstack-neutron | 05:00 | |
| *** nati_ueno has quit IRC | 05:03 | |
| *** nati_uen_ is now known as nati_ueno | 05:04 | |
| *** mihgen has joined #openstack-neutron | 05:06 | |
| *** LiJiansheng has quit IRC | 05:14 | |
| *** jecarey has quit IRC | 05:18 | |
| *** Jianyong has joined #openstack-neutron | 05:18 | |
| *** Jianyong has quit IRC | 05:25 | |
| *** Jianyong has joined #openstack-neutron | 05:30 | |
| *** mihgen has quit IRC | 05:35 | |
| *** changbl has joined #openstack-neutron | 05:38 | |
| *** steven-weston has joined #openstack-neutron | 05:44 | |
| *** LiJiansheng has joined #openstack-neutron | 05:44 | |
| *** bvandenh has joined #openstack-neutron | 05:47 | |
| *** mihgen has joined #openstack-neutron | 05:53 | |
| *** suresh12 has joined #openstack-neutron | 06:03 | |
| prometheanfire | idella4: yes, yes I do | 06:15 |
| idella4 | huh | 06:15 |
| prometheanfire | huh | 06:18 |
| *** gongysh has joined #openstack-neutron | 06:20 | |
| *** prometheanfire has quit IRC | 06:24 | |
| openstackgerrit | Dane LeBlanc proposed a change to openstack/neutron: Improve unit test coverage for Cisco plugin nexus code https://review.openstack.org/57100 | 06:24 |
| *** prometheanfire has joined #openstack-neutron | 06:26 | |
| *** suresh12 has quit IRC | 06:37 | |
| *** gongysh has quit IRC | 06:37 | |
| *** suresh12 has joined #openstack-neutron | 06:38 | |
| *** matsuhashi has quit IRC | 06:41 | |
| *** matsuhashi has joined #openstack-neutron | 06:42 | |
| *** suresh12 has quit IRC | 06:42 | |
| *** suresh12 has joined #openstack-neutron | 06:45 | |
| *** matsuhashi has quit IRC | 06:46 | |
| *** matsuhashi has joined #openstack-neutron | 06:48 | |
| *** jianingy has joined #openstack-neutron | 06:48 | |
| *** suresh12 has quit IRC | 06:51 | |
| *** suresh12 has joined #openstack-neutron | 06:51 | |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/56864 | 06:52 |
| *** suresh12 has quit IRC | 06:55 | |
| *** matsuhashi has quit IRC | 07:09 | |
| *** matsuhashi has joined #openstack-neutron | 07:10 | |
| *** matsuhas_ has joined #openstack-neutron | 07:13 | |
| *** matsuhashi has quit IRC | 07:14 | |
| *** gdubreui has quit IRC | 07:18 | |
| *** jprovazn has joined #openstack-neutron | 07:31 | |
| openstackgerrit | Evgeny Fedoruk proposed a change to openstack/neutron: Adds tests, fixes Radware LBaaS driver as a result https://review.openstack.org/54155 | 07:34 |
| *** amuller has joined #openstack-neutron | 07:39 | |
| openstackgerrit | Zhang Hua proposed a change to openstack/neutron: Fwaas can't run in operating system without namespace feature https://review.openstack.org/56893 | 07:44 |
| openstackgerrit | Zhongyue Luo proposed a change to openstack/neutron: (WIP) DO NOT MERGE https://review.openstack.org/57111 | 07:51 |
| *** suresh12 has joined #openstack-neutron | 08:02 | |
| *** mihgen has quit IRC | 08:03 | |
| *** marun has quit IRC | 08:06 | |
| *** suresh12 has quit IRC | 08:07 | |
| *** z0rkito has quit IRC | 08:07 | |
| *** z0rkito has joined #openstack-neutron | 08:10 | |
| *** jlibosva has joined #openstack-neutron | 08:11 | |
| *** mestery_ has joined #openstack-neutron | 08:15 | |
| *** mestery has quit IRC | 08:19 | |
| *** marun has joined #openstack-neutron | 08:24 | |
| *** marun has quit IRC | 08:25 | |
| *** coolsvap has quit IRC | 08:30 | |
| *** afazekas_ has joined #openstack-neutron | 08:34 | |
| *** mihgen has joined #openstack-neutron | 08:40 | |
| *** steven-weston has quit IRC | 08:41 | |
| openstackgerrit | Isaku Yamahata proposed a change to openstack/neutron: Implement service vm framework(WIP): https://review.openstack.org/56892 | 08:46 |
| *** enikanorov_ has quit IRC | 08:53 | |
| *** enikanorov_ has joined #openstack-neutron | 08:53 | |
| openstackgerrit | Sascha Peilicke proposed a change to openstack/neutron: Support building wheels (PEP-427) https://review.openstack.org/57124 | 08:57 |
| *** nati_ueno has quit IRC | 08:59 | |
| *** yamahata_ has joined #openstack-neutron | 09:01 | |
| *** salv-orlando has joined #openstack-neutron | 09:05 | |
| *** rossella_s has joined #openstack-neutron | 09:08 | |
| *** ygbo has joined #openstack-neutron | 09:09 | |
| *** jpich has joined #openstack-neutron | 09:09 | |
| *** matsuhas_ has quit IRC | 09:25 | |
| *** matsuhashi has joined #openstack-neutron | 09:25 | |
| openstackgerrit | Sascha Peilicke proposed a change to openstack/python-neutronclient: Support building wheels (PEP-427) https://review.openstack.org/57153 | 09:27 |
| *** yamahata_ has quit IRC | 09:30 | |
| *** matsuhashi has quit IRC | 09:30 | |
| *** matsuhashi has joined #openstack-neutron | 09:30 | |
| sgran | anyone feel like helping me debug an sqlalchemy problem? | 09:35 |
| sgran | adding this: http://paste.debian.net/66578/ | 09:36 |
| sgran | results in: | 09:37 |
| sgran | InvalidRequestError: Table 'sessionpersistences' is already defined for this MetaData instance. Specify 'extend_existing=True' to redefine options and columns on an existing Table object. | 09:37 |
| sgran | my sqlalchemy knowledge is not good enough to figure this one out | 09:38 |
| *** alexpilotti has joined #openstack-neutron | 09:51 | |
| *** networkstatic has quit IRC | 09:52 | |
| *** LiJiansheng has quit IRC | 09:55 | |
| enikanorov_ | sgran: i' | 09:59 |
