Thursday, 2016-03-17

openstackgerritSteve McLellan proposed openstack/searchlight: Add neutron subnets and routers  https://review.openstack.org/29377100:59
*** bpokorny has quit IRC01:04
*** lakshmiS has joined #openstack-searchlight01:52
*** lei-zh has joined #openstack-searchlight02:07
*** tsufiev has quit IRC02:19
*** tsufiev has joined #openstack-searchlight02:25
openstackgerritLakshmi N Sampath proposed openstack/searchlight: Swift plugin  https://review.openstack.org/27162202:43
TravTlakshmiS: can you take a look at the subnets and routers patch?02:43
lakshmiSok02:44
*** GB21 has joined #openstack-searchlight03:09
*** GB21 has quit IRC03:15
*** yingjun has joined #openstack-searchlight03:55
*** yingjun has quit IRC03:57
openstackgerritMerged openstack/searchlight: Swift plugin  https://review.openstack.org/27162205:06
*** lakshmiS has quit IRC05:13
*** lakshmiS has joined #openstack-searchlight05:18
*** GB21 has joined #openstack-searchlight05:22
*** GB21 has quit IRC05:31
*** GB21 has joined #openstack-searchlight05:33
*** GB21 has quit IRC05:51
*** GB21 has joined #openstack-searchlight05:51
*** GB21 has quit IRC05:57
*** GB21 has joined #openstack-searchlight05:59
*** lakshmiS has quit IRC06:01
*** itisha has quit IRC06:19
*** GB21 has quit IRC06:19
*** GB21 has joined #openstack-searchlight06:21
*** GB21 has quit IRC06:49
*** GB21 has joined #openstack-searchlight06:49
*** GB21 has quit IRC07:13
*** GB21 has joined #openstack-searchlight07:13
*** GB21 has quit IRC07:24
*** GB21 has joined #openstack-searchlight08:00
*** GB21 has quit IRC08:15
*** GB21 has joined #openstack-searchlight08:15
*** pcaruana has joined #openstack-searchlight08:24
*** GB21 has quit IRC08:28
*** GB21 has joined #openstack-searchlight08:30
*** GB21 has quit IRC09:04
*** GB21 has joined #openstack-searchlight09:28
*** preetika has joined #openstack-searchlight09:36
*** GB21 has quit IRC09:44
*** lei-zh has quit IRC09:57
*** GB21 has joined #openstack-searchlight10:11
*** GB21 has quit IRC10:52
*** GB21 has joined #openstack-searchlight10:52
*** sjmc7 has quit IRC10:56
*** sjmc7 has joined #openstack-searchlight10:57
*** GB21 has quit IRC10:58
*** GB21 has joined #openstack-searchlight10:58
*** GB21 has quit IRC11:07
*** GB21 has joined #openstack-searchlight11:08
*** preetika has quit IRC11:37
*** GB21 has quit IRC11:47
*** openstackgerrit has quit IRC11:48
*** GB21 has joined #openstack-searchlight11:48
*** openstackgerrit has joined #openstack-searchlight11:48
*** GB21 has quit IRC11:53
*** GB21 has joined #openstack-searchlight11:55
*** GB21 has quit IRC12:01
*** GB21 has joined #openstack-searchlight12:03
*** GB21 has quit IRC12:24
*** sigmavirus24_awa is now known as sigmavirus2413:47
*** lakshmiS_ has joined #openstack-searchlight14:25
*** GB21 has joined #openstack-searchlight14:25
*** pcaruana has quit IRC14:28
*** sigmavirus24 is now known as sigmavirus24_awa14:40
*** sigmavirus24_awa is now known as sigmavirus2414:41
*** lei-zh has joined #openstack-searchlight14:53
TravTCourtesy Searchlight meeting reminder in one minute in #openstack-meeting-4: lakshmiS, nikhil_k, rosmaita, TravT, david-lyle, sjmc7, abhijeetm, itisha, GB21, briancline, lei-zh, yingjun, RickA-HP14:59
*** yingjun has joined #openstack-searchlight15:00
*** GB21 has quit IRC15:15
*** bpokorny has joined #openstack-searchlight15:50
*** lakshmiS_ has left #openstack-searchlight15:54
*** yingjun has quit IRC15:56
*** lei-zh has left #openstack-searchlight15:56
*** yingjun has joined #openstack-searchlight15:56
*** khushbu has joined #openstack-searchlight15:56
*** yingjun has quit IRC16:01
*** khushbu has quit IRC16:02
*** khushbu has joined #openstack-searchlight16:14
*** pcaruana has joined #openstack-searchlight16:22
*** lakshmiS has joined #openstack-searchlight16:24
khushbuhi nikhil16:28
