Wednesday, 2019-03-20

*** lseki has quit IRC00:09
*** dviroel_ has quit IRC01:17
*** senrique__ has joined #openstack-manila01:19
*** whoami-rajat has joined #openstack-manila01:29
*** vh_ has joined #openstack-manila02:00
*** senrique__ has quit IRC02:04
openstackgerritMerged openstack/manila master: NetApp cDOT assume disabled compression on empty result  https://review.openstack.org/61957002:08
*** senrique__ has joined #openstack-manila02:20
*** senrique__ has quit IRC02:27
*** vh_ has quit IRC02:43
openstackgerritMerged openstack/manila stable/rocky: Fix driver filter to not check share_backend_name  https://review.openstack.org/64376703:44
openstackgerritGoutham Pacha Ravi proposed openstack/manila stable/queens: Fix driver filter to not check share_backend_name  https://review.openstack.org/64472303:54
*** e0ne has joined #openstack-manila07:17
*** pcaruana has joined #openstack-manila07:20
*** e0ne has quit IRC07:27
*** tosky has joined #openstack-manila07:29
*** e0ne has joined #openstack-manila07:42
*** e0ne has quit IRC08:15
openstackgerritMerged openstack/manila master: Check all_tenants value in security_service api  https://review.openstack.org/62019709:58
openstackgerritHelen Walsh proposed openstack/manila stable/rocky: VMAX manila - deprecate old tags correctly  https://review.openstack.org/64480910:12
openstackgerritMerged openstack/manila master: Check all_tenants value in share_networks api  https://review.openstack.org/61899110:14
openstackgerritMerged openstack/puppet-manila master: Prepare Stein RC1  https://review.openstack.org/64382610:15
openstackgerritHelen Walsh proposed openstack/manila stable/queens: VMAX manila doc - use of correct VMAX tags  https://review.openstack.org/64481410:29
openstackgerritHelen Walsh proposed openstack/manila stable/queens: VMAX manila doc - SSL Support  https://review.openstack.org/64481510:33
openstackgerritHelen Walsh proposed openstack/manila stable/queens: Manila VMAX docs -  clarify driver_handles_share_servers  https://review.openstack.org/64481910:35
*** e0ne has joined #openstack-manila10:45
*** carlos_silva has joined #openstack-manila11:06
toabctltbarron, gouthamr is anybody still using run_tests.sh ?11:07
openstackgerritThomas Bechtold proposed openstack/manila master: Drop run_tests.sh and tools/colorizer.py  https://review.openstack.org/64482611:10
*** vh_ has joined #openstack-manila11:15
*** vh_ has left #openstack-manila11:18
*** vh_ has joined #openstack-manila11:19
openstackgerritHelen Walsh proposed openstack/manila stable/queens: Manila VMAX docs -  clarify snapshot support  https://review.openstack.org/64482911:22
*** thgcorrea has joined #openstack-manila11:27
*** pcaruana has quit IRC11:38
*** dviroel_ has joined #openstack-manila11:44
openstackgerritHelen Walsh proposed openstack/manila stable/queens: Manila VMAX docs -  improve pre-configurations on VMAX section  https://review.openstack.org/64483811:52
openstackgerritHelen Walsh proposed openstack/manila stable/queens: Manila VMAX docs -  differences between quotas  https://review.openstack.org/64484011:56
*** e0ne has quit IRC12:07
*** pcaruana has joined #openstack-manila12:16
*** eharney has joined #openstack-manila12:42
*** rraja has joined #openstack-manila12:48
*** rraja has quit IRC12:54
tbarrontoabctl: no run_tests.sh is deprecated12:55
tbarrontoabctl: has been for several releases12:56
toabctltbarron, oki. I created already a changeset to remove it to see what the CI says12:56
toabctlhttps://review.openstack.org/64482612:56
tbarrontoabctl: did you grep through our doc to see if we still mention it anywhere?12:56
toabctlI didn't find anything12:57
*** jmlowe has quit IRC12:58
tbarrontoabctl: thanks13:00
*** lseki has joined #openstack-manila13:09
*** thgcorrea has quit IRC13:15
*** e0ne has joined #openstack-manila13:16
openstackgerritMerged openstack/manila stable/queens: Fix driver filter to not check share_backend_name  https://review.openstack.org/64472313:28
