Monday, 2017-02-06

*** FL1SK has quit IRC00:10
*** FL1SK has joined #openstack-manila00:12
*** gouthamr has quit IRC00:12
*** mtanino has joined #openstack-manila00:40
*** gouthamr has joined #openstack-manila00:47
*** gouthamr has quit IRC00:55
*** mtanino has quit IRC01:08
openstackgerritMerged openstack/manila master: Replaces yaml.load() with yaml.safe_load()  https://review.openstack.org/42915601:35
*** gcb has joined #openstack-manila01:45
*** xinyanzhang has quit IRC01:45
*** gcb has quit IRC01:49
*** gcb has joined #openstack-manila01:50
*** ianychoi has quit IRC01:52
*** zhugaoxiao has joined #openstack-manila01:53
*** ianychoi has joined #openstack-manila01:54
*** mtanino has joined #openstack-manila02:02
*** xinyanzhang has joined #openstack-manila02:04
*** xinyanzhang has quit IRC02:10
*** xinyanzhang has joined #openstack-manila02:11
*** tuanluong has joined #openstack-manila02:12
openstackgerritTommyLike proposed openstack/manila-specs master: Add spec for share backup  https://review.openstack.org/33030602:43
*** tommylikehu has joined #openstack-manila03:27
*** ianychoi has quit IRC03:37
*** ianychoi has joined #openstack-manila03:39
*** mriedem has quit IRC03:46
*** ianychoi has quit IRC04:12
*** ianychoi has joined #openstack-manila04:14
openstackgerritMerged openstack/manila-ui master: Imported Translations from Zanata  https://review.openstack.org/42919404:34
*** david-lyle has joined #openstack-manila04:46
*** david-lyle_ has joined #openstack-manila04:50
*** david-lyle has quit IRC04:50
*** david-lyle_ is now known as david-lyle04:52
*** david-lyle has quit IRC05:21
*** shausy has joined #openstack-manila05:25
tommylikehuping vponomaryov05:51
openstackgerritzhongshengping proposed openstack/puppet-manila master: Add rpc_response_timeout option  https://review.openstack.org/42951406:07
*** lpetrut has joined #openstack-manila06:14
rhagartytbarron, bswartz: FYI - my nova_net_id issue when a creating share network went away after I pulled latest manila_ui. Thanks06:20
*** rraja has joined #openstack-manila06:42
*** mtanino has quit IRC06:46
*** nkrinner_afk is now known as nkrinner07:07
*** gcb has quit IRC07:13
*** gcb has joined #openstack-manila07:14
*** lpetrut has quit IRC07:17
*** jprovazn has joined #openstack-manila07:24
*** gcb has quit IRC07:52
*** gcb has joined #openstack-manila07:54
*** pcaruana has joined #openstack-manila07:55
*** nkrinner has quit IRC08:00
*** nkrinner has joined #openstack-manila08:00
*** daidv has joined #openstack-manila08:06
*** pcaruana has quit IRC08:07
*** pcaruana has joined #openstack-manila08:07
*** daidv has quit IRC08:07
*** databus23_ has joined #openstack-manila08:39
*** lpetrut has joined #openstack-manila08:42
*** zengyingzhe has joined #openstack-manila09:11
*** yumiriam has joined #openstack-manila09:48
*** ganso has joined #openstack-manila09:48
openstackgerritJan Provaznik proposed openstack/manila master: User Messages  https://review.openstack.org/31354909:50
*** gcb has quit IRC09:54
*** yumiriam1 has joined #openstack-manila09:59
openstackgerritTom Barron proposed openstack/manila master: Fix devstack manila nfs install for fedora  https://review.openstack.org/41223210:00
openstackgerritJan Provaznik proposed openstack/python-manilaclient master: Add commands for user messages  https://review.openstack.org/42961410:00
*** yumiriam1 has quit IRC10:02
*** yumiriam has quit IRC10:02
*** yumiriam has joined #openstack-manila10:03
*** tuanluong has quit IRC10:26
*** mkoderer has quit IRC10:36
openstackgerritJan Provaznik proposed openstack/python-manilaclient master: Add commands for user messages  https://review.openstack.org/42961410:40
tommylikehuping tbarron10:46
tbarrontommylikehu: hi10:46