khushbuI am writing a heat stack plugin16:28
khushbufor a bug16:28
openstackgerritTravis Tripp proposed openstack/searchlight: Simple Script for Generating Resources  https://review.openstack.org/23069716:29
*** itisha has joined #openstack-searchlight16:32
*** khushbu has quit IRC16:33
*** preetika has joined #openstack-searchlight16:36
*** sigmavirus24 is now known as sigmavirus24_awa16:45
*** sigmavirus24_awa is now known as sigmavirus2416:45
openstackgerritRick Aulino proposed openstack/searchlight: Fix duplicate entries displayed w/"index sync".  https://review.openstack.org/29416816:55
openstackgerritRick Aulino proposed openstack/searchlight: Fix duplicate entries displayed w/"index sync".  https://review.openstack.org/29416817:01
david-lyleTravT: are you sure you don't want independent for release model on searchlight-ui?17:21
david-lylemost other plugins follow that17:21
david-lyleaside from trove and sahara17:23
david-lylebut those are more mature17:24
david-lylejust a thought17:24
TravTdavid-lyle: that might be a good idea.17:38
TravTi want to be able to state X release is compatible with X release of openstack17:39
sjmc7lakshmiS: about neutron interface add… i don’t see the router change when i add a port17:39
TravTand i think it may quickly get out of sync.17:40
lakshmiSok thats good then it jsut the port create17:40
sjmc7i’m not sure what ‘routes’ is17:40
sjmc7after a week of looking at neutron i’m still partially baffled :)17:40
TravTi guess the first time it depends on a horizon change in the next release, we'd branch stable17:40
sjmc7the subnet doesn’t seem to change either17:40
lakshmiSso what field in router has the information about its interfaces?17:41
sjmc7none17:41
sjmc7there’s a separate API command for it17:41
sjmc7it seems to be port-based17:41
sjmc7the DHCP ports are also a bit of a problem; not sure where they come from17:42
sjmc7i guess from subnet creation17:42
sjmc7will check that too. but neutron seems to get everything off the port info17:42
sjmc7‘device_owner’ gets set differently for these magic ports17:42
lakshmiSyeah the router interface could be either subnet or port17:45
sjmc7right, but the subnet is just a shortcut i think17:45
sjmc7it sets up a port for you17:45
sjmc7versus using an existing one17:46
sjmc7we’ll have to test it a bit more next week to see if there are any crucial edge cases17:46
sjmc7i’m about 95% sure i’ve got it right17:46
lakshmiSwould be good if we can get a neutron core or sme to review it too17:46
sjmc7rick might be friendly with some of our neutron folks in FtC17:48
sjmc7i’ll add some documentation too noting any odd bits17:49
sjmc7good catch on the ha flag too17:57
sjmc7weird that that one’s not shown17:57
*** sigmavirus24 is now known as sigmavirus24_awa18:00
lakshmiSi am sure the admin fields might be hidden in some neutron code like swift did. that might be a better souce to find out than staring at cmd line output :)18:08
sjmc7i diffed the API output :)18:09
sjmc7yeah, i’m sure it is somewhere18:09
TravTwell too devstack failures in a row...18:10
TravTtwo18:10
TravTgeez18:10
sjmc7:(18:10
TravT2016-03-17 18:02:13.963 | ovs-ofctl: br-tun is not a bridge or a socket18:10
TravT2016-03-17 18:02:13.983 | ovs-ofctl: br-tun is not a bridge or a socket18:10
lakshmiSyou are being lucky18:10
sjmc7you should fix your brtun18:11
TravTget your own brtun18:11
sjmc7i’ve got plenty!18:11
TravTwell, i'm gonna go back to old stack and just do an offline update with latest from horizon and searchlight.18:12
TravTi have an update to the local.conf example i think i'll push shortly18:12
*** nehap has joined #openstack-searchlight18:12
nehapHi sjmc718:15
sjmc7hello18:16