*** senrique__ has joined #openstack-manila13:42
*** senrique__ has quit IRC13:44
*** enriquetaso has joined #openstack-manila13:44
*** thgcorrea has joined #openstack-manila13:48
*** jgrosso has joined #openstack-manila13:56
*** eharney has quit IRC14:03
openstackgerritMerged openstack/python-manilaclient master: Update master for stable/stein  https://review.openstack.org/64416414:04
openstackgerritMerged openstack/python-manilaclient stable/stein: Update .gitreview for stable/stein  https://review.openstack.org/64416214:13
openstackgerritMerged openstack/manila stable/queens: Destroy type quotas when a share type is deleted  https://review.openstack.org/64458514:14
openstackgerritGoutham Pacha Ravi proposed openstack/manila stable/pike: Destroy type quotas when a share type is deleted  https://review.openstack.org/64487714:37
openstackgerritMerged openstack/manila stable/rocky: VMAX manila - deprecate old tags correctly  https://review.openstack.org/64480914:45
openstackgerritMerged openstack/manila stable/queens: VMAX manila doc - use of correct VMAX tags  https://review.openstack.org/64481414:45
openstackgerritMerged openstack/manila stable/queens: VMAX manila doc - SSL Support  https://review.openstack.org/64481514:45
openstackgerritMerged openstack/manila stable/queens: Manila VMAX docs -  clarify driver_handles_share_servers  https://review.openstack.org/64481914:45
openstackgerritGoutham Pacha Ravi proposed openstack/manila stable/pike: Fix driver filter to not check share_backend_name  https://review.openstack.org/64488614:52
tbarroncarlos_silva: lseki: how is it going with https://review.openstack.org/#/c/644560/ ?14:58
*** jmlowe has joined #openstack-manila15:00
*** enriquetaso has quit IRC15:08
carlos_silvalseki: is going to propose a new patch in few hours15:15
carlos_silvawe are waiting for our internal CI output before creating a new patch15:16
carlos_silvas/a new patch/a new upstream patch15:18
carlos_silvatbarron lseki changed some things in manila-tempest-plugin and the changes seems effective15:24
tbarroncarlos_silva: lseki cool, just checking :)15:28
*** eharney has joined #openstack-manila15:40
openstackgerritLucio Seki proposed openstack/manila-tempest-plugin master: Delete servers before deleting its share network  https://review.openstack.org/64490515:54
openstackgerritCarlos Eduardo proposed openstack/manila master: Fix server delete attempt along with share net deletion  https://review.openstack.org/64456016:01
tbarroncarlos_silva: lseki fingers crossed :)16:02
lsekitbarron, carlos_silva great!16:04
carlos_silva:)16:07
*** enriquetaso has joined #openstack-manila16:11
openstackgerritjacky06 proposed openstack/python-manilaclient master: Update hacking version  https://review.openstack.org/64139916:12
openstackgerritMerged openstack/python-manilaclient stable/stein: Update UPPER_CONSTRAINTS_FILE for stable/stein  https://review.openstack.org/64416316:21
*** jmlowe has quit IRC16:21
*** altlogbot_2 has joined #openstack-manila16:40
*** altlogbot_2 has quit IRC16:41
*** jmlowe has joined #openstack-manila16:48
*** thgcorrea has quit IRC16:55
*** thgcorrea has joined #openstack-manila16:57
*** e0ne has quit IRC17:07
*** altlogbot_1 has joined #openstack-manila17:24
*** altlogbot_1 has quit IRC17:24
*** altlogbot_2 has joined #openstack-manila17:26
*** altlogbot_2 has quit IRC17:28
*** altlogbot_0 has joined #openstack-manila17:29
*** raissa has joined #openstack-manila17:31
openstackgerritSofia Enriquez proposed openstack/manila-specs master: Manila support in python-openstackclient  https://review.openstack.org/64421817:55
openstackgerritLucio Seki proposed openstack/manila-tempest-plugin master: Delete servers before deleting its share network  https://review.openstack.org/64490518:00
