| *** yangyape_ has quit IRC | 00:05 | |
| *** ayma2 has quit IRC | 00:07 | |
| *** lpetrut has quit IRC | 00:09 | |
| *** akshai has quit IRC | 00:14 | |
| *** gouthamr_ has joined #openstack-manila | 00:16 | |
| *** gouthamr has quit IRC | 00:19 | |
| *** openstackgerrit_ has quit IRC | 00:23 | |
| *** openstackgerrit_ has joined #openstack-manila | 00:24 | |
| *** openstackgerrit_ has quit IRC | 00:48 | |
| *** openstackgerrit_ has joined #openstack-manila | 00:49 | |
| *** chlong has joined #openstack-manila | 00:52 | |
| *** yangyapeng has joined #openstack-manila | 01:11 | |
| *** yangyapeng has quit IRC | 01:15 | |
| *** a_ta has joined #openstack-manila | 01:18 | |
| *** furlongm has quit IRC | 01:20 | |
| *** esker has quit IRC | 01:21 | |
| *** yangyapeng has joined #openstack-manila | 01:22 | |
| *** a_ta has quit IRC | 01:22 | |
| *** esker has joined #openstack-manila | 01:22 | |
| *** resker has joined #openstack-manila | 01:35 | |
| *** esker has quit IRC | 01:37 | |
| openstackgerrit | zhongjun proposed openstack/manila: Create share API without checking share network in DHSS=true https://review.openstack.org/256353 | 01:42 |
|---|---|---|
| bswartz | ganso: ping | 01:43 |
| gouthamr_ | zhongjun: ping | 01:44 |
| bswartz | hey gouthamr | 01:44 |
| gouthamr_ | hey bswartz | 01:44 |
| zhongjun | gouthamr_: hi | 01:44 |
| bswartz | are we running migration tests in the gate jobs as far as you know? | 01:44 |
| gouthamr_ | bswartz: i guess on the lvm job. | 01:44 |
| gouthamr_ | bswartz: can check | 01:44 |
| bswartz | it's not urgent, but I'd like to know | 01:45 |
| gouthamr_ | bswartz: http://logs.openstack.org/43/244843/17/check/gate-manila-tempest-dsvm-neutron-postgres-lvm-multibackend/529c8fc/console.html | 01:45 |
| gouthamr_ | bswartz: yep | 01:45 |
| gouthamr_ | zhongjun: hello! :) I wanted to check with you about your snapshot_instances API work | 01:46 |
| gouthamr_ | zhongjun: are you planning on working on it for Newton? | 01:46 |
| zhongjun | gouthamr_: yes, I will continue to working on it. | 01:46 |
| gouthamr_ | zhongjun: awesome. we should sync up sometime.. I am implementing similar APIs for share replicas.. it would be duplication of a bunch of code; if you like, we can work on it together. | 01:47 |
| gouthamr_ | zhongjun: similar APIs as in, i would like to list snapshot_instances that are associated with replicas.. | 01:48 |
| zhongjun | gouthamr_: This is great, I am so happy to cooperate with you. :) | 01:49 |
| gouthamr_ | zhongjun: awesome. I guess i'll reach out to you over email .. would this email work? -> zhongjun <jun.zhongjun@huawei.com> | 01:51 |
| zhongjun | yes, I always use this email. | 01:52 |
| gouthamr_ | thanks zhongjun.. | 01:52 |
| zhongjun | thank you gouthamr. :) | 01:52 |
| *** gouthamr_ is now known as gouthamr | 01:53 | |
| *** mtanino has quit IRC | 02:02 | |
| *** houming has joined #openstack-manila | 02:03 | |
| openstackgerrit | zhongjun proposed openstack/manila: Create share API without checking share network in DHSS=true mode https://review.openstack.org/256353 | 02:05 |
| *** dalgaaf has quit IRC | 02:08 | |
| *** baojg has joined #openstack-manila | 02:24 | |
| openstackgerrit | zhongjun proposed openstack/manila: Create share API without checking share network in DHSS=true mode https://review.openstack.org/256353 | 02:59 |
| openstackgerrit | zhongjun proposed openstack/manila: Create share API without checking share network in DHSS=true mode https://review.openstack.org/256353 | 03:08 |
| *** baojg has quit IRC | 03:14 | |
| *** baojg has joined #openstack-manila | 03:16 | |
| *** mdnadeem_ has joined #openstack-manila | 03:21 | |
| *** baojg has quit IRC | 03:27 | |
| *** krotscheck has quit IRC | 03:30 | |
| *** mdnadeem has quit IRC | 03:30 | |
| *** krotscheck has joined #openstack-manila | 03:31 | |
| *** cknight has joined #openstack-manila | 03:33 | |
| *** cknight has quit IRC | 03:36 | |
| *** amrith is now known as _amrith_ | 04:10 | |
| *** baojg has joined #openstack-manila | 04:11 | |
| *** a_ta has joined #openstack-manila | 04:22 | |
| *** a_ta has quit IRC | 04:26 | |
| *** a_ta has joined #openstack-manila | 04:31 | |
| *** gouthamr has quit IRC | 04:32 | |
| *** a_ta has quit IRC | 04:36 | |
| *** baojg has quit IRC | 04:54 | |
| *** baojg has joined #openstack-manila | 04:54 | |
| *** nidhimittalhada has joined #openstack-manila | 04:57 | |
| *** nidhimittalhada has quit IRC | 05:02 | |
| *** nidhimittalhada has joined #openstack-manila | 05:03 | |
| *** jasonsb has joined #openstack-manila | 05:16 | |
| *** houming has quit IRC | 05:19 | |
| *** yangyape_ has joined #openstack-manila | 05:48 | |
| *** yangyapeng has quit IRC | 05:49 | |
| *** alyson_ has quit IRC | 05:56 | |
| *** krotscheck has quit IRC | 06:01 | |
| *** alyson_ has joined #openstack-manila | 06:02 | |
| *** vponomaryov has quit IRC | 06:03 | |
| *** vponomaryov has joined #openstack-manila | 06:04 | |
| *** krotscheck has joined #openstack-manila | 06:09 | |
| *** openstackgerrit_ has quit IRC | 06:17 | |
| *** openstackgerrit_ has joined #openstack-manila | 06:18 | |
| openstackgerrit | zhongjun proposed openstack/manila: Create share API without checking share network in DHSS=true mode https://review.openstack.org/256353 | 06:33 |
| *** baojg has quit IRC | 06:43 | |
| *** baojg has joined #openstack-manila | 06:43 | |
| *** lpetrut has joined #openstack-manila | 06:44 | |
| *** dalgaaf has joined #openstack-manila | 06:49 | |