| enikanorov_ | sgran: i'll look into it | 09:59 |
| sgran | great, thanks | 10:00 |
| enikanorov_ | can you upload your patch to gerrit so i could just pull it and run it myself? | 10:01 |
| enikanorov_ | (you can mark it as WIP) | 10:01 |
| sgran | sure | 10:02 |
| sgran | sec | 10:02 |
| openstackgerrit | Stephen Gran proposed a change to openstack/neutron: Enforce unique constraint on neutron pool members https://review.openstack.org/56815 | 10:03 |
| *** coolsvap has joined #openstack-neutron | 10:03 | |
| sgran | enikanorov_: https://review.openstack.org/56815 | 10:03 |
| enikanorov_ | yeah, pulling | 10:03 |
| sgran | thanks | 10:04 |
| *** feleouet has joined #openstack-neutron | 10:08 | |
| *** nati_ueno has joined #openstack-neutron | 10:09 | |
| *** nati_ueno has quit IRC | 10:11 | |
| *** LiJiansheng has joined #openstack-neutron | 10:11 | |
| *** jprovazn has quit IRC | 10:16 | |
| *** Jianyong has quit IRC | 10:22 | |
| idella4 | what's happened @ https://bugs.launchpad.net/neutron/+bug/1251657 so I can't add any further comment "Post comment" cmd button is greyed out | 10:29 |
| *** yamahata_ has joined #openstack-neutron | 10:38 | |
| *** jprovazn has joined #openstack-neutron | 10:41 | |
| sgran | status has been marked as invalid, at a guess | 10:44 |
| idella4 | I added to it already under invalid | 10:46 |
| idella4 | and it should NOT be invalid | 10:47 |
| idella4 | the 2 wrongs still don't make a right | 10:47 |
| idella4 | prometheanfire: also added to it post set to invalid | 10:48 |
| idella4 | but as you said, at a guess, and that's a fair one, I guess | 10:49 |
| *** marun has joined #openstack-neutron | 10:49 | |
| *** marun has quit IRC | 10:49 | |
| *** safchain has joined #openstack-neutron | 10:50 | |
| *** jlibosva is now known as jlibosva|1unch | 10:57 | |
| *** matsuhashi has quit IRC | 10:58 | |
| *** matsuhashi has joined #openstack-neutron | 10:59 | |
| enikanorov_ | sgran: replied in the review. I think i got rid of sqlalchemy errors in your patch | 11:01 |
| enikanorov_ | by putting comma after UniqueConstraint | 11:01 |
| sgran | aha! | 11:01 |
| sgran | let me try | 11:01 |
| sgran | thanks | 11:02 |
| enikanorov_ | which means that we specify a tuple for __table_args__, not a single object | 11:02 |
| *** jlibosva|1unch is now known as jlibosva | 11:03 | |
| *** matsuhashi has quit IRC | 11:04 | |
| *** coolsvap has quit IRC | 11:07 | |
| *** amritanshu_RnD has quit IRC | 11:23 | |
| *** enikanorov_ has quit IRC | 11:31 | |
| *** marun has joined #openstack-neutron | 11:31 | |
| *** enikanorov_ has joined #openstack-neutron | 11:31 | |
| *** pcm_ has joined #openstack-neutron | 11:34 | |
| *** pcm_ has quit IRC | 11:34 | |
| *** alagalah has joined #openstack-neutron | 11:34 | |
| *** pcm_ has joined #openstack-neutron | 11:35 | |
| sgran | right, makes sense | 11:35 |
| *** enikanorov_ has quit IRC | 11:39 | |
| *** enikanorov_ has joined #openstack-neutron | 11:40 | |
| *** enikanorov_ has quit IRC | 11:48 | |
| *** enikanorov_ has joined #openstack-neutron | 11:48 | |
| *** ben_duyujie has joined #openstack-neutron | 11:48 | |
| *** amotoki has quit IRC | 12:03 | |
| *** kashyap has quit IRC | 12:08 | |
| *** yfried has joined #openstack-neutron | 12:13 | |
| *** kashyap has joined #openstack-neutron | 12:13 | |
| yfried | marun: why isn't it affecting my rhos setup? | 12:14 |
| marun | it was removed in icehouse | 12:14 |
| marun | rhos is havana | 12:14 |
| marun | An fyi for everyone - we have a regression wrt security groups: https://bugs.launchpad.net/devstack/+bug/1252620 | 12:14 |
| yfried | marun: devstack is Icehouse? | 12:15 |
| marun | yfried: Your test that validates security group behavior should be merged but skipped with a reference to the bug in question. | 12:15 |
| marun | yfried: When the bug gets fixed, the test can be unskipped | 12:15 |
| marun | yfried: The upstream gate is master, which became icehouse mid-Oct after rc havana was branched | 12:16 |
| marun | 327384 | 12:17 |
| yfried | marun: maybe just remove the negative checks and wait for the bug to be fixed before submitting them separately? | 12:17 |
| * marun shakes fist at yubikey | 12:17 | |
| marun | yfried: no, having them in is useful | 12:17 |
| marun | yfried: people who are working on the fix then have something to test against | 12:17 |
| marun | yfried: and people know not to reimplement the test independently | 12:17 |
| yfried | ok. can you show me how to mark them for skipping? | 12:18 |
| marun | yes, i'll find an example | 12:18 |
| marun | yfried: check tempest/scenario/test_stamp_pattern.py:144 | 12:19 |
| marun | yfried: there's a decorator that allows referencing a bug number | 12:20 |
| yfried | and it's for the whole test? can't I put it only on the check_block function? | 12:20 |
| marun | it's for the whole test | 12:21 |
| marun | there's no way to skip a partial test | 12:21 |
| marun | hmm | 12:21 |
| marun | which review is affected again? | 12:21 |
| yfried | marun: https://review.openstack.org/#/c/55101/ | 12:22 |
| marun | Is there any point in running the test if security groups are inoperative? | 12:23 |
| *** enikanorov_ has quit IRC | 12:24 | |
| *** enikanorov_ has joined #openstack-neutron | 12:24 | |
| *** nati_ueno has joined #openstack-neutron | 12:24 | |