openstackgerritGoutham Pacha Ravi proposed openstack/manila stable/pike: Destroy type quotas when a share type is deleted  https://review.openstack.org/64487718:00
*** enriquetaso has quit IRC18:02
openstackgerritCarlos Eduardo proposed openstack/manila master: Fix server delete attempt along with share net deletion  https://review.openstack.org/64456018:09
*** e0ne has joined #openstack-manila18:17
*** kopecmartin is now known as kopecmartin|off18:19
*** irclogbot_1 has quit IRC18:21
openstackgerritTom Barron proposed openstack/manila master: Only allow IP access type for CephFS NFS  https://review.openstack.org/64018518:23
*** irclogbot_3 has joined #openstack-manila18:24
*** irclogbot_3 has quit IRC18:27
*** irclogbot_3 has joined #openstack-manila18:28
*** amito has quit IRC18:31
*** amito has joined #openstack-manila18:33
tbarrongouthamr: carlos_silva lseki bswartz carthaca ganso jgrosso toabctl xyang here are our outstanding reviews on master:18:35
tbarronhttps://review.openstack.org/#/q/status:open+project:openstack/manila+branch:master18:35
tbarronThe only one left that we're trying to get in to rc1 is: https://review.openstack.org/#/c/644560/18:35
tbarronwhich depends on manila-tempest-plugin fix https://review.openstack.org/#/c/644905/18:36
tbarroni'm ok with the changes from code inspection18:38
gouthamrtbarron: ack18:38
tbarronvkmc: gouthamr ^^ have you had a chance to look at the tempest-plugin changes proposed?18:39
gouthamrtbarron: although, i want to ask a question of ganso about what he thought about this...18:39
gouthamrtbarron: technically, when a share server is unmanaged, you should be able to delete the share network, and not delete the unmanaged share server...18:39
gouthamrtbarron: don't understand entirely *why* we'd want to stop users from removing it..18:40
gouthamris it because of port allocations from neutron that need to go away?18:40
tbarrongouthamr: that was my assumption18:41
gouthamrthat'd be wrong, wouldn't it... i mean, we don't own those ports, the administrator does..18:42
gouthamri'll just recheck that code path, afaiu, case 1) Share server is created by manila, but admin unmanaged a share - so, share server isn't going to be deleted... admin needs to do it, it makes sense to keep the share network around, unless the administrator wants to unmanage or delete the share server first...18:45
tbarronif we delete the share network then it frees up the allocations so we no longer have neutron reservation of those resources18:45
gouthamra lesser privileged user in the project may want to cleanup the share network, but can't because there's a share server they don't know about that is blocking the share network deletion18:45
tbarrongouthamr: ok follow up with ganso and if there's an issue here file a bug18:48
gouthamri think in this situation, we can prevent the share network deletion but let the error message be a little more user friendly: i.e, "there may be shares on this share network that we don't know about. Please contact your administrator to clarify"18:48
tbarrongouthamr: oh, you are suggesting a tweak.18:48
tbarrongouthamr: post it right away then.18:49
tbarrongouthamr: we have a couple of hours while https://review.openstack.org/#/c/644560/ and https://review.openstack.org/#/c/644905/ make their way through the system18:49
tbarrongouthamr: we can look at that tweak in parallel18:50
gouthamrack18:50
tbarrononce we merge these though I plan to propose rc118:50
tbarronother projects have already released ...18:50
tbarrondon't want to get caught in a logjam18:50
tbarronany bug fixes are fair to merge until rc1 is proposed though18:51
tbarronIf there's nothing else on rc1 let's talk backports since I want to propose stable/* releases this week as well18:52
*** e0ne has quit IRC18:53
tbarronwe don't have anything pending on stable/rocky any more: https://review.openstack.org/#/q/status:open+project:openstack/manila+branch:stable/rocky18:53