| *** nkrinner has joined #openstack-manila | 06:50 | |
| *** baojg has quit IRC | 06:56 | |
| *** baojg has joined #openstack-manila | 06:57 | |
| nidhimittalhada | . | 06:59 |
| *** nidhimittalhada has quit IRC | 07:03 | |
| *** ayma has joined #openstack-manila | 07:08 | |
| *** ayma has quit IRC | 07:10 | |
| *** alyson_ has quit IRC | 07:10 | |
| *** absubram has joined #openstack-manila | 07:18 | |
| *** absubram_ has joined #openstack-manila | 07:19 | |
| *** ayma has joined #openstack-manila | 07:21 | |
| *** alyson_ has joined #openstack-manila | 07:21 | |
| openstackgerrit | Tom Patzig proposed openstack/manila: Update quota of proper user on resource delete https://review.openstack.org/289992 | 07:23 |
| *** absubram has quit IRC | 07:23 | |
| *** absubram_ is now known as absubram | 07:23 | |
| *** lpetrut has quit IRC | 07:30 | |
| *** absubram has quit IRC | 07:38 | |
| *** ayma has quit IRC | 07:47 | |
| *** alyson_ has quit IRC | 07:47 | |
| *** ayma has joined #openstack-manila | 07:53 | |
| *** alyson_ has joined #openstack-manila | 07:53 | |
| *** ayma has left #openstack-manila | 07:56 | |
| *** nidhimittalhada has joined #openstack-manila | 08:03 | |
| *** ayma1 has joined #openstack-manila | 08:14 | |
| *** ayma1 has left #openstack-manila | 08:14 | |
| *** aovchinnikov has joined #openstack-manila | 08:25 | |
| *** houming has joined #openstack-manila | 08:32 | |
| *** a_ta has joined #openstack-manila | 08:33 | |
| *** lpetrut has joined #openstack-manila | 08:35 | |
| *** a_ta has quit IRC | 08:37 | |
| *** pcaruana has joined #openstack-manila | 08:41 | |
| *** lpetrut has quit IRC | 08:44 | |
| *** chlong has quit IRC | 08:49 | |
| *** ayma has joined #openstack-manila | 08:58 | |
| *** ayma1 has joined #openstack-manila | 08:59 | |
| *** ayma has quit IRC | 09:02 | |
| *** rraja has joined #openstack-manila | 09:08 | |
| *** ayma1 has quit IRC | 09:23 | |
| mkoderer___ | vponomaryov: ping | 09:31 |
| mkoderer___ | toabctl: ping | 09:31 |
| toabctl | mkoderer___: pong | 09:32 |
| openstackgerrit | Tom Patzig proposed openstack/manila: Update quota of proper user on resource delete https://review.openstack.org/289992 | 09:33 |
| mkoderer___ | toabctl: I have a real issue with depreacted opts for keystone v3. keystone1 only support deprecated opts for sessions objects | 09:34 |
| mkoderer___ | but not for auth objects | 09:34 |
| *** nidhimittalhada has quit IRC | 09:35 | |
| mkoderer___ | toabctl: https://github.com/openstack/keystoneauth/blob/master/keystoneauth1/loading/session.py#L191 | 09:36 |
| mkoderer___ | toabctl: https://github.com/openstack/keystoneauth/blob/master/keystoneauth1/loading/conf.py#L97 | 09:37 |
| mkoderer___ | means I can't support the old options like neutron_admin_password | 09:37 |
| *** houming has quit IRC | 09:38 | |
| mkoderer___ | not sure if this is acceptable | 09:38 |
| toabctl | mkoderer___: I try to have a look at 1 pm cest or so. | 09:38 |
| mkoderer___ | toabctl: ok I will process without supporting the old opts | 09:39 |
| mkoderer___ | s/process/proceed/ | 09:39 |
| *** houming has joined #openstack-manila | 09:40 | |
| openstackgerrit | Alexey Ovchinnikov proposed openstack/manila: Fix for share id usage in LXC/LXD driver https://review.openstack.org/291273 | 09:41 |
| vponomaryov | mkoderer___: pong | 09:42 |
| mkoderer___ | vponomaryov: my question.. do I have to support old options like neutron_admin_password for the v3 fix? | 09:43 |
| mkoderer___ | vponomaryov: seems we are missing a way to pass them to keystoneauth lib | 09:43 |
| vponomaryov | mkoderer___: how are you going to use "special" creds for Neutron? | 09:43 |
| mkoderer___ | the new way is [neutron] passwort = foo | 09:44 |
| mkoderer___ | so each client has it's own section | 09:44 |
| vponomaryov | mkoderer___: and manilaclient? | 09:44 |
| mkoderer___ | vponomaryov: mhhh never thought about manilaclient | 09:45 |
| mkoderer___ | but should be the same | 09:45 |
| vponomaryov | it is expected in manila.conf? | 09:45 |
| toabctl | isn't that independent from manilaclient? | 09:45 |
| vponomaryov | section "[neutron] password = foo" | 09:45 |
| mkoderer___ | vponomaryov: so with my adaptation of the devstack plugin you get all that new sections | 09:46 |
| vponomaryov | mkoderer___: new sections where? | 09:46 |
| mkoderer___ | in manila.conf | 09:46 |
| mkoderer___ | vponomaryov: if you have a nova.conf somewhere you see the neutron section | 09:47 |
| vponomaryov | mkoderer___: right | 09:47 |
| mkoderer___ | vponomaryov: that how it look in manila.conf after devstack run: | 09:48 |
| mkoderer___ | http://paste.openstack.org/show/490112/ | 09:48 |
| vponomaryov | mkoderer___: sounds ok | 09:48 |
| mkoderer___ | vponomaryov: ok fine.. I will upload the fix ASAP | 09:49 |
| vponomaryov | mkoderer___: about old options - need to leave them for some time and throw warning not to set up them | 09:49 |
| vponomaryov | mkoderer___: notifying that they will not be used and how people should set up them now | 09:49 |
| mkoderer___ | vponomaryov: ok, will do that | 09:50 |
| vponomaryov | mkoderer___: BUT, removing them from config sample generation, adding new ones | 09:50 |
| mkoderer___ | vponomaryov: +1 | 09:51 |
| vponomaryov | mkoderer___: ok )) | 09:51 |
| vponomaryov | mkoderer___: ty for working on it | 09:51 |
| *** houming has quit IRC | 10:06 | |
| *** lpetrut has joined #openstack-manila | 10:08 | |
| *** tpsilva has joined #openstack-manila | 10:09 | |