nehapsjmc7: Can you please review this https://review.openstack.org/#/c/288853/ and let me know if I am going in the right direction18:17
*** pcaruana has quit IRC18:18
sjmc7yeah, it’s the right direction. we won’t merge this for the mitaka release, but i’ll try to review it once we’re done with the remaining mitaka work18:18
nehapsjmc7: Thanks. For my outreachy application, should I work on some other bug as well18:20
nehapsjmc7: Is there any other easy bug I should work on18:22
TravTnehap, let me take a look18:22
sjmc7just looking at the bug list18:22
sjmc7one sec18:22
sjmc7https://bugs.launchpad.net/searchlight/+bug/1558606 shouldn’t be too hard18:23
openstackLaunchpad bug 1558606 in OpenStack Search (Searchlight) "Remove duplicate project/tenant facets from cinder plugins" [Low,New]18:23
sjmc7https://bugs.launchpad.net/searchlight/+bug/155049418:23
openstackLaunchpad bug 1550494 in OpenStack Search (Searchlight) "Generated version numbers are returned as strings" [Low,New] - Assigned to Ruchi Parmar (ruchiparmar143)18:23
sjmc7you may also discover your own bugs :)18:25
nehapsjmc7: I had looked at the bug https://bugs.launchpad.net/searchlight/+bug/1550494 before18:26
openstackLaunchpad bug 1550494 in OpenStack Search (Searchlight) "Generated version numbers are returned as strings" [Low,New] - Assigned to Ruchi Parmar (ruchiparmar143)18:26
nehapsjmc7: I just needed one clarification: truncate_timestamp should be returned as string. Am I  right18:27
*** sigmavirus24_awa is now known as sigmavirus2418:28
sjmc7umm.. maybe? truncate_timestamp is a string. the function should return a number rather than a string18:28
*** pcaruana has joined #openstack-searchlight18:30
nehapsjmc7: ok, thanks a lot18:33
sjmc7sure. sorry we’re so busy. things should be a bit less crazy next week18:34
sjmc7release weeks are always a scramble18:35
openstackgerritSteve McLellan proposed openstack/searchlight: Add neutron subnets and routers  https://review.openstack.org/29377118:40
sjmc7lakshmiS: put a new patch up. i need to eat something, and i want to take a look at dhcp port events. i’ll put a -1 workflow on for now18:41
lakshmiSok sounds good18:42
lakshmiSwill check it out later18:43
sjmc7everything should be working though18:46
sjmc7just tested interface create/delete18:46
*** nehap has quit IRC18:48
*** nehap has joined #openstack-searchlight18:52
*** nehap has quit IRC18:56
*** yingjun has joined #openstack-searchlight18:56
*** nehap has joined #openstack-searchlight18:58
*** yingjun has quit IRC19:01
openstackgerritRick Aulino proposed openstack/searchlight: Zero Downtime Re-indexing Error Handling.  https://review.openstack.org/29307919:03
*** pcaruana has quit IRC20:17
openstackgerritSteve McLellan proposed openstack/searchlight: Add neutron subnets and routers  https://review.openstack.org/29377120:23
sjmc7^ is good for reviews. i’m gonna figure out DHCP ports as a separate patch; it’s kind of a bug against networks & ports20:28
*** pcaruana has joined #openstack-searchlight20:30
*** matt-borland has joined #openstack-searchlight20:32
matt-borlando/20:32
sjmc7TravT, matt-borland : got a neutron question20:32
matt-borlandk20:32
TravThey20:33
sjmc7there are some neutron ports created to implement magical things like router interfaces and DHCP20:33
sjmc7i don’t *think* horizon shows these20:33
sjmc7and i’m wondering whether instead of cobbling hacks together to implement them, we should ignore them20:33
TravThmm20:33
sjmc7oo, wait, i’m wrong, it does show them20:34
sjmc7grrr20:34
TravTi think you see them in the network details page20:34
matt-borlandI seem to recall a few magical ports showing up in Horizo...20:34
matt-borlandyeah.20:34
sjmc7that’s unfortunate. they seem like useless information20:34
sjmc7oh well. i guess i’ll actually have to do some work for once20:34