gouthamrack on rocky18:54
tbarronstable queens has no critical backports: https://review.openstack.org/#/q/status:open+project:openstack/manila+branch:stable/queens18:54
tbarronthere's a couple doc backports and one can debate whether they should be backported or not but they're harmless18:55
gouthamrack on queens as well, the ones you workflowed don't need to be in a release; i.e, they'll update https://docs.openstack.org/manila/queens18:55
tbarrongouthamr: right18:55
tbarronstable/pike is going to go EOM soon18:56
gouthamrhttps://review.openstack.org/#/q/status:open+project:openstack/manila+branch:stable/pike has a couple of my backports that would be nice before marking EM18:56
gouthamrEOM != EM :)18:56
tbarronyeah18:56
tbarronpicky picky picky18:56
gouthamr:P18:57
tbarronbut correct!18:57
tbarronLet's see if we can get those two merged18:57
tbarrongouthamr: actually I'd like to merge https://review.openstack.org/#/c/635887/ as well18:58
tbarronit's test-only but might as well include it in the last release18:59
gouthamrtbarron: ack, so we can fix the tempest run within the ceph devstack plugin?18:59
tbarrongouthamr: that's a separate fix, will need to be done in branchelss devstack-plugin-ceph19:00
*** altlogbot_0 has quit IRC19:00
tbarrongouthamr: to have it run tempest in the way pike wants on pike and the way it currently runs on more recent branches19:00
tbarronugly19:00
gouthamrtbarron: ack, i wonder if you can write a condition on ZUUL_BRANCH and run tox -eplugin instead of "tempest run"19:01
tbarronbut this fix will get devstack completing instead of failing b/c of conflicting ceph versions19:01
tbarrongouthamr: maybe, I'll have to pick your brain19:01
gouthamrtbarron: +1, W'ed19:01
tbarrongouthamr: thanks19:01
*** altlogbot_3 has joined #openstack-manila19:02
tbarronOK, I think that does it for backports?19:02
tbarronI'll start cutting releases after we see the stuff we mentioned merge (and after my workday meetings).19:03
gouthamryes19:03
tbarrongouthamr: thanks for the help!19:03
gouthamrthank you tbarron19:03
gouthamrwe might need to work on the rc1 blocker a bit more.. the latest change breaks tempest for the dummy driver19:03
*** altlogbot_3 has quit IRC19:06
*** altlogbot_0 has joined #openstack-manila19:07
*** jmlowe has quit IRC19:15
*** irclogbot_3 has quit IRC19:16
*** irclogbot_2 has joined #openstack-manila19:17
*** raissa has quit IRC19:26
*** irclogbot_2 has quit IRC19:30
*** jmlowe has joined #openstack-manila19:30
*** irclogbot_3 has joined #openstack-manila19:31
*** altlogbot_0 has quit IRC19:34
*** altlogbot_3 has joined #openstack-manila19:35
gansogouthamr: hey19:39
gansogouthamr: what is that you would like to ask?19:39
gansogouthamr: I was in a meeting, saw the notification just now19:40
*** eharney has quit IRC19:50
*** thgcorrea has quit IRC19:59
gouthamrganso: hey, thanks for the ping.. was in a meeting myself20:00
gouthamrganso: have you seen the bug: https://bugs.launchpad.net/manila/+bug/182011820:01
openstackLaunchpad bug 1820118 in Manila "Manila unexpectedly attempts to delete a share server when deleting a share network" [Medium,In progress] - Assigned to Carlos Eduardo (silvacarlose)20:01
gansogouthamr: yes20:01
gouthamrganso: and the fix proposed: https://review.openstack.org/#/c/644905/20:01
gansogouthamr: no20:02
* ganso looks20:02
gouthamrganso: cool, please review it if possible.. i feel it leaks out the share server abstraction a bit20:02
*** eharney has joined #openstack-manila20:02
gansogouthamr: why20:02
gouthamrganso: i'd like to know what you think, and how we should handle it..20:03
gouthamrganso: by default, in our perception of how this is used... non-privileged users can delete share networks since they can create them20:03
*** vh_ has quit IRC20:03
gouthamrganso: but now we have something that an admin does (unmanaging a share / managing a share server into the user's share network) that prevents them from deleting their share networks20:04