tommylikehutbarron: you just had your new year gala show in USA yesterday :)10:51
tbarrontommylikehu: yes, interesting timing I guess :)10:52
tommylikehutbarron: why?10:52
tbarrontommylikehu: I don't think I got a chance to wish you happy new year, but I do hope you have a great one!10:53
tbarrontommylikehu: I just mean that we do superbowl at lunar new year10:53
tommylikehutbarron: yes it's your lunar new year :)10:53
tbarrontommylikehu: maybe there's a universal need for a lunar new year so we end up having an unofficial holiday then10:54
tommylikehutbarron:that is true, but you don't have holidays for that now, right?10:55
tbarrontommylikehu: no we don't10:55
openstackgerritRodrigo Barbieri proposed openstack/manila master: [DNM] Testing cfgs for 3rd Party CI Scenario tests  https://review.openstack.org/42964410:56
tbarrontommylikehu: I haven't read your revised backup spec yet but will try to take a look soon.10:56
tommylikehutbarron:thanks so much, and luckily I think I  have a chance  to wish you happy new year(wish it's not too late :) )10:58
tbarrontommylikehu: never too late, thanks.10:58
tommylikehutbarron, do you happen to know which of us is working on implement manila command in OSC?11:05
tbarrontommylikehu: Mauricio Lima has been working on a spec for this: https://review.openstack.org/#/c/395775/11:08
tommylikehutbarron: thanks~11:08
tbarrontommylikehu: I think there are two main challenges for making progress on manila with OSC.11:11
tbarrontommylikehu: #1) microversions - OSC currently doesn't handle them.11:11
tbarron#2) desire to be able to run manila standalone, w/o the rest of OpenStack: means there needs to be some kind of plugin capability in OSC so that OSC and a manila-only shell could share common core.11:12
*** dsariel has quit IRC11:17
tommylikehutbarron: so we probably not going to support OSC until these two main concerns are fixed/supported, at lease we will support this in low priority?11:25
tbarrontommylikehu: I think that is the prevailing opinion among manila developers.11:27
tbarrontommylikehu: that said, I know we have customers who see it as a sign that manila is less mature & established than projects which are integrated into OSC.11:28
tommylikehutbarron: that could be kind of prejudice.11:36
openstackgerritTom Barron proposed openstack/manila master: DNM: Update our tempest pin  https://review.openstack.org/42823011:41
tbarrontommylikehu: well, I can see operators point of view who want to be able to use a single client with consistent syntax to be able to manage their whole openstack cloud.11:43
openstackgerritTom Barron proposed openstack/manila master: Do not merge - baseline  https://review.openstack.org/42846311:46
openstackgerritRodrigo Barbieri proposed openstack/manila-specs master: Update Share Migration Ocata Improvements Spec  https://review.openstack.org/40714511:57
*** dsariel has joined #openstack-manila12:02
openstackgerritValeriy Ponomaryov proposed openstack/manila master: Disable share groups APIs by default  https://review.openstack.org/42884012:10
openstackgerritValeriy Ponomaryov proposed openstack/manila master: Disable share groups APIs by default  https://review.openstack.org/42884012:11
*** surabujin has quit IRC12:14
openstackgerritMerged openstack/puppet-manila master: Prepare ocata-rc1 release  https://review.openstack.org/42908912:22
*** surabujin has joined #openstack-manila12:26
tommylikehuping vponomaryov12:27
vponomaryovtommylikehu: зщтп12:50
vponomaryovtommylikehu: pong12:50
*** _fatalis is now known as aimtsou12:50
arnewiebalck_tbarron: tommylikehu: for our cloud, the lack of Manila OSC support meant additional clients, creation of pre-configured docker images, additional documentation, and certainly user surprise that they have to use a dedicated client while we tell them since a while to move to the OSC for other projects.12:55