| *** enikanorov__ has joined #openstack-neutron | 12:26 | |
| *** enikanorov_ has quit IRC | 12:26 | |
| *** salv-orlando has quit IRC | 12:34 | |
| *** salv-orlando has joined #openstack-neutron | 12:43 | |
| marun | markmcclain: ping | 12:57 |
| *** gfa has joined #openstack-neutron | 13:00 | |
| *** jlibosva has quit IRC | 13:00 | |
| *** afazekas_ has quit IRC | 13:02 | |
| *** bvandenh has quit IRC | 13:08 | |
| *** afazekas_ has joined #openstack-neutron | 13:14 | |
| *** richardboswell has quit IRC | 13:14 | |
| *** salv-orlando has quit IRC | 13:17 | |
| *** marun has quit IRC | 13:20 | |
| *** richardboswell has joined #openstack-neutron | 13:22 | |
| *** jlibosva has joined #openstack-neutron | 13:24 | |
| *** richardboswell has quit IRC | 13:36 | |
| *** richardboswell has joined #openstack-neutron | 13:37 | |
| *** bvandenh has joined #openstack-neutron | 13:39 | |
| *** yamahata_ has quit IRC | 13:48 | |
| *** yamahata_ has joined #openstack-neutron | 13:49 | |
| *** marun has joined #openstack-neutron | 13:49 | |
| *** wenjianhn has quit IRC | 13:51 | |
| *** afazekas_ has quit IRC | 14:02 | |
| *** coolsvap has joined #openstack-neutron | 14:02 | |
| *** dkliban has quit IRC | 14:07 | |
| *** julim has joined #openstack-neutron | 14:12 | |
| *** ben_duyujie has quit IRC | 14:14 | |
| *** yamahata_ has quit IRC | 14:18 | |
| *** carl_baldwin has joined #openstack-neutron | 14:19 | |
| sgran | if I want code into icehouse, when is the deadline? | 14:22 |
| *** yamahata_ has joined #openstack-neutron | 14:23 | |
| mestery_ | sgran: Look here https://wiki.openstack.org/wiki/Icehouse_Release_Schedule | 14:24 |
| *** mestery_ is now known as mestery | 14:24 | |
| *** changbl has quit IRC | 14:25 | |
| sgran | so, basically 2 weeks to icehouse-1 | 14:26 |
| sgran | hopefully reviews start picking up in speed | 14:27 |
| sgran | if it takes 3 or 4 weeks to get a new driver reviewed, there's no way it'll make icehouse | 14:27 |
| *** peristeri has joined #openstack-neutron | 14:29 | |
| mestery | sgran: New driver for what? | 14:31 |
| sgran | I have started a new LBaaS driver | 14:31 |
| mestery | sgran: Cool! Vendor based or open source? | 14:32 |
| sgran | but I'm waiting on https://review.openstack.org/#/c/40381/ | 14:32 |
| *** jianingy is now known as jianingy_afk | 14:32 | |
| mestery | sgran: Only reason I ask is due to the new testing requirements for new drivers (all new drivers need 3rd party tempest tests before being accepted, among other requirements). | 14:32 |
| sgran | it's an appliance, but the vendor has told me (not in writing yet) that I can open source it | 14:32 |
| sgran | I'd like to push it back up | 14:32 |
| mestery | sgran: Most awesome. | 14:32 |
| sgran | sure, I've talked to the vendor about it | 14:32 |
| mestery | cool | 14:33 |
| sgran | 40381 is taking quite a long time to get through, and spending most of it's time waiting for review | 14:33 |
| sgran | so this is making me nervous | 14:33 |
| sgran | I guess we'll just run with a local driver for icehouse, but that seems a bit crap | 14:33 |
| mestery | 40381 is a larger review, those tend to take longer to work their way through. | 14:33 |
| sgran | hope you're right :) | 14:34 |
| *** jecarey has joined #openstack-neutron | 14:35 | |
| *** afazekas_ has joined #openstack-neutron | 14:35 | |
| *** thedodd has joined #openstack-neutron | 14:46 | |
| *** unicell has joined #openstack-neutron | 14:56 | |
| *** richardboswell has quit IRC | 14:57 | |
| *** richardboswell has joined #openstack-neutron | 14:57 | |
| *** salv-orlando has joined #openstack-neutron | 15:02 | |
| *** dkliban has joined #openstack-neutron | 15:04 | |
| *** armax has joined #openstack-neutron | 15:07 | |
| markmcclain | sgran: the new driver will require 3rd party tests | 15:10 |
| markmcclain | have you begun work on those? | 15:10 |
| *** mihgen has quit IRC | 15:12 | |
| anteaya | hello markmcclain | 15:20 |
| *** peristeri has quit IRC | 15:21 | |
| *** peristeri has joined #openstack-neutron | 15:23 | |
| markmcclain | anteaya: hi | 15:24 |
| *** nati_uen_ has joined #openstack-neutron | 15:25 | |
| anteaya | how goes your day so far? | 15:26 |
| sgran | markmcclain: I've started talking to riverbed about it | 15:26 |
| sgran | I have not started writing gate tests or anything | 15:26 |
| markmcclain | anteaya: it's good…yours? | 15:27 |
| *** otherwiseguy has quit IRC | 15:27 | |
| *** otherwiseguy has joined #openstack-neutron | 15:27 | |
| markmcclain | sgran: you shouldn't have to write any extra tests.. just configure a setup to run tempest against it | 15:27 |
| sgran | oh, even easier | 15:27 |
| anteaya | good thanks, beautiful sunrise this morning | 15:28 |
| *** nati_ueno has quit IRC | 15:28 | |
| *** idella4 has quit IRC | 15:29 | |
| anteaya | sgran: don't stay stuck on something, speak up and share what you are doing | 15:29 |
| anteaya | then if someone can help they can jump in | 15:29 |
| sgran | anteaya: nah, it's not that I'm stuck, it's that I'm waiting for the LBaaS agent code shape to shake itself out | 15:30 |
| anteaya | sgran: what or who needs to do something about the LBaaS agent code shape? | 15:31 |