gansogouthamr: I agree with https://review.openstack.org/#/c/644560, but the error message should be user-facing. Users don't know what share servers are20:04
gouthamrganso: yes, 1) i think the error message can be more appropriate...20:05
openstackgerritMerged openstack/manila master: Drop run_tests.sh and tools/colorizer.py  https://review.openstack.org/64482620:05
openstackgerritMerged openstack/manila stable/queens: Manila VMAX docs -  clarify snapshot support  https://review.openstack.org/64482920:05
openstackgerritMerged openstack/manila stable/queens: Manila VMAX docs -  improve pre-configurations on VMAX section  https://review.openstack.org/64483820:05
openstackgerritMerged openstack/manila stable/queens: Manila VMAX docs -  differences between quotas  https://review.openstack.org/64484020:05
gansogouthamr: I don't agree with the test patch. I don't think the core needs to be changed (as the patch is doing), but a separate test should be created to test the new error message20:07
gouthamrganso: 2) i'm thinking where this might go wrong.. maybe this stuff can be clarified by documentation somewhere... unmanaging a share from a share server or having a managed share server on a share network does not allow users to delete the underlying share network20:07
gansogouthamr: and the test patch should depend on the fix. Nothing else should change20:08
gansogouthamr: /the core needs to be changed/the base class needs to be changed20:08
gansogouthamr: and I think it is better for us to agree on the proper solution here first before -1'ing the patch20:09
gouthamrganso: yes, unsure if lseki, carlos_silva are still around20:10
gansogouthamr: I don't see a problem with #2 you mentioned above20:10
carlos_silvayeah, I'm here20:10
gansogouthamr: as long as the error message is clear20:10
gansogouthamr: something along the lines of "there may be shares on this share network that we don't know about. Please contact your administrator to clarify" as you suggested should be fine20:11
lsekiganso: > I don't think the base class needs to be changed20:12
lsekiI needed to change it to make sure that the share-servers are deleted before the share-network upon testclass teardown20:12
gouthamrganso: cool, that works for me... i think it'd be prudent to also log a warning message in the share manager when a share server is managed or when a share is unmanaged that the share network will not be deleted unless the associated share server is unmanaged or deleted20:13
lsekiotherwise it would raise the exception carlos_silva's patch added, sporadically20:13
gouthamrganso: hmmm, i suggested they handle this cleanup in the base class, do you think we should rather do this everywhere the conditions above are true? (share unmanaged from share server, or share server is managed)20:14
gansolseki: it isn't a problem for any of the current tests. And by making sure you delete the share server in all of the manage/unmanage tests (where is_auto_deletable will be False) will be making sure of that. The change as you proposed affects all tests and then all other tests will stop testing the deletion of share servers when deleting a share network (when all share servers are is_auto_deletable = True)20:14
gansogouthamr: yes20:15
gansogouthamr: I don't think the warning is needed, but it is ok to add20:15
*** eharney has quit IRC20:22
openstackgerritMerged openstack/manila stable/pike: Only run the needed services for CephFS jobs  https://review.openstack.org/63588720:22
lsekiganso: gouthamr  we realized the need to adapt the functional tests when several tests failed while cheking carlos_silva's patch, including those tests that don't manage/unmanage things at all (such as AvailabilityZonesTest, ShareLimitsTest, etc).20:27
gansowhy? why carlos_silva's change broke other tests where is_auto_deletable=True ?20:28