tbarronarnewiebalck_: good data point - you have that little CERN cloud, right? -D12:58
arnewiebalck_tbarron: yes :)12:58
tbarronarnewiebalck_: Are by any chance going to be at the PTG in Atlanta?12:59
arnewiebalck_tbarron: no, I won’t; there will be someone from our team, though; I’ll be at the summit.13:01
tbarronarnewiebalck_: ok, I'm just interested in getting real deployers interacting with the manila developers.13:03
*** catintheroof has joined #openstack-manila13:03
arnewiebalck_tbarron: I’ve submitted a talk for the summit :)13:03
tbarronarnewiebalck_: good!13:04
tbarronarnewiebalck_: subject? (besides the mandatory LHC slides, etc.)13:05
arnewiebalck_tbarron: basically how we moved Manila into production13:06
tbarronarnewiebalck_: awesome.13:06
arnewiebalck_tbarron: we’re about to do some scale/stress tests maybe 1’000 concurrent clients (if we manage), suggestions what to test or look for are more than welcome!13:07
arnewiebalck_tbarron: I was about to ask that but figured that all of you are pretty busy atm ;)13:07
vponomaryovarnewiebalck_: just curios, which backends do you use in prod? or it is secret?13:14
arnewiebalck_vponomaryov: we use CephFS13:15
vponomaryovarnewiebalck_: you use UI part too?13:16
arnewiebalck_vponomaryov: yes (hence my patches in the past months ;-))13:17
arnewiebalck_vponomaryov: for clarification, Manila is kind of in pre-prod atm, it should have made to full production by summit time13:18
vponomaryovarnewiebalck_: good to know, thank you13:18
vponomaryovarnewiebalck_: also, if not secret, what is use case for using Manila?13:21
vponomaryovarnewiebalck_: in your case13:21
arnewiebalck_vponomaryov: at least ath moment, Manila and CephFS are closely linked in our case13:23
vponomaryovarnewiebalck_: I mean waht do you do with "shares"13:24
arnewiebalck_vponomaryov: we have users that request CephFS (e.g. HPC for POSIX) and Manila is a convenient way to hand out shares13:24
arnewiebalck_vponomaryov: CephFS is also a candidate to provide reliable shared storage13:25
arnewiebalck_vponomaryov: we have NFS for this at the moment13:25
vponomaryovarnewiebalck_: so, it is end-user storage place? Not back-up place, not DB-layout?13:25
arnewiebalck_vponomaryov: yes13:25
arnewiebalck_vponomaryov: we plan to have running services on top of this13:26
arnewiebalck_vponomaryov: services and HPC applications13:27
vponomaryovarnewiebalck_: interesting, thank you )13:27
arnewiebalck_vponomaryov: but it is not end-user in the sense of home directory13:27
vponomaryovarnewiebalck_: under "end-user" I mean VM/machine that is used by concrete person for his own needs, not by service nodes13:29
arnewiebalck_vponomaryov: it’s meant for services (e.g. our HTCondor masters, our GitLab nodes, the JIRA cluster, …)13:32
vponomaryovok )13:32
vponomaryovarnewiebalck_:and now you use "shares" too, but provisioned by other means than Manila, right?13:33
arnewiebalck_vponomaryov: yes, we have several storage solutions to serve the various needs; they are prvisioned in different ways; we try now to provide CephFS exclusively through Manila13:35
vponomaryovarnewiebalck_: and what about network segregation?13:37
vponomaryovarnewiebalck_: do you need it in your cloud?13:37
vponomaryovarnewiebalck_: or usage of same network for all is ok?13:37
arnewiebalck_vponomaryov: segregation between?13:37
vponomaryovarnewiebalck_: tenants or reguions...13:37
vponomaryovs/reguions/regions/13:38
arnewiebalck_vponomaryov: anyway, the answer is “no" :-D13:38
vponomaryovarnewiebalck_: ok ))13:38
*** dustins has joined #openstack-manila13:49
*** gouthamr has joined #openstack-manila13:58
tommylikehuthanks vponomaryov, I think your comments have been addressed: https://review.openstack.org/#/c/406776/, also paste my configuration for dummy drivers there :)14:00