| sgran | there's a refactor going on, when that's done, it needs patching so more than one agent can manage a loadbalancer backend, and the new driver needs to get refactored to match | 15:31 |
| anteaya | who is working on the refactor? | 15:31 |
| sgran | it's mostly just waiting for the same thing I've been waiting for - 40381 | 15:31 |
| sgran | obondarev | 15:31 |
| anteaya | yeah, the waiting is what gets me down | 15:31 |
| sgran | I'm sure he's on it - he has been quick getting to things so far | 15:31 |
| sgran | it's just that it seems like a day or two between every upload/comment cycle, which is starting to take forever | 15:32 |
| anteaya | let's see if he is around, obondarev are you avaiable for a quick LBaaS agent re-write update? | 15:32 |
| anteaya | sgran: yes | 15:32 |
| anteaya | that cycle is much shorter in other projects because they talk to each other in irc | 15:32 |
| anteaya | which is what I am trying to promote here | 15:33 |
| anteaya | so we can begin to close up that turn around time | 15:33 |
| anteaya | I have seen agreement on a new patch take 5 or 10 minutes if everyone is in the discussion | 15:33 |
| anteaya | having to wait weeks and take 15+ patchsets for a new patch is really painful | 15:34 |
| obondarev | anteaya: yep, I'm here | 15:37 |
| obondarev | update on https://review.openstack.org/#/c/40381 - working on Akihiro's comments right now | 15:37 |
| *** carl_baldwin has quit IRC | 15:37 | |
| anteaya | awesome thanks obondarev | 15:38 |
| obondarev | will come up with a new patch set soon | 15:38 |
| anteaya | thanks for letting us know | 15:38 |
| sgran | great, thanks | 15:38 |
| anteaya | great work | 15:38 |
| sgran | sorry to be a pest | 15:38 |
| anteaya | never a pest when your work hinges on someone elses | 15:38 |
| anteaya | good to know rather than wonder | 15:38 |
| *** richardboswell has quit IRC | 15:39 | |
| anteaya | markmcclain: what is your take on patch length? | 15:40 |
| *** richardboswell has joined #openstack-neutron | 15:40 | |
| anteaya | there was a very good ml post by lifeless sometime back on diminishing returns after a certain line length for patches | 15:41 |
| anteaya | I think the number under discussion was 200 lines per patch | 15:41 |
| anteaya | not throwing a wrench in the works on 40381 but it is +1300, -700 net 600 lines of code | 15:42 |
| anteaya | that is a significant amount of mental resources to apply to this patch for each reviewer | 15:42 |
| anteaya | perhaps if there was an approach to line length for patches, encouraging features to be split into smaller patches, the review cycle would go faster | 15:43 |
| anteaya | so a series of dependant patches rather than one large patch | 15:44 |
| *** salv-orlando has quit IRC | 15:50 | |
| *** afazekas_ has quit IRC | 15:50 | |
| *** mordred has quit IRC | 15:52 | |
| *** Sam-I-Am has quit IRC | 15:52 | |
| anteaya | he posted it in August: http://lists.openstack.org/pipermail/openstack-dev/2013-August/013692.html | 15:57 |
| *** carl_baldwin has joined #openstack-neutron | 15:57 | |
| *** salv-orlando has joined #openstack-neutron | 15:58 | |
| *** mordred has joined #openstack-neutron | 16:00 | |
| openstackgerrit | Oleg Bondarev proposed a change to openstack/neutron: LBaaS: unify haproxy-on-host plugin driver and agent https://review.openstack.org/40381 | 16:00 |
| *** devlaps has joined #openstack-neutron | 16:01 | |
| anteaya | sgran ^ | 16:02 |
| sgran | \o/ | 16:03 |
| anteaya | :D | 16:04 |
| markmcclain | anteaya: that number is very dependent on what the patch is doing.. some patches have be larger because they're adding a single functional component and others we can break down into smaller patches | 16:04 |
| anteaya | are large patches ever asked to become a series of smaller patches? | 16:05 |
| anteaya | or if the author submits a large patch, does it stay large? | 16:05 |
| markmcclain | we've rejected patches for being too large | 16:06 |
| anteaya | awesome, thanks | 16:06 |
| anteaya | good to know | 16:06 |
| markmcclain | many times with plugins the first commit is large because we require the core API be implemented | 16:06 |
| anteaya | okay | 16:07 |
| anteaya | trying to find ways to reduce the amount of time a patch takes to get merged | 16:07 |
| anteaya | since that builds momentum | 16:08 |
| anteaya | going to head out for a walk, back in a bit | 16:09 |
| markmcclain | the patch that started this is discussion is larger refactor.. I don't think that Oleg could broken it up much more without losing the context of the direction of hte patch | 16:09 |
| anteaya | oh there you are | 16:09 |
| anteaya | fair enough | 16:09 |
| anteaya | just trying to understand the foundational principles for development in the channel | 16:10 |
| anteaya | and this helps, thanks | 16:10 |
| * anteaya puts on her coat and hat | 16:11 | |
| ygbo | Hi all, can someone help me with VPNaaS? My issue is that When doing https://wiki.openstack.org/wiki/Neutron/VPNaaS/HowToInstall, I get no errors, everything "seems" wonderfull, security groups allowing everything in all directions, the 2 routers can ping each other, but the from the routers respective namespaces (or vms on their lan) I can't ping remote side lans (routers and VM ping successfully each routers external IP). In /opt/stack/data/neutron/ | 16:12 |