matt-borlandfrom my perspective, we're here to write what is useful20:34
sjmc7sorry to interrupt you, i should’ve checked more carefully20:34
TravTyeah, what's wrong with you sjmc7... get back in your dungeon20:35
sjmc7yes master20:35
matt-borlandnow the entirety of Young Frankenstein is playing in my head20:35
sjmc7yeah, maybe the UI would choose to hide them but i think we have to provide the option20:36
sjmc7haha, now THAT is torture20:36
matt-borlandWHAT?20:36
sjmc7TravT: the router patch includes magic router ports but i’m gonna do DHCP separately because it’s hard20:36
* matt-borland loves that movie20:36
*** matt-borland has left #openstack-searchlight20:36
TravTthat was interesting cameo from matt20:37
TravTsjmc7: that's fine.20:37
sjmc7i summoned him20:39
sjmc7beetlejuice!20:39
openstackgerritNeha proposed openstack/searchlight: Remove duplicate project facets from cinder plugins Closes-Bug: #1558606  https://review.openstack.org/29428521:05
openstackbug 1558606 in OpenStack Search (Searchlight) "Remove duplicate project/tenant facets from cinder plugins" [Low,New] https://launchpad.net/bugs/1558606 - Assigned to Neha (neha-m-pande)21:05
TravT@lakshmiS @sjmc721:17
sjmc7lakshmiS: good catch on the neutron policy stuff21:17
TravTjust read that as well.21:17
TravTi think that adds additional sharing21:17
sjmc7i’m not sure about trying to implement that at this stage21:18
sjmc7i think i have missed some public/private rules, which i’ll look at21:18
sjmc7i think though for newton we need to standardize how we’re doing additional rbac21:19
TravTthe extra sharing can be put in as a bug.21:19
lakshmiSlooks like a pattern on shared objects across services21:19
TravTsecurity is always high and we have until mar 28.21:19
lakshmiSatleast neutron is less fine grained than swift21:19
sjmc7yeah, they’re all basically the same tenant:user/tenant:*21:20
sjmc7yeah21:20
sjmc7let’s make that a target for next release. i think public/shared is just based on the network state?21:20
TravTi think base feature that is more restrictive can go in.21:20
sjmc7TravT: can you file a bug for ports to alter the rbac to include ports on public networks?21:21
sjmc7if indeed that is correct21:21
TravTyep21:21
sjmc7which i’m not 100% sure about21:21
sjmc7i’ll check for subnets too21:21
TravThmm, really21:21
sjmc7it sounds sensible, i think?21:21
lakshmiSi think ports are not shared21:21
sjmc7hmmm21:22
sjmc7but subnets are?21:22
lakshmiSyes21:22
sjmc7how magical21:22
lakshmiSsharing ports would be a secuirty breach21:22
TravTlakshmiS you spent a bit of time in your past on openstack networking and are digging in more now21:23
sjmc7hahaha21:23
lakshmiSwho knows how it comes useful21:23
sjmc7:D21:23
lakshmiSbut neutron is a beast21:23
sjmc7yeah, i’m learning all kidns of stuff21:23
TravTso i think i'd like to defer to your judgment on how to handle this21:24
sjmc7can i make a private subnet on a shared network?21:24
lakshmiShaven't tried that21:24
TravTmy perspective is that if we support base concepts that is okay so long as the rbac policies can't make them more open21:24
lakshmiSagree its just more restrictive21:25
sjmc7ok, you’re right - shared subnets are visible, ports are not21:25
sjmc7how wondrous!21:25
sjmc7ok, that’s an easy fix, good catch. yeah, it looks like those policies can only add access, not take it away21:26
lakshmiSright21:26
lakshmiSseparate bug would be ok for it21:26
sjmc7give me 10 minutes to fix the subnet sharing21:28
lakshmiShave a hard stop. will check it out in evening21:28
TravTif everybody actually could use the same policy concepts as keystone, it seems all this sharing could just be implemented as a common database schema and library21:29
sjmc7yeah, i’ve got to go in about 45 minutes21:29