tommylikehuarnewiebalck_: thanks for your explanation14:01
arnewiebalck_tommylikehu: np!14:02
openstackgerritValeriy Ponomaryov proposed openstack/manila master: Fix default approach for share group snapshot creation  https://review.openstack.org/42781114:04
*** jprovazn has quit IRC14:05
openstackgerritValeriy Ponomaryov proposed openstack/manila master: Fix default approach for share group snapshot creation  https://review.openstack.org/42781114:07
*** chlong has joined #openstack-manila14:19
*** breitz has quit IRC14:20
*** eharney has joined #openstack-manila14:25
*** jprovazn has joined #openstack-manila14:27
openstackgerritValeriy Ponomaryov proposed openstack/manila master: Add 'consistent_snapshot_support' attr to 'share_groups' DB model  https://review.openstack.org/42874814:29
*** eharney has quit IRC14:31
*** Yogi1 has joined #openstack-manila14:41
*** eharney has joined #openstack-manila14:44
openstackgerritMatt Riedemann proposed openstack/manila master: WIP: Revert "Fix Generic driver DHSS=False setup"  https://review.openstack.org/42972914:45
*** rraja has quit IRC15:07
*** gcb has joined #openstack-manila15:17
*** shausy has quit IRC15:18
*** breitz has joined #openstack-manila15:18
*** jcsp has joined #openstack-manila15:19
*** wiebalck has joined #openstack-manila15:33
*** xyang_ has joined #openstack-manila15:33
*** cdelatte has joined #openstack-manila15:38
*** xyang_ has quit IRC15:46
openstackgerritMerged openstack/manila master: Fix devstack manila nfs install for fedora  https://review.openstack.org/41223215:53
*** xyang_ has joined #openstack-manila15:55
*** xyang_ has quit IRC15:56
openstackgerritTom Barron proposed openstack/manila master: DNM: Update our tempest pin  https://review.openstack.org/42823015:58
*** wiebalck has quit IRC16:16
bswartzvponomaryov: jenkins gave you 2 -1s16:17
*** breitz has quit IRC16:21
*** lpetrut has quit IRC16:29
*** jprovazn is now known as jprovazn_bbl16:30
*** gcb has quit IRC16:33
*** shausy has joined #openstack-manila16:36
*** pcaruana has quit IRC16:38
*** crushil has joined #openstack-manila16:39
vponomaryovbswartz: I see, one is false and one is correct16:46
*** mriedem has joined #openstack-manila17:03
mriedemvponomaryov: i reverted the workaround for the manila devstack setup for that nova server list bug, and ci all passed https://review.openstack.org/#/c/429729/17:04
vponomaryovmriedem: have you seen my comment there?17:04
mriedemso we might have already fixed whatever the issue was - we had several fixes going into nova last week related to that code, so something might have already got it17:04
mriedemvponomaryov: so manila devstack wasn't creating the cells v2 stuff before?17:04
mriedemif so, then the bug is invalid17:04
vponomaryovmriedem: the bug is about "hiding" servers17:05
vponomaryovmriedem: and it exists17:05
mriedemvponomaryov: cells v2 is required17:05
mriedemand the cells v2 setup is required17:05
mriedemso if that wasn't happening before, then that was at least part of the issue17:05
vponomaryovmriedem: how is ti related for hiding servers?17:05
mriedemvponomaryov: if there was an issue, i'm saying it might have been fixed last week when we fixed several other related issues,17:06
vponomaryovmriedem: no, devstack creates cell now earlier17:06
mriedemand since i've reverted that workaround for devstack setup in manila, and it all passed, it seems it's not an issue17:06
vponomaryovmriedem: we do not reproduce bug steps anymore17:06
vponomaryovmriedem: looks like you haven't read my comment in that revert commit yet17:07
mriedem"devstack creates cell now earlier" - was there a change in devstack for that or in manila?17:07