| ygbo | ipsec/<router_id>/* apart from configuration and process pids there are no logs at all. A guide on how to debug and find logs would be appreciated because ipsec processes are visible with ps, I get no error in neutron, so without logs nor errors I don't know where to start searching. PS: I know openvpn off by hart, but not ipsec... | 16:12 |
| ygbo | Is OpenVPN not planned for VPNaaS? | 16:12 |
| markmcclain | OpenVPN is something we're working on during Icehouse | 16:12 |
| *** arahal has joined #openstack-neutron | 16:13 | |
| ygbo | markmcclain: cool :-) | 16:13 |
| ygbo | markmcclain: do you have any blueprint or link towards something started on OpenVPN? | 16:13 |
| markmcclain | it's a little early for nati_uen_ to be active, but debugging tips would be good for docs | 16:14 |
| markmcclain | https://blueprints.launchpad.net/neutron/+spec/neutron-ssl-vpn | 16:14 |
| markmcclain | ygbo: also this etherpad: https://etherpad.openstack.org/p/NeutronVPNaaSIceHouse | 16:15 |
| ygbo | markmcclain: thanks for the links | 16:17 |
| *** jp_at_hp has joined #openstack-neutron | 16:20 | |
| ygbo | openvpn comes with easy_rsa scripts to easily build the PKI (CA and certificates), will we re_use these? Is a PKI management planned or will users have to provide their pre-made certificates whith their own PKI management? | 16:21 |
| *** nati_uen_ has quit IRC | 16:24 | |
| markmcclain | ygbo: not sure all of those issues have been solved yet | 16:24 |
| *** yfried has quit IRC | 16:25 | |
| markmcclain | most of the VPN team is US-west, so might have to ping the room later to get more info | 16:25 |
| *** jp_at_hp has quit IRC | 16:38 | |
| *** changbl has joined #openstack-neutron | 16:38 | |
| *** jp_at_hp has joined #openstack-neutron | 16:39 | |
| *** rossella_s_ has joined #openstack-neutron | 16:44 | |
| *** rossella_s has quit IRC | 16:44 | |
| *** rossella_s_ is now known as rossella_s | 16:44 | |
| *** terrylhowe has joined #openstack-neutron | 16:46 | |
| terrylhowe | I'm working on the unified CLI and I'm wondering what the reaction might be to generalizing some things in the neutron cli | 16:46 |
| terrylhowe | For instance, things like using getattr on parsed_args so the unified CLI doesn't have to have command line args for somethings. E.g.: | 16:47 |
| terrylhowe | page_size = getattr(parsed_args, 'page_size', None) | 16:47 |
| markmcclain | I think that would be helpful | 16:47 |
| markmcclain | our current CLI implementation has some optional args that we don't predeclare so this would fit in | 16:48 |
| terrylhowe | I'd sooner reuse as much of that code as possible, but the implementation is a lot different. Some nice extra features for sure, but we'd need to work to get them across all the other clients first | 16:48 |
| *** dhellmann has joined #openstack-neutron | 16:49 | |
| markmcclain | yeah.. we're also working on refactoring some of the underlying code | 16:49 |
| markmcclain | so if there are particular pain points for the unified CLI it would be good to know them | 16:49 |
| terrylhowe | is there a blue print? | 16:50 |
| markmcclain | not an updated one | 16:50 |
| markmcclain | still working on incorporating the feedback on pain and scaling points of the neutron client | 16:50 |
| markmcclain | the first batch of changes will mostly target the API interactions and leave the CLI classes un modified | 16:51 |
| markmcclain | but if there's items to fix on the CLI we can carve up the work differently | 16:51 |
| terrylhowe | should I create a blueprint at least for discussion? | 16:53 |
| markmcclain | yes please do | 16:53 |
| markmcclain | I think a separate bp for CLI changes would be useful | 16:53 |
| terrylhowe | cool | 16:53 |
| *** steven-weston has joined #openstack-neutron | 16:54 | |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Optionally delete namespaces when they are no longer needed https://review.openstack.org/56114 | 16:57 |
| *** chandankumar has quit IRC | 16:59 | |
| *** networkstatic has joined #openstack-neutron | 17:02 | |
| *** changbl has quit IRC | 17:05 | |
| *** amuller has quit IRC | 17:06 | |
| *** afazekas_ has joined #openstack-neutron | 17:08 | |
| terrylhowe | https://blueprints.launchpad.net/python-neutronclient/+spec/unifiedcli | 17:09 |
| *** alexpilotti has quit IRC | 17:10 | |
| *** lskyw has joined #openstack-neutron | 17:10 | |
| *** steven-weston has quit IRC | 17:11 | |
| markmcclain | terrylhowe: thanks for creating it | 17:11 |
| openstackgerrit | Stephen Gran proposed a change to openstack/neutron: Enforce unique constraint on neutron pool members https://review.openstack.org/56815 | 17:13 |
| *** salv-orlando has quit IRC | 17:14 | |
| *** salv-orlando has joined #openstack-neutron | 17:22 | |
| *** changbl has joined #openstack-neutron | 17:26 | |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Optionally delete namespaces when they are no longer needed https://review.openstack.org/56114 | 17:31 |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/neutron: Enable polling minimization https://review.openstack.org/57290 | 17:38 |
| *** salv-orlando has quit IRC | 17:39 | |
| *** salv-orlando has joined #openstack-neutron | 17:47 | |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Refactor _spawn/destroy_metadata_proxy to not require full router_info https://review.openstack.org/57293 | 17:51 |