| *** houming has joined #openstack-manila | 10:14 | |
| *** yangyapeng has joined #openstack-manila | 10:37 | |
| *** yangyape_ has quit IRC | 10:40 | |
| *** yangyapeng has quit IRC | 10:41 | |
| *** ganso has joined #openstack-manila | 10:50 | |
| -openstackstatus- NOTICE: Gerrit is going to be restarted due to bad performance | 10:53 | |
| *** ChanServ changes topic to "Gerrit is going to be restarted due to bad performance" | 10:53 | |
| tpsilva | ^ good... I thought it was my internet | 10:55 |
| tpsilva | ganso: ^ | 10:56 |
| ganso | great! =D | 10:56 |
| openstackgerrit | Marc Koderer proposed openstack/manila: Fix keystone v3 issues in neutron client https://review.openstack.org/290511 | 10:57 |
| *** ChanServ changes topic to "OpenStack Manila Project" | 10:59 | |
| -openstackstatus- NOTICE: Gerrit has been restarted successfully | 10:59 | |
| *** baojg has quit IRC | 11:04 | |
| openstackgerrit | Marc Koderer proposed openstack/manila: Fix keystone v3 issues in neutron client https://review.openstack.org/290511 | 11:08 |
| *** ociuhandu has quit IRC | 11:11 | |
| openstackgerrit | Marc Koderer proposed openstack/manila: Fix keystone v3 issues in neutron client https://review.openstack.org/290511 | 11:18 |
| *** a_ta has joined #openstack-manila | 11:28 | |
| *** mlima_ has joined #openstack-manila | 11:32 | |
| *** a_ta has quit IRC | 11:32 | |
| *** nkrinner has quit IRC | 11:36 | |
| *** nkrinner has joined #openstack-manila | 11:38 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Fix Share Migration access rule mapping https://review.openstack.org/291671 | 11:38 |
| *** houming has quit IRC | 11:39 | |
| *** houming has joined #openstack-manila | 11:42 | |
| *** houming has quit IRC | 11:42 | |
| *** erlon has joined #openstack-manila | 11:44 | |
| *** _amrith_ is now known as amrith | 11:44 | |
| *** ociuhandu has joined #openstack-manila | 11:48 | |
| openstackgerrit | Tom Patzig proposed openstack/manila: Update quota of proper user on resource delete https://review.openstack.org/289992 | 11:55 |
| *** merooney has joined #openstack-manila | 12:02 | |
| *** merooney has quit IRC | 12:12 | |
| openstackgerrit | Alexey Ovchinnikov proposed openstack/manila: Fix for share id usage in LXC/LXD driver https://review.openstack.org/291273 | 12:13 |
| *** JoseMello has joined #openstack-manila | 12:14 | |
| *** rraja has quit IRC | 12:14 | |
| *** merooney has joined #openstack-manila | 12:17 | |
| *** amrith is now known as _amrith_ | 12:33 | |
| *** yangyapeng has joined #openstack-manila | 12:42 | |
| *** mdnadeem has joined #openstack-manila | 12:46 | |
| *** _amrith_ is now known as amrith | 12:47 | |
| *** mdnadeem_ has quit IRC | 12:48 | |
| *** baojg has joined #openstack-manila | 12:50 | |
| *** kaisers1 has joined #openstack-manila | 12:54 | |
| *** nkrinner has left #openstack-manila | 12:56 | |
| *** merooney has quit IRC | 12:59 | |
| *** amrith is now known as _amrith_ | 13:00 | |
| *** merooney has joined #openstack-manila | 13:05 | |
| *** martyturner has joined #openstack-manila | 13:05 | |
| *** porrua has joined #openstack-manila | 13:07 | |
| *** gouthamr has joined #openstack-manila | 13:13 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Disallow scheduling multiple replicas on a given pool https://review.openstack.org/291287 | 13:15 |
| *** akshai has joined #openstack-manila | 13:19 | |
| mkoderer___ | bswartz: can we merge https://review.openstack.org/#/c/289281/ | 13:23 |
| *** timcl has joined #openstack-manila | 13:26 | |
| gouthamr | vponomaryov, bswartz: could you guys take a look at the tempest tests for share replication: https://review.openstack.org/#/c/244843/ ZFS driver's reported on the latest patch.. | 13:27 |
| vponomaryov | gouthamr: a little bit later | 13:29 |
| *** akshai has quit IRC | 13:29 | |
| gouthamr | vponomaryov: thank you.. Want to get the netapp-ci to start reporting those tests :) | 13:30 |
| *** akshai has joined #openstack-manila | 13:30 | |
| *** merooney has quit IRC | 13:32 | |
| *** merooney has joined #openstack-manila | 13:38 | |
| *** eharney has quit IRC | 13:42 | |
| *** akerr has joined #openstack-manila | 13:44 | |
| *** akerr_ has joined #openstack-manila | 13:46 | |
| *** akerr has quit IRC | 13:49 | |
| *** akshai has quit IRC | 13:51 | |
| openstackgerrit | Merged openstack/manila: Fix unstable DB migration tests https://review.openstack.org/291397 | 13:55 |
| *** timcl has quit IRC | 13:55 | |
| tpatzig | vponomaryov: what do you mean with your comment https://review.openstack.org/#/c/289992/10/manila/tests/api/v1/test_share_networks.py ? | 13:55 |
| *** akshai has joined #openstack-manila | 13:56 | |
| tpatzig | does test_create_valid_cases need to be adjusted? | 13:56 |
| *** timcl has joined #openstack-manila | 13:56 | |
| *** mlima_ has quit IRC | 13:58 | |
| mkoderer___ | gouthamr: hi there, how are you? ;) | 14:06 |
| mkoderer___ | gouthamr: question to your netapp-ci.. if I change the manila.conf structre do you have to adpapt anything? I adapted the devstack pluging accordingly | 14:07 |
| openstackgerrit | Tom Patzig proposed openstack/manila: update quota of origin user on share extend/shrink https://review.openstack.org/291725 | 14:08 |
| openstackgerrit | Marc Koderer proposed openstack/manila: Fix keystone v3 issues in neutron client https://review.openstack.org/290511 | 14:08 |
| vponomaryov | tpatzig: yes, see example: http://paste.openstack.org/show/490164/ | 14:10 |
| *** merooney has quit IRC | 14:10 | |