lsekiganso my suspect is that teardown of those test classes is trying to delete the share-network before share-servers are deleted20:30
gansolseki: then that problem would already be there and unrelated to carlos_silva's changes20:30
lsekiganso yes, but the error was not being raised, until carlos_silva's change started to block it at the api20:31
gansolseki: perhaps this needs further invesgation then, as it wasn't supposed to happen20:33
lsekiand by forcing the share-servers' deletion before the share-network deletion, modifying the base class, those errors stopped appearing20:34
carlos_silvaganso: you can check the tests output before lseki changes here: http://13.56.179.158/logs/19/1019/3/downstream-functional/manila-cDOT-ss/418763a/console.html.gz20:35
gansolseki: what if you delete the share server within the test manila_tempest_tests.tests.api.admin.test_share_manage.ManageNFSShareTest ?20:36
gansolseki: you are probably going to need 3 patches20:37
gansolseki: 1) to change how share servers are deleted on any test that causes is_auto_deletable to be Fals20:37
ganso2) carlos_silva's changes20:38
ganso3) new test that validates carlos_silva's changes20:38
lseki> what if you delete the share server within the test ...20:39
lsekiganso: it would make the ManageNFSShareTest tests pass, but probably keep the other tests (that doesn't manage/unmanage at all) failing20:39
gansolseki: I think it is worth a shot. The reason other tests which seem unrelated are failing is most likely because they are sharing the same share_network as the ManageNFSTest class20:40
gansolseki: I didn't check the code of those tests to see if they are creating a share_network of their own, but I would assume they are not20:42
*** irclogbot_3 has quit IRC20:45
*** irclogbot_0 has joined #openstack-manila20:47
*** e0ne has joined #openstack-manila20:52
*** altlogbot_3 has quit IRC20:53
*** altlogbot_1 has joined #openstack-manila20:55
openstackgerritCarlos Eduardo proposed openstack/manila master: Fix server delete attempt along with share net deletion  https://review.openstack.org/64456020:58
openstackgerritSoledad Kuczala proposed openstack/manila master: Adding documentation for user messages in Manila  https://review.openstack.org/64501021:14
*** eharney has joined #openstack-manila21:19
openstackgerritSoledad Kuczala proposed openstack/manila master: Adding documentation for user messages in Manila  https://review.openstack.org/64501021:19
lsekiganso: yeah, I tried with one of positive tests here and your approach worked21:24
lsekiso I'll change every single test that unmanages share/share-server and delete the share-server manually before the test ends21:25
*** jgrosso has quit IRC21:25
lsekiand remove the change in the base class21:27
lsekiand cross fingers hoping that all other tests will continue passing21:27
*** whoami-rajat has quit IRC21:29
*** e0ne has quit IRC22:00
*** pcaruana has quit IRC22:11
openstackgerritMerged openstack/manila stable/pike: Destroy type quotas when a share type is deleted  https://review.openstack.org/64487722:26
openstackgerritMerged openstack/manila stable/pike: Fix driver filter to not check share_backend_name  https://review.openstack.org/64488622:38
openstackgerritTom Barron proposed openstack/manila master: Only allow IP access type for CephFS NFS  https://review.openstack.org/64018522:45
openstackgerritGoutham Pacha Ravi proposed openstack/manila master: DNM: Test Python3 LVM Job  https://review.openstack.org/62306123:01
openstackgerritGoutham Pacha Ravi proposed openstack/manila master: DNM: Test Python3 LVM Job  https://review.openstack.org/62306123:04
*** tosky has quit IRC23:31
openstackgerritCarlos Eduardo proposed openstack/manila master: Fix server delete attempt along with share net deletion  https://review.openstack.org/64456023:41

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!