| *** richardboswell2 has joined #openstack-neutron | 17:54 | |
| *** richardboswell has quit IRC | 17:57 | |
| *** rossella_s has quit IRC | 18:00 | |
| *** insanidade has joined #openstack-neutron | 18:04 | |
| *** jsergent has joined #openstack-neutron | 18:05 | |
| *** ygbo has quit IRC | 18:08 | |
| *** yfried has joined #openstack-neutron | 18:10 | |
| *** afazekas_ has quit IRC | 18:12 | |
| *** steven-weston has joined #openstack-neutron | 18:15 | |
| *** jprovazn is now known as jprovazn_afk | 18:15 | |
| *** suresh12 has joined #openstack-neutron | 18:16 | |
| *** decede has quit IRC | 18:37 | |
| *** suresh12 has quit IRC | 18:38 | |
| *** safchain has quit IRC | 18:38 | |
| *** suresh12 has joined #openstack-neutron | 18:39 | |
| *** decede has joined #openstack-neutron | 18:40 | |
| *** yfried has quit IRC | 18:43 | |
| *** jp_at_hp has quit IRC | 18:45 | |
| anteaya | the -infra meeting is in 13 minutes in -meeting if anyone wants to listen in on the current state of the gate | 18:47 |
| *** mihgen has joined #openstack-neutron | 18:49 | |
| insanidade | anteaya: what meeting is happening there right now ? | 18:49 |
| anteaya | keystone | 18:50 |
| insanidade | anteaya: thanks. | 18:50 |
| anteaya | np | 18:50 |
| insanidade | in terms of routing, what is the difference between a standard and a distributed router ? | 18:55 |
| openstackgerrit | Amir Sadoughi proposed a change to openstack/neutron: Added check on plugin.supported_extension_aliases https://review.openstack.org/47432 | 18:55 |
| mestery | salv-orlando: ping | 18:57 |
| salv-orlando | hi mestery | 18:57 |
| mestery | salv-orlando: In this review you just pushed (https://review.openstack.org/#/c/57290/), I notice that ML2 doesn't support config for minimize_polling. | 18:58 |
| mestery | I think it should, shall I file a bug for that? | 18:58 |
| mestery | The agent either way works with both I guess. | 18:58 |
| mestery | ACtually, looking at it, it's only odd because the agent config is in the openvswitch section of code. | 18:59 |
| salv-orlando | yes, I've sent that too | 18:59 |
| salv-orlando | it's just a matter of adding the defaults into ml2_conf.ini I think | 18:59 |
| salv-orlando | because it's not that the configuration is not supported, it's just that it's implicitly taking the default value for each parameter | 18:59 |
| anteaya | insanidade: I don't know the answer to your question, hopefully someone else does | 19:00 |
| *** sadasu has joined #openstack-neutron | 19:01 | |
| *** jprovazn_afk is now known as jprovazn | 19:02 | |
| *** matrohon_ has quit IRC | 19:02 | |
| *** jlibosva has quit IRC | 19:02 | |
| mestery | salv-orlando: Exactly, thanks for confirming! | 19:02 |
| *** yfried has joined #openstack-neutron | 19:11 | |
| *** rkukura has quit IRC | 19:20 | |
| *** mlavalle has joined #openstack-neutron | 19:21 | |
| openstackgerrit | Dane LeBlanc proposed a change to openstack/neutron: Improve unit test coverage for Cisco plugin nexus code https://review.openstack.org/57100 | 19:25 |
| *** rkukura has joined #openstack-neutron | 19:26 | |
| salv-orlando | otherwiseguy: ping | 19:26 |
| otherwiseguy | salv-orlando: pong | 19:27 |
| salv-orlando | otherwiseguy: a little bird told me you're happy to take over on bug 1224001 ? | 19:27 |
| otherwiseguy | salv-orlando: those little birds sure do get around. :) | 19:28 |
| otherwiseguy | salv-orlando: Love to. :) | 19:28 |
| salv-orlando | are you already up-to-date? otherwise I can fill you in with all the details | 19:28 |
| otherwiseguy | salv-orlando: I think I'm pretty up to date. minimizing polling good to make default, needs polling interval set as a precondition? Anything else I'm missing? | 19:29 |
| salv-orlando | nope. I think this will reduce if not eliminate failures for bug 1224001. Basically what happens is that the VM is not reachable via a floating IP | 19:30 |
| salv-orlando | but the floating IP is configured, it's the VM which does not get an IP | 19:30 |
| salv-orlando | now I've seen the DHCP agent being update 1-2 seconds after the vm boots, but that should not be an issue, because in those seconds the vm is likely still loading the ramdisk | 19:31 |
| salv-orlando | I've also noticed that in some cases it might take up to 15 seconds for the ovs agent to wire the port, and that might cause cirrus, which boots very quickly, to timeout on the dhcp request | 19:32 |
| otherwiseguy | salv-orlando: makes sense. my natural tendency is to cringe when I hear things like "X updates a bit after Y, but Y will most likely not notice" :) | 19:34 |
| salv-orlando | there's a situation where we can hardly enforce a synchronisation point. | 19:34 |
| *** reed has joined #openstack-neutron | 19:34 | |
| otherwiseguy | yeah, the world is a messy place. ;) | 19:35 |
| salv-orlando | it's one of those borderline situations where "eventual consistency" means more "finger crossed" | 19:35 |
| *** alexpilotti has joined #openstack-neutron | 19:37 | |
| marun | otherwiseguy: one last thing, make sure to depend on your 'kill child process' patch! | 19:38 |
| *** marun has quit IRC | 19:39 | |
| otherwiseguy | marun: will do. hopefully it can make it through the dying/overloaded testing. | 19:39 |