| *** dustins has joined #openstack-manila | 14:11 | |
| vponomaryov | tpatzig: it is for testing new line 251 from https://review.openstack.org/#/c/289992/11/manila/api/v1/share_networks.py | 14:12 |
| vponomaryov | tpatzig: so, right now, if we have bug on that line, we will not know it | 14:13 |
| vponomaryov | tpatzig: agree? | 14:13 |
| vponomaryov | tpatzig: for example, if by mistake, there was written project_id instead of user_id, quotas will be calculated for nonexistent user with id that is project_id indeed | 14:14 |
| mkoderer___ | vponomaryov: did you run a test coverage tool? ;) | 14:14 |
| vponomaryov | mkoderer___:? | 14:14 |
| *** merooney has joined #openstack-manila | 14:14 | |
| mkoderer___ | vponomaryov: tox -e coverage or something | 14:15 |
| vponomaryov | mkoderer___: in which context is this question? | 14:15 |
| vponomaryov | mkoderer___: if in scope of tpatzig's question, then answer is: "executed code" does not mean "tested code" | 14:15 |
| mkoderer___ | vponomaryov: ok | 14:16 |
| mkoderer___ | vponomaryov: though you asking for more test coverage | 14:16 |
| vponomaryov | mkoderer___: and my example above describes possible problem | 14:16 |
| gouthamr | mkoderer___: do you mean changing the plugin.sh in manila? akerr_ might know. | 14:17 |
| mkoderer___ | gouthamr: I changed some config blocks with https://review.openstack.org/#/c/290511/5 and I am not sure if CI systems have an problem with it | 14:17 |
| *** porrua has quit IRC | 14:18 | |
| *** gouthamr has quit IRC | 14:18 | |
| *** merooney has quit IRC | 14:18 | |
| *** akshai has quit IRC | 14:18 | |
| *** akerr_ has quit IRC | 14:18 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Disallow scheduling multiple replicas on a given pool https://review.openstack.org/291287 | 14:18 |
| tpatzig | vponomaryov: i think i got it. we have to compare if the given user_id is same as share['user_id'] | 14:18 |
| *** akerr has joined #openstack-manila | 14:19 | |
| vponomaryov | tpatzig: not really | 14:19 |
| vponomaryov | tpatzig: you work with "share_network" | 14:19 |
| tpatzig | yes | 14:19 |
| *** gouthamr has joined #openstack-manila | 14:19 | |
| vponomaryov | tpatzig: it has one user_id, but you can create lots of shares by lots of users | 14:20 |
| tpatzig | i mean share_network['user_id'] | 14:20 |
| *** akshai has joined #openstack-manila | 14:20 | |
| akerr | mkoderer___: we try to change as many settings as possible through the local.conf / localrc as possible instead of directly editing manila.conf after the fact | 14:20 |
| akerr | mkoderer: sorry lost connection for a moment if you answered about what you're changing | 14:20 |
| vponomaryov | tpatzig: having all of them be related to single project | 14:20 |
| mkoderer___ | akerr: ok you might have a look at https://review.openstack.org/#/c/290511/5 and -1 if you see an issue | 14:20 |
| *** cknight has joined #openstack-manila | 14:21 | |
| mkoderer___ | akerr: that would be awesome | 14:21 |
| vponomaryov | mkoderer___: I like to -1 things, want me to review this patch? )) | 14:21 |
| *** xyang1 has joined #openstack-manila | 14:21 | |
| mkoderer___ | vponomaryov: as long you don't -2 it! :) | 14:21 |
| *** kaisers1 has quit IRC | 14:22 | |
| *** porrua has joined #openstack-manila | 14:22 | |
| akerr | mkoderer___: I don't see anything that jumps out at me, we do change some neutron settings, but not manila.conf neutron settings | 14:22 |
| akerr | mkoderer___: lets try running our ci against it and see what happens :) | 14:23 |
| mkoderer___ | akerr: yeah sure :) | 14:23 |
| vponomaryov | 5 cents it will work | 14:24 |
| akerr | rubles or dollars? | 14:24 |
| vponomaryov | if it works in general )) | 14:24 |
| gouthamr | vponomaryov: lol. | 14:24 |
| vponomaryov | akerr: my national currency is hryvna | 14:24 |
| akerr | for now... :P | 14:25 |
| *** timcl has quit IRC | 14:25 | |
| bswartz | 5 kopiykas = 5 cents in ukraine I guess | 14:25 |
| vponomaryov | 1 dollar is about 26 hryvnas today, was 8 two years ago | 14:26 |
| vponomaryov | so, 5 kopiykas = almost nothing | 14:26 |
| *** timcl has joined #openstack-manila | 14:31 | |
| *** _amrith_ is now known as amrith | 14:32 | |
| *** cknight has quit IRC | 14:35 | |
| openstackgerrit | Merged openstack/manila: Fix status update for replicas https://review.openstack.org/289634 | 14:35 |
| *** cknight has joined #openstack-manila | 14:36 | |
| *** cknight has quit IRC | 14:41 | |
| *** sgotliv has joined #openstack-manila | 14:43 | |
| tpatzig | vponomaryov: i cannot verify your scenario, if i exchange user_id with project_id on the QUOTA.reserve, the test in L267 will fail | 14:47 |
| tpatzig | https://review.openstack.org/#/c/289992/11/manila/tests/api/v1/test_share_networks.py | 14:47 |
| *** a_ta has joined #openstack-manila | 15:02 | |
| *** aovchinnikov has quit IRC | 15:10 | |
| *** sgotliv has quit IRC | 15:18 | |
| ganso | bswartz: ping | 15:23 |
| bswartz | hey ganso! | 15:30 |
| bswartz | how was your vacation | 15:30 |
| ganso | bswartz: hey! =D | 15:31 |
| ganso | bswartz: it was good! thank you! =) | 15:31 |
| ganso | bswartz: I got a question for you regarding LVM | 15:31 |
| bswartz | ganso: we're drowning in bugs -- https://launchpad.net/manila/+milestone/mitaka-rc1 -- please help us! | 15:31 |
| ganso | bswartz: I am working on it =) submitted 5 fixes already, gotta discuss one in particular with you | 15:32 |
| bswartz | k | 15:32 |