mriedemi just did17:07
vponomaryovmriedem: I haven't searched for it, but I see order of steps in devstack logs17:07
*** breitz has joined #openstack-manila17:07
vponomaryovmriedem: now it creates cell much earlier17:07
vponomaryovmriedem: and manila runs now after it17:07
vponomaryovmriedem: so, devstack setup was fixed17:08
vponomaryovbut not Nova API17:08
vponomaryovmriedem: bug still exist17:08
mriedemi don't see anything in https://github.com/openstack/manila/commits/master/devstack that shows a change like that17:08
mriedemor https://github.com/openstack-dev/devstack/commits/master/lib/nova17:08
vponomaryovmriedem: why are you looking in manila code?17:08
mriedembecause you guys have a devstack plugin17:09
vponomaryovmriedem: cell creation runs earlier17:09
vponomaryovmriedem: manila does not touch it17:09
mriedemthere is nothing after 1/24 in either of those places which changes when the cell is created17:09
mriedemso, it's possible that one of the fixes that went into nova last week fixed this issue17:10
vponomaryovmriedem: Matt ))17:10
vponomaryovmriedem: you mean this -> https://review.openstack.org/#/c/427394/2/nova/compute/api.py ?17:13
mriedemthat was one of them but you're not running cells v1 in your manila dsvm jobs,17:14
mriedemso i don't think that would make any difference here17:14
mriedemthat was to fix a problem we were seeing in the cells v1 job17:14
vponomaryovright17:15
mriedemif you didn't have cells setup properly when you hit this, you'd see this error in the logs https://review.openstack.org/#/c/427394/2/nova/compute/api.py@253617:15
mriedemdo you have a link to a patch that failed with this bug?17:15
vponomaryovmriedem: looking for it17:17
vponomaryovmriedem: http://logs.openstack.org/64/355264/26/check/gate-manila-tempest-dsvm-generic-no-share-servers-ubuntu-xenial-nv/889af0f/logs/devstacklog.txt.gz17:19
mriedemi don't see simple_cell_setup run in there17:20
vponomaryovmriedem: right17:20
vponomaryovmriedem: it was ruinning at that time right before "test-config" section in devstack17:20
*** harlowja has joined #openstack-manila17:20
vponomaryovmriedem: so, noone really could create servers during devtack run17:21
vponomaryovmriedem: now can17:21
vponomaryovmriedem: because this step was moved earlier17:21
vponomaryovmriedem: cell creation step17:21
mriedemit wasn't moved in devstack,17:21
vponomaryovmriedem: but it does not remove bug in Nova itself17:21
mriedemso i'm not sure what changed17:21
mriedemvponomaryov: cells v2 must be setup before you can do anything with nova-api in ocata17:21
mriedemperiod17:21
mriedemit's an upgrade implication17:22
mriedemcells v2 is required in ocata17:22
vponomaryovmriedem: it is one of possible reproduce approaches17:22
vponomaryovmriedem: bug can be reproduced even with set up cells17:22
vponomaryovmriedem: just need to get error before assigning cell to server17:22
vponomaryovmriedem: or list servers before it got schecduled17:22
vponomaryovmriedem: it is concurrency-prone17:22
mriedemand i already said we fixed a few bugs with races in that space last week17:22
mriedemconcurrency issues17:23
openstackgerritMerged openstack/puppet-manila master: Add rpc_response_timeout option  https://review.openstack.org/42951417:23
mriedemlikely this https://github.com/openstack/nova/commit/8ba92778fe14b47ad4ff5b53022e0550a93f37d317:23
mriedemactually that shouldn't be related to your issue,17:24
mriedemsince your flow isn't deleting the server before it's active17:24
vponomaryovmriedem: yes, bug being descussed is only API-related and listing scheduled servers17:25
mriedemhttps://review.openstack.org/#/c/319379/ merged on 1/27 it's possible something was going wrong between those 2 changes17:25