| anteaya | any new devstack-gate jobs have to be cleared with clarkb for a while, he is doing a consolidation | 19:48 |
| anteaya | and we are currently producting logs at the rate of just under 1T per week | 19:53 |
| salv-orlando | we openstack or we neutron? | 19:57 |
| salv-orlando | marun, otherwise guy: do we need a devstack change to pass the right minimize_polling options in? | 19:58 |
| anteaya | I would hope you were both | 19:58 |
| anteaya | are you not openstack? | 19:58 |
| * salv-orlando thinks anteaya is trying to read an openstack vs neutron thing in my statement is perhaps | 20:00 | |
| salv-orlando | that's not the case. I just wanted to know if the neutron logs were taking so much space | 20:00 |
| anteaya | help me understand your question then | 20:00 |
| anteaya | ah | 20:00 |
| anteaya | sorry | 20:01 |
| anteaya | all of openstack is creating 1T of logs per week | 20:01 |
| anteaya | currently | 20:01 |
| salv-orlando | I've attended the infra meeting; It seems it's not critical at the moment; we (openstack) have capacity. Is that correct? | 20:02 |
| anteaya | yes, fungi is going to increase volume for static log storage | 20:02 |
| anteaya | just sharing info on our growth rate | 20:02 |
| anteaya | we are growing fast | 20:02 |
| anteaya | there was an expectation log growth would level out but not only did that not happen it has grown faster | 20:03 |
| anteaya | and thanks for attending the meeting | 20:03 |
| *** mlavalle has quit IRC | 20:29 | |
| *** yfried has quit IRC | 20:33 | |
| *** aymenfrikha has joined #openstack-neutron | 20:50 | |
| *** sadasu has quit IRC | 20:52 | |
| openstackgerrit | A change was merged to openstack/neutron: LBaaS: fix reported binary name of a loadbalancer agent https://review.openstack.org/55377 | 20:52 |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Call _destroy_metadata_proxy from _destroy_router_namespaces https://review.openstack.org/57293 | 20:57 |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Call _destroy_metadata_proxy from _destroy_router_namespaces https://review.openstack.org/57293 | 20:58 |
| *** aymenfrikha has quit IRC | 20:59 | |
| *** yamahata_ has quit IRC | 21:12 | |
| openstackgerrit | Carl Baldwin proposed a change to openstack/neutron: Optionally delete namespaces when they are no longer needed https://review.openstack.org/56114 | 21:13 |
| *** jprovazn has quit IRC | 21:19 | |
| *** lskyw has quit IRC | 21:21 | |
| *** yamahata_ has joined #openstack-neutron | 21:34 | |
| *** armax has left #openstack-neutron | 21:39 | |
| anteaya | markmcclain: are you ready to use rootwrap by i-2? | 21:45 |
| markmcclain | yeah.. I really don't have any objection to ttx's plan | 21:48 |
| markmcclain | we're running the version from oslo so I expect the migration to relatively painless | 21:48 |
| *** nati_ueno has joined #openstack-neutron | 21:53 | |
| *** same5336_ has quit IRC | 21:54 | |
| *** bvandenh has quit IRC | 21:55 | |
| anteaya | markmcclain: awesome | 21:57 |
| *** insanidade has quit IRC | 21:59 | |
| markmcclain | most of the hard word was done in Havana harmonizing our old fork with oslo | 21:59 |
| *** peristeri has quit IRC | 22:00 | |
| *** idella4 has joined #openstack-neutron | 22:01 | |
| *** richardboswell2 has quit IRC | 22:03 | |
| anteaya | wonderful | 22:05 |
| anteaya | in other news, in the current graph of the gate illness, neutron is not playing the most important role | 22:06 |
| anteaya | as a matter of fact, the gate neutron tests are passing | 22:06 |
| anteaya | to give you a moment of levity | 22:06 |
| *** dkliban has quit IRC | 22:08 | |
| markmcclain | nice to hear, but we still have substantial work to complete quickly | 22:08 |
| markmcclain | the race conditions have gone in waves | 22:08 |
| anteaya | yes | 22:08 |
| anteaya | just wanted to share the moment | 22:09 |
| anteaya | fleeting though it maybe | 22:09 |
| anteaya | may be | 22:09 |
| anteaya | markmcclain: is there anything in addition to what I am doing that can help with that | 22:09 |
| anteaya | do you think? | 22:09 |
| idella4 | g'day all | 22:10 |
| anteaya | hey idella4 | 22:10 |
| *** nati_ueno has quit IRC | 22:10 | |
| idella4 | it's mornin' | 22:10 |
| idella4 | ho markmcclain | 22:10 |
| anteaya | it's really morning | 22:10 |
| anteaya | what time for you 6am? | 22:10 |
| idella4 | yes | 22:10 |
| idella4 | 6.10 | 22:11 |
| markmcclain | anteaya: not much new right now | 22:11 |
| idella4 | sun's been up a good while | 22:11 |
| markmcclain | idella4: hi! | 22:11 |
| anteaya | markmcclain: okay don't keep it a secret if you think of something | 22:11 |
| idella4 | have you passed some woeds onto your counterpart at this point markmcclain | 22:11 |
| *** pcm_ has quit IRC | 22:12 | |
| *** yamahata_ has quit IRC | 22:12 | |
| idella4 | words | 22:12 |
| idella4 | anteaya here briefed me on some fundamentals | 22:12 |
| markmcclain | anteaya: too much to do to stockpile work items | 22:14 |
| idella4 | I'm re-running this neutron testsuite seeking a consistent fail rate of 1 | 22:14 |
| markmcclain | idella4: which test? | 22:14 |
| anteaya | markmcclain: indeed | 22:15 |
| idella4 | which I got around 28 hours ago | 22:15 |
| idella4 | testsuite we were discussing, neutron-2013.2 | 22:15 |