| ganso | bswartz: first, this: Logical volume lvm-shares/share-ac94a3d4-caf2-49f8-a8ef-3e85ce3d900a contains a filesystem in use. | 15:32 |
| ganso | bswartz: I remember seeing a patch that already addressed this, isn't this the problem that is caused by neutron conflict? | 15:33 |
| openstackgerrit | Alex O'Rourke proposed openstack/manila: 3PAR remove file tree on delete when using nested shares https://review.openstack.org/290209 | 15:34 |
| *** vbellur has quit IRC | 15:34 | |
| vponomaryov | tpatzig: ok, lets leave it as is | 15:36 |
| vponomaryov | tpatzig: tests not flexible enough to cover more cases | 15:37 |
| *** Yogi1 has joined #openstack-manila | 15:38 | |
| bswartz | ganso: where did you see that message | 15:38 |
| ganso | bswartz: LVM CI | 15:38 |
| ganso | bswartz: it is still a bit unstable | 15:38 |
| bswartz | have a link? | 15:38 |
| *** vbellur has joined #openstack-manila | 15:39 | |
| ganso | bswartz: https://review.openstack.org/#/c/291366/ | 15:39 |
| bswartz | ganso: yes I'm trying to replace it with the job that doesn't use neutron -- https://review.openstack.org/#/c/291546/ | 15:39 |
| ganso | bswartz: oh ok, I thought it had been addressed in the one that is running neutron | 15:40 |
| bswartz | ganso: we probably could add a workaround to the LVM driver that would make it immune to these neutron issues | 15:41 |
| ganso | bswartz: I remember you talking about domain_names that are used both by LVM and neutron | 15:41 |
| vponomaryov | bswartz: I would say we should do it | 15:41 |
| bswartz | in reality the issue is caused by trusty | 15:41 |
| vponomaryov | bswartz: only in trusty? | 15:41 |
| bswartz | any newer version of Ubuntu doesn't have the mount namespace problems | 15:42 |
| bswartz | yes | 15:42 |
| bswartz | it's a perfect storm of things not working well together | 15:42 |
| bswartz | I can look into adding a workaround for LVM | 15:42 |
| bswartz | the practical problem will go away though when my LVM job change merges | 15:43 |
| ganso | bswartz: from what I have seen LVM driver is almost perfect, just that outstanding issue popping up and making it a bit unstable | 15:43 |
| ganso | bswartz: well, then we should not concentrate effort if it is going to be fixed in the new job | 15:44 |
| vponomaryov | ganso: the same can be said about each driver )) | 15:44 |
| ganso | vponomaryov: yea hehe, LXD and ZFS | 15:44 |
| bswartz | well ZFS is pretty solid right now | 15:44 |
| bswartz | and LXD has a few known issues | 15:44 |
| vponomaryov | ganso: "perfect, but with little nuance - it doesn't work" )) | 15:45 |
| bswartz | LXD concurrency still doesn't work, LXD has version issues, LXD has a serious bug upstream that we have an ugly workaround for | 15:45 |
| vponomaryov | bswartz: "busy" thing appears for ZFS from time to time | 15:45 |
| ganso | well, we are getting there =) | 15:46 |
| bswartz | vponomaryov: when? | 15:46 |
| vponomaryov | bswartz: one sec | 15:46 |
| ganso | we are now much better than we were few weeks ago | 15:46 |
| bswartz | vponomaryov: it should be completely fixed | 15:47 |
| *** vbellur has quit IRC | 15:47 | |
| bswartz | vponomaryov: the only issue I've seen lately with zfs is related to share/unshare commands | 15:47 |
| ganso | bswartz: http://logs.openstack.org/07/289507/3/check/gate-manila-tempest-dsvm-postgres-zfsonlinux-multibackend/427a4ed/logs/screen-m-shr.txt.gz?level=ERROR | 15:47 |
| vponomaryov | bswartz: oh,right, | 15:47 |
| vponomaryov | bswartz: this one appeared on my local lab | 15:48 |
| bswartz | !!!!!!!!!!!! | 15:48 |
| openstack | bswartz: Error: "!!!!!!!!!!!" is not a valid command. | 15:48 |
| ganso | lol ^ | 15:48 |
| * bswartz slaps openstack around a bit with a large trout | 15:48 | |
| vponomaryov | ganso: ty, I remember I saw it )) | 15:49 |
| *** mtanino has joined #openstack-manila | 15:49 | |
| bswartz | this is inconceivable! | 15:49 |
| ganso | vponomaryov: =) | 15:49 |
| *** ayma has joined #openstack-manila | 15:50 | |
| ganso | bswartz: so, there is another controversial fix that I would like your opinion and vponomaryov 's on, this: https://review.openstack.org/#/c/291279/ | 15:51 |
| bswartz | ganso: I'm still trying to understand the above line | 15:52 |
| bswartz | above link | 15:52 |
| bswartz | when did that test run, and what did it run against | 15:52 |
| ganso | bswartz: this was the test https://review.openstack.org/#/c/289507/3 | 15:52 |
| bswartz | the traceback has line numbers that don't correspond to the code in master | 15:52 |
| ganso | bswartz: it is from 2 days ago | 15:52 |
| openstackgerrit | Merged openstack/manila: NetApp cDOT APIs may get too little data https://review.openstack.org/290559 | 15:53 |
| vponomaryov | bswartz: hm, you are right, it has "nsenter" stuff in logs | 15:53 |
| bswartz | ganso: oh, my fix hadn't merged yet | 15:53 |
| bswartz | if you rebase/recheck you should never see that error again | 15:53 |
| ganso | bswartz: does removing nsenter fix this? | 15:54 |
| bswartz | ganso, I implement a different workaround | 15:54 |
| bswartz | in the same patch I removed the nsenter workaround because it created an unwanted dependency | 15:54 |
| bswartz | If anyone finds failed ZFS jobs after my patch merged I want to see them | 15:55 |
| ganso | bswartz: ok | 15:55 |
| * bswartz wishes it were possible to browse test results by job name | 15:55 | |
| ganso | bswartz: sorry about the false alarm | 15:55 |
| vponomaryov | ganso: I guess, "update_access" also fails if share_server is errored | 15:55 |