openstackgerritMatt Riedemann proposed openstack/manila master: Revert "Fix Generic driver DHSS=False setup"  https://review.openstack.org/42972917:26
vponomaryovmriedem: no need to revert, current approach uses less API calls17:27
mriedemok abandoned17:28
vponomaryovmriedem: ok, considering that, bug I reported is low-priority, we can leave it unfixed, even though I haven't convinced anybody in its existence )17:30
*** zhonghua has quit IRC17:37
*** zhonghua has joined #openstack-manila17:38
*** crushil has quit IRC17:41
*** crushil has joined #openstack-manila17:42
*** harlowja has quit IRC17:49
*** shausy has quit IRC17:50
*** lpetrut has joined #openstack-manila17:57
openstackgerritValeriy Ponomaryov proposed openstack/manila master: Add 'consistent_snapshot_support' attr to 'share_groups' DB model  https://review.openstack.org/42874817:58
*** mriedem has left #openstack-manila18:08
openstackgerritMerged openstack/manila master: Improve test coverage for share migration  https://review.openstack.org/41855918:13
*** crushil has quit IRC18:22
*** breitz has quit IRC18:30
openstackgerritRodrigo Barbieri proposed openstack/manila-specs master: Update Share Migration Ocata Improvements Spec  https://review.openstack.org/40714518:35
*** pcaruana has joined #openstack-manila18:36
*** nkrinner is now known as nkrinner_afk18:40
*** harlowja has joined #openstack-manila18:44
*** crushil has joined #openstack-manila18:45
*** xyang1 has joined #openstack-manila18:48
openstackgerritRodrigo Barbieri proposed openstack/manila master: Add Share Migration devref docs  https://review.openstack.org/37577118:51
*** gouthamr has quit IRC18:52
*** chlong has quit IRC18:52
*** gouthamr has joined #openstack-manila18:58
*** yumiriam has quit IRC19:00
*** crushil has quit IRC19:00
*** porrua has joined #openstack-manila19:05
*** chlong has joined #openstack-manila19:08
*** pcaruana has quit IRC19:11
*** crushil has joined #openstack-manila19:12
*** kberger has joined #openstack-manila19:37
*** kberger has quit IRC19:37
*** kberger has joined #openstack-manila19:38
*** dsariel has quit IRC19:39
*** mtanino has joined #openstack-manila20:11
*** dsariel has joined #openstack-manila20:17
*** breitz has joined #openstack-manila20:45
*** breitz has quit IRC20:45
*** breitz_ has joined #openstack-manila20:45
*** porrua has quit IRC20:46
*** cdelatte has quit IRC20:46
*** chlong has quit IRC20:49
*** gouthamr has quit IRC20:56
openstackgerritTom Barron proposed openstack/manila master: DNM: Update our tempest pin  https://review.openstack.org/42823021:13
*** ganso has quit IRC21:14
*** gouthamr has joined #openstack-manila21:16
*** jprovazn_bbl has quit IRC21:19
*** Yogi1 has quit IRC21:27
*** crushil has quit IRC21:43
*** eharney has quit IRC21:53
*** dustins has quit IRC22:15
*** gouthamr has quit IRC22:17
*** jcsp has quit IRC22:33
*** cdelatte has joined #openstack-manila22:53
openstackgerritTom Barron proposed openstack/manila master: Do not merge - baseline  https://review.openstack.org/42846322:53
*** cdelatte has quit IRC22:55
*** lpetrut has quit IRC23:02
*** crushil has joined #openstack-manila23:18
*** zhonghua2 has joined #openstack-manila23:24
*** zengyingzhe_ has joined #openstack-manila23:25
*** DuncanT_ has joined #openstack-manila23:27
*** xyang1 has quit IRC23:28
*** cargonza_ has joined #openstack-manila23:28
*** netapp_ has joined #openstack-manila23:31
*** zhonghua has quit IRC23:34
*** zengyingzhe has quit IRC23:34
*** DuncanT has quit IRC23:34
*** netapp has quit IRC23:34
*** cargonza has quit IRC23:34
*** netapp_ is now known as netapp23:34
*** DuncanT_ is now known as DuncanT23:36
*** cargonza_ is now known as cargonza23:37
*** catintheroof has quit IRC23:54
*** gouthamr has joined #openstack-manila23:56

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