| idella4 | It's difficult wrking it with the essential logging or coverage reporting turned off | 22:16 |
| openstackgerrit | A change was merged to openstack/neutron: Replace mox in unit tests with mock https://review.openstack.org/54044 | 22:17 |
| *** banix has joined #openstack-neutron | 22:17 | |
| markmcclain | hmmm.. testr produces the failures for me | 22:18 |
| idella4 | oh it DOES? | 22:18 |
| markmcclain | yes | 22:18 |
| idella4 | that's significant | 22:18 |
| markmcclain | sorry when I make a failing tests | 22:18 |
| idella4 | well we're on the same pages or two | 22:18 |
| markmcclain | I get a traceback | 22:19 |
| idella4 | oh well, tell me when you're perfect | 22:19 |
| *** ljjjusti1 has joined #openstack-neutron | 22:19 | |
| *** nati_ueno has joined #openstack-neutron | 22:20 | |
| idella4 | hmm; well I got | 22:21 |
| idella4 | Binary content: traceback (test/plain; charset="utf8") Ran 12233 tests in 2123.316s FAILED (id=0, failures=2, skips=171) | 22:21 |
| idella4 | and that's it. It's a traceback I suppose | 22:22 |
| *** ljjjustin has quit IRC | 22:23 | |
| idella4 | well, anyway, is there much diff between the tests in the static tarball and the live repo? | 22:24 |
| idella4 | the havana live repo? | 22:24 |
| markmcclain | there shouldn't be any difference | 22:36 |
| idella4 | right, thought so | 22:36 |
| *** networkstatic has quit IRC | 22:37 | |
| *** nati_ueno has quit IRC | 22:37 | |
| idella4 | that's quite inconsequential at this point, I think the totals varied only by about 5 | 22:37 |
| *** arahal has quit IRC | 22:38 | |
| *** nati_ueno has joined #openstack-neutron | 22:38 | |
| idella4 | anyway, was there a fix to the ncira hardware stuff in any history to find? | 22:38 |
| idella4 | or does this represent new info? | 22:38 |
| markmcclain | seems like new info | 22:39 |
| idella4 | ah ok | 22:39 |
| markmcclain | need to fine a good way to replicate in master | 22:39 |
| markmcclain | s/fine/find/ | 22:39 |
| idella4 | right | 22:39 |
| markmcclain | so that we can file the proper bug | 22:39 |
| idella4 | so I'm being a bit helpful in finding something that needs a fix or inprove | 22:39 |
| idella4 | improve | 22:40 |
| markmcclain | yes, this is helpful | 22:40 |
| idella4 | ok. well my last best result was a fail of 1 | 22:40 |
| idella4 | bit with no backtrace of any type, I'm blinded | 22:41 |
| markmcclain | strange you're not getting any output | 22:41 |
| markmcclain | testr run --parallel --concurrency=3 | 22:41 |
| markmcclain | will produce output for a failing test for me | 22:42 |
| idella4 | ah ok | 22:42 |
| idella4 | well that's helpful markmcclain I can add the --concurrency=3. after I let the current complete | 22:42 |
| idella4 | as I said I'm basically a testr newbie | 22:43 |
| idella4 | more practiced with nose and nosetests, so I'm learnin' more too | 22:43 |
| *** jecarey has quit IRC | 22:43 | |
| *** same5336_ has joined #openstack-neutron | 22:44 | |
| *** fkak has quit IRC | 22:45 | |
| markmcclain | the concurrency flag limits the number of parallel processes that testr will use to run | 22:45 |
| markmcclain | very similar to the -j flag in make | 22:46 |
| *** alagalah has quit IRC | 22:46 | |
| idella4 | ok thx | 22:46 |
| idella4 | so I ocan re-add the --parallel option, ok that's 'good' news | 22:47 |
| idella4 | well after taking out the ncira folder, I still had one left and I'm still trying to figure what it is | 22:48 |
| markmcclain | ok | 22:50 |
| *** alagalah has joined #openstack-neutron | 22:55 | |
| *** jecarey has joined #openstack-neutron | 22:56 | |
| *** mihgen has quit IRC | 22:57 | |
| *** thedodd has quit IRC | 23:03 | |
| *** julim has quit IRC | 23:03 | |
| *** jpich has quit IRC | 23:04 | |
| *** salv-orlando has quit IRC | 23:07 | |
| *** salv-orlando has joined #openstack-neutron | 23:07 | |
| idella4 | hmm at this point (still running) I have | 23:09 |
| idella4 | FAIL: neutron.tests.unit.openvswitch.test_agent_scheduler.OvsAgentSchedulerTestCaseXML.test_router_schedule_with_candidates | 23:09 |
| idella4 | does this pass for you? | 23:09 |
| *** nati_ueno has quit IRC | 23:17 | |
| markmcclain | yes | 23:37 |
| openstackgerrit | stephen-ma proposed a change to openstack/neutron: Delete DHCP port without DHCP server on a network node https://review.openstack.org/56740 | 23:38 |
| idella4 | oh ok thx | 23:38 |
| anteaya | okay big converstation happened in infra about the gate | 23:39 |
| anteaya | items being considered include removing or reducing use of recheck/reverify, having tests run more than once for the same patch | 23:40 |
| anteaya | and having gate critical bugs be identified and announced project wide | 23:40 |
| anteaya | join the conversation if you would like: https://etherpad.openstack.org/p/future-of-recheck | 23:41 |
| idella4 | what is the 'gate' exactly? | 23:46 |
| *** changbl has quit IRC | 23:48 | |
| *** banix has left #openstack-neutron | 23:49 | |
| salv-orlando | idella4: it's the last door you have to go through before your patch can be merged. | 23:49 |
| salv-orlando | the whole workflow is here: https://wiki.openstack.org/wiki/Gerrit_Workflow | 23:49 |
| salv-orlando | idella4: this at least from a developer perspective | 23:49 |
| idella4 | ok thx | 23:52 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!