| vponomaryov | ganso: faced such problem today | 15:55 |
| ganso | vponomaryov: that should be handled by driver implementation, shouldn't it? | 15:56 |
| vponomaryov | ganso: no, manager calls "update_access" and driver fails to reach out share_server that is expected | 15:56 |
| vponomaryov | ganso: because it is errored | 15:56 |
| *** sgotliv has joined #openstack-manila | 15:56 | |
| ganso | vponomaryov: how does the share exist if its share_server is errored? | 15:56 |
| vponomaryov | ganso: DB record exists | 15:57 |
| vponomaryov | ganso: not real resource | 15:57 |
| vponomaryov | ganso: so, right now, using generic driver, it is impossible to delete DB record of a share on failed share_server | 15:57 |
| ganso | vponomaryov: but share creation would fail if share server is errored, then update_access would not succeed, unless it is invoked on a share that is not created | 15:57 |
| ganso | vponomaryov: because of update_access? | 15:58 |
| vponomaryov | ganso: http://paste.openstack.org/show/490187/ | 15:59 |
| vponomaryov | ganso: it is attempt to delete share on errored share_server using generic driver | 15:59 |
| vponomaryov | ganso: where share_server is not available | 15:59 |
| ganso | vponomaryov: that seems easier to address... do not run delete_share if share_server.status == error | 16:00 |
| bswartz | vponomaryov: https://review.openstack.org/291546 | 16:00 |
| bswartz | going to grab lunch bbiab | 16:00 |
| ganso | bswartz: ok | 16:00 |
| *** eharney has joined #openstack-manila | 16:01 | |
| ganso | vponomaryov: is it on CI or lab? is it possible to track if share_server was available and then got into error state before share could be deleted? | 16:02 |
| vponomaryov | ganso: it is on lab, server was never avilable | 16:03 |
| vponomaryov | ganso: statement "do not run delete_share if share_server.status == error" sounds really strange to me | 16:03 |
| vponomaryov | ganso: really wanted to say that you did say? | 16:04 |
| akerr | mkoderer___: seems to work fine for us: https://review.openstack.org/#/c/290511/5 | 16:05 |
| vponomaryov | ganso: it means "share DB record will stay forever" if share server failed to be created | 16:06 |
| ganso | vponomaryov: if user tried to create a share, manager tried to create a share_server but failed, share_server remains in error, so share remains in error as well | 16:07 |
| vponomaryov | ganso: yes | 16:07 |
| ganso | vponomaryov: so, user will invoke delete_share... then manager code's delete_share will invoke update_access before driver's delete_share | 16:07 |
| vponomaryov | ganso: and ti si incorrect | 16:07 |
| ganso | vponomaryov: this manager code's delete_share should not do any of these if share_server status is error | 16:08 |
| vponomaryov | ganso: how are you going to apply update of access rules of failed server? | 16:08 |
| ganso | vponomaryov: skip it | 16:08 |
| vponomaryov | ganso: right | 16:08 |
| vponomaryov | ganso: so, you confirm, that it is bug? | 16:08 |
| ganso | vponomaryov: yes, user cannot remove share from DB | 16:09 |
| ganso | vponomaryov: it is weird that this is a new bug, I was able to remove the share before when server failed to be created | 16:09 |
| vponomaryov | ganso: right, it is exactly new bug | 16:09 |
| vponomaryov | ganso: and, I suspect, it is related to all DHSS=True drivers | 16:10 |
| ganso | vponomaryov: so, maybe we should discuss https://review.openstack.org/#/c/291279/ | 16:10 |
| ganso | vponomaryov: because an alternate solution (not the one I am proposing) would fix that | 16:11 |
| vponomaryov | ganso: need to look closely at it | 16:11 |
| *** a_ta has quit IRC | 16:13 | |
| vponomaryov | ganso: this part - https://review.openstack.org/#/c/291279/2/manila/share/access.py not covered with unit tests | 16:13 |
| vponomaryov | ganso: also, you sure, that there will not be "ShareNotFound" exception in addition to "ShareResourceNotFound"? | 16:14 |
| ganso | vponomaryov: my unit test coverage reports it is. This test: test_delete_share_instance_not_found addresses that because it mocks driver's update_access to throw exception | 16:15 |
| *** vbellur has joined #openstack-manila | 16:15 | |
| vponomaryov | ganso: ok | 16:16 |
| vponomaryov | ganso: you say it is "partial" bug | 16:16 |
| vponomaryov | ganso: because of driver's part? | 16:17 |
| ganso | vponomaryov: ShareNotFound is not related to ShareResourceNotFound, one is not subclass of the other | 16:17 |
| ganso | vponomaryov: yes, this is the controversial part | 16:17 |
| ganso | vponomaryov: I would like other opinion on whether drivers should acknowledge to this, or we take a different approach | 16:17 |
| ganso | vponomaryov: there is driver effort involved | 16:17 |
| vponomaryov | ganso: saying about "ShareNotFound" I do not mean inheritance, I mean exactly raise of different exception | 16:17 |
| ganso | vponomaryov: guideline should say explicitely that ShareResourceNotFound should be raised | 16:18 |
| vponomaryov | ganso: I am ok having it as is | 16:18 |
| vponomaryov | ganso: two parts | 16:18 |
| *** vbellur has quit IRC | 16:19 | |
| vponomaryov | ganso: is it correct to say, that now, we should update all driver to raise "ShareResourceNotFound" in "update_access"? | 16:21 |
| vponomaryov | s/driver/drivers/ | 16:21 |
| *** cknight has joined #openstack-manila | 16:21 | |
| vponomaryov | ganso: as it is done in https://review.openstack.org/#/c/291358/1/manila/share/drivers/hitachi/hds_hnas.py | 16:21 |