TravTyeah, that's fine.21:29
sjmc7yeah.. like the quota mailing list discussion21:29
sjmc7it’s mad everyone does everything themeslves21:29
TravTi'm not feeling particularly well today...21:30
TravTprobably going to drop off soon as well21:30
lakshmiSwe should introduce common rbac client since we are doing it for everyone anyway :)21:30
*** nehap has quit IRC21:30
TravTthere already is one...21:30
lakshmiS?21:30
TravTpolicy enforcer21:30
TravTbut it has to be done on post result filtering21:30
TravTwhich just don't work for performance21:31
TravTthe keystone policy language is pretty powerful.21:31
lakshmiSoh yeah, that needs more logic which is a little difficult with some projects like glance but should be converged21:31
TravTanyway21:31
TravTgreat work guys21:32
TravTlakshmi, definitely please give another round of reviews later.21:32
lakshmiSsure21:32
TravTonce sjmc7 posts, i'll also take a look21:32
TravTi have some ui bugs i'm trying to figure out as well.21:33
*** lakshmiS has quit IRC21:34
sjmc7ok. written the test (12 minutes), will upload after i write the code (2 minutes) :)21:47
sjmc7gotta love TDD!21:47
openstackgerritSteve McLellan proposed openstack/searchlight: Add neutron subnets and routers  https://review.openstack.org/29377122:00
sjmc7TravT: subnet patch ^22:01
TravTokay cool. will pull now22:01
sjmc7sounds like ports doesn’t need altering?22:01
sjmc7i.e. a port on a shared network isn’t shared22:01
TravTthat's what lakshmi said22:01
TravTbut i'd rather confirms22:01
TravThe22:01
TravTwow22:01
sjmc7ok. i need to go shortly, but will just visually compare horizon to SL subnet results22:02
TravTok22:02
sjmc7if it knows what’s good for it, it’ll look right22:03
sjmc7otherwise… why i oughtta...22:03
sjmc7i get the right number of networks, so let’s call that a win22:04
sjmc7is it weird that after over 15 years in this industry i’m still surprised when stuff i wrote works? :)22:04
TravTno22:05
sjmc7good22:06
sjmc7subnets look right22:06
sjmc7yay22:06
TravThttps://s-media-cache-ak0.pinimg.com/736x/7a/2a/2d/7a2a2d80f6155fa309fdebc739908599.jpg22:07
sjmc7ok, so the only remaining thing is ports for DHCP which i think will have to wait til tomorrow22:07
sjmc7in terms of tagging RC1, i think we could consider doing it once subnets/routers is merged22:08
sjmc7unless you would rather clear the last couple of criticals22:08
TravTjust looking at the list22:09
TravTi'm not seeing anything particular worth holding the tag for after this22:10
TravTalthough we have to go through stable branch fun after that22:10
TravTmaster -> cherry pick22:10
sjmc7the duplicate entries thing we can probably get merged first thing tomorrow22:11
sjmc7dhcp ports i think we can do as a cherrypick22:11
TravTduplicate entries is down to nitpicks22:11
sjmc7yeah, it works fine22:11
sjmc7now we’ve got to find whatever other bugs we created this week :)22:11
sjmc7ok, i should go. i’ll keep an eye out for feedback on that patch and try to get any changes done tonight22:12
TravTok22:12
TravTput an ice pack on your brain22:13
TravTit probably needs it22:13
TravTits been going on overdrive22:13
sjmc7i feel like 25% of it’s dedicated to being a python compiler22:14
*** bpokorny_ has joined #openstack-searchlight22:31
*** bpokorny has quit IRC22:34
*** sigmavirus24 is now known as sigmavirus24_awa22:40
*** TravT has quit IRC23:11
*** TravT has joined #openstack-searchlight23:12
*** bpokorny_ has quit IRC23:31
*** bpokorny has joined #openstack-searchlight23:32
*** bpokorny has quit IRC23:32
*** bpokorny has joined #openstack-searchlight23:33
openstackgerritTravis Tripp proposed openstack/searchlight: Update Docs for Mitaka Release  https://review.openstack.org/29433323:59

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