| ganso | vponomaryov: if driver vendor cares to fix this problem, yes. Not sure if "we" should do it | 16:21 |
| ganso | vponomaryov: regarding being a "partial-bug", I am not sure if I consider it closed only with the first one implemented and no driver implementation | 16:22 |
| vponomaryov | ganso: bug is not addressed with only first part | 16:22 |
| vponomaryov | ganso; that why I ask | 16:22 |
| vponomaryov | ganso: all other drivers will still fail | 16:23 |
| ganso | vponomaryov: yes, thus "partial-bug" | 16:23 |
| ganso | vponomaryov: and not sure if this approach can be considered valid, it does not fix anything by itself... not sure if this is the right way, maybe we can think of something else | 16:24 |
| ganso | vponomaryov: what I do not want to do is have drivers saying they applied rules successfully when in fact should be throwing an exception because share does not exist | 16:24 |
| *** absubram has joined #openstack-manila | 16:24 | |
| ganso | vponomaryov: so, most drivers have this bug. NetApp fixed this by swallowing the exception and saying it applied successfully. Other drivers that have this bug should take one approach to fix it. If we merge this one, they have this way of fixing it | 16:26 |
| *** absubram_ has joined #openstack-manila | 16:26 | |
| vponomaryov | ganso: I think, we should make our "force-delete" API really delete things anyway | 16:27 |
| vponomaryov | ganso: even this one fails to do it | 16:27 |
| vponomaryov | ganso: so, we can say that second part of fix would be possibility to just delete DB record anyway | 16:28 |
| vponomaryov | ganso: and allow drivers to do it in the way to do | 16:28 |
| vponomaryov | s/to do/you do/ | 16:28 |
| ganso | vponomaryov: that's a good idea, it will unblock everyone having that problem | 16:29 |
| *** absubram has quit IRC | 16:29 | |
| *** absubram_ is now known as absubram | 16:29 | |
| vponomaryov | ganso: it also will address mentioned bug with share servers | 16:29 |
| ganso | vponomaryov: yes. Thanks for your input, I will improve the change with your suggestion. =) | 16:30 |
| ganso | vponomaryov: seems the right way to go | 16:30 |
| vponomaryov | ganso: ok | 16:30 |
| ganso | vponomaryov: s/seems/seems to be | 16:30 |
| *** sgotliv has quit IRC | 16:32 | |
| openstackgerrit | Alex Meade proposed openstack/manila: NetApp cDOT: Fix status updates for replicas https://review.openstack.org/291803 | 16:33 |
| ganso | vponomaryov: do you have bug filed/ | 16:33 |
| ganso | vponomaryov: s/'/'/? | 16:34 |
| vponomaryov | ganso: no | 16:34 |
| vponomaryov | ganso: was busy distracting to lots of things )) | 16:34 |
| *** cknight has quit IRC | 16:39 | |
| *** cknight has joined #openstack-manila | 16:40 | |
| openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Fix HNAS snapshot creation on deleted shares https://review.openstack.org/291806 | 16:42 |
| *** cknight has quit IRC | 16:43 | |
| *** a_ta has joined #openstack-manila | 16:46 | |
| *** pcaruana has quit IRC | 16:50 | |
| * bswartz is back | 16:54 | |
| *** absubram has quit IRC | 16:55 | |
| *** absubram has joined #openstack-manila | 16:55 | |
| *** zhonghua has quit IRC | 16:55 | |
| tpsilva | are all the bugs' assignees actually working on the bugs? | 16:56 |
| bswartz | tpsilva: ni | 16:56 |
| bswartz | no | 16:56 |
| *** shausy has joined #openstack-manila | 16:56 | |
| *** zhonghua-lee has joined #openstack-manila | 16:56 | |
| bswartz | for some older bugs there are patchsets with -1 and no activity for months | 16:56 |
| tpsilva | bswartz: nice... I'll hunt for some of these and try to help | 16:57 |
| bswartz | others have no patchset at all | 16:57 |
| vponomaryov | ganso: what happend to your +2 button? )) | 16:58 |
| vponomaryov | ganso; you broke it and use old +1? )) | 16:58 |
| bswartz | vponomaryov: thanks for +1 but we still need to talk about check jobs | 17:02 |
| bswartz | vponomaryov: why do we have vendor CI systems and all the different drivers run API tests? many API tests are able to pass with completely broken drivers | 17:02 |
| bswartz | seems like the set of tests we run on each driver should be limited to tests that actually exercise the drivers | 17:03 |
| vponomaryov | bswartz: tests that work with broken driver are fast | 17:03 |
| vponomaryov | bswartz: and it is good to have statistics when we break something | 17:03 |
| vponomaryov | bswartz: not related to driver | 17:04 |
| bswartz | you think that if we removed the non-driver-exercising tests that the job times would be the same? | 17:04 |
| bswartz | like quota tests, for example | 17:04 |
| *** shausy has quit IRC | 17:04 | |
| vponomaryov | bswartz: it will be less than common time deviation of test runs | 17:05 |
| vponomaryov | bswartz: making jobs not install redundant services - this is the thing we shoudl do | 17:05 |
| bswartz | the deviations between test run times has more to do with which cloud provider got the job than actual randomness | 17:05 |
| bswartz | some nodes are very fast and other nodes are very slow | 17:06 |
| bswartz | if we can decrease the test duration on a consistent configuration though then we should do it | 17:06 |
| bswartz | also, I think our test timeouts cause problems for extremely slow nodes | 17:07 |
| bswartz | on the fast nodes our timeout should be like 50 minutes and on the slow nodes our test timeout should be like 150 minutes | 17:07 |
| vponomaryov | bswartz: for generic driver only | 17:08 |
| *** absubram has quit IRC | 17:08 | |
| vponomaryov | bswartz: and maybe windows | 17:08 |
| *** baojg has quit IRC | 17:09 | |
| *** a_ta has quit IRC | 17:14 | |
| bswartz | vponomaryov: I have seen LVM jobs take more than an hour | 17:17 |
| vponomaryov | bswartz: it is while we have redundant services installed | 17:17 |
| vponomaryov | bswartz: current minimal-lvm takes 15 minutes | 17:18 |
| bswartz | vponomaryov: is it time to move LXD and ZFS to "minimal" job configurations? | 17:18 |
| vponomaryov | bswartz: I think all drivers, why those two? | 17:18 |
| bswartz | well LXD will still require neutron | 17:18 |
| bswartz | generic will require neutron+cinder+nova+glance | 17:19 |
| bswartz | so there is less win in those cases | 17:19 |
| bswartz | scenario job will require neutron+cinder+nova+glance | 17:19 |
| vponomaryov | then template should allow us to set those for each separate job as we need | 17:19 |
| bswartz | vponomaryov: you think that's possible? | 17:20 |
| bswartz | the template will get pretty complex | 17:20 |
| vponomaryov | bswartz: kind of complex, yes | 17:20 |
| openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Fix HNAS snapshot creation on deleted shares https://review.openstack.org/291806 | 17:20 |
| vponomaryov | bswartz: but not complex enough to avoid such approach | 17:20 |
| bswartz | let's not spend time on that until after RC1 | 17:21 |
| bswartz | anyone know IRC handle for Alex O'Rourke? | 17:22 |
| bswartz | oh duh | 17:22 |
| bswartz | aorourke: ping | 17:22 |
| aorourke | bswartz, hey | 17:22 |
| bswartz | https://review.openstack.org/#/c/290209/ <-- this add a bunch of config options | 17:22 |
| bswartz | typically that's a not allowed after FF | 17:23 |
| bswartz | can you explain why you need 4 new config options to fix a bug? | 17:23 |
| openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Remove default values for update_access() https://review.openstack.org/291153 | 17:24 |
| aorourke | bswartz, in order to delete the nested share for CIFS (this is not an issue with NFS), I have to mount the share. And I need the CIFS admin username, password, and domain to do the mount | 17:24 |
| aorourke | bswartz, the mount path was just to allow the option to decide where to mount the share. If you did not want to use /mnt/ | 17:25 |
| bswartz | but these options weren't needed anywhere else in the driver? | 17:25 |
| bswartz | I guess I'm concerned because this seems like another case of putting m-shr in the data path | 17:25 |
| *** dustins_ has joined #openstack-manila | 17:26 | |
| aorourke | bswartz, they were not needed before, no. Only for doing the mount required to delete the nested shares that hang around | 17:26 |
| bswartz | I don't have an issue with allowing drivers to delete files in this way, but it feels like it should be done through a common service, not inside the driver | 17:26 |
| aorourke | is there currently a common service that this can be done with? | 17:27 |
| bswartz | I guess I wonder if this change should wait until newton | 17:27 |
| bswartz | no there isn't yet | 17:27 |
| bswartz | what is the damage if we leave this bug unfixed in mitaka? | 17:28 |
| aorourke | bswartz, when a nested share is deleted (which nested shares are used by default), the entire content of that share still exists. | 17:28 |
| bswartz | so space leaks? | 17:28 |
| aorourke | bswartz, basically. The data is not being wiped | 17:29 |
| bswartz | are nested shares necessary? can a workaround be to not use nested shares for CIFS in mitaka? | 17:29 |
| *** dustins has quit IRC | 17:29 | |
| aorourke | this bug fixes that and allows us to delete the content in the share | 17:29 |
| aorourke | we have limitations without nested shares on the 3PAR. the amount we can create is very very limited | 17:30 |
| bswartz | right but the current stable driver suffers from the same limitation doesn't it? | 17:31 |
| bswartz | for reference, a similar issue came up with glusterfs driver in liberty and we made them find a way to fix the bug without adding new config options | 17:31 |
| bswartz | I'm trying to be fair to everyone about preventing stealth features-as-bugs after feature freeze | 17:31 |
| aorourke | the current stables driver will use the nested shares by default, so these 3PAR limitations are not really an issue. The issue we do have is the shares contents not being deleted when manila delete is called | 17:32 |
| aorourke | i understand your trying to be fair. I do think it is a bug that when delete is called, the shares contents are not actually deleted. they still exist on the backend | 17:32 |
| bswartz | I don't disagree that a bug exists | 17:32 |
| bswartz | but we tend to classify new config options as new features | 17:33 |
| bswartz | and this fix is a great illustration of the tension that causes | 17:33 |
| aorourke | bswartz, I don't have a way to mount the CIFS share without having the that info, though | 17:34 |
| *** krotscheck has quit IRC | 17:34 | |
| bswartz | well for future reference, it's better to get this kind of change in before feature freeze | 17:35 |
| bswartz | I'll need to think about what to do there | 17:35 |
| *** pcaruana has joined #openstack-manila | 17:35 | |
| aorourke | bswartz, understandable. thank you | 17:36 |
| *** krotscheck has joined #openstack-manila | 17:37 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Disallow scheduling multiple replicas on a given pool https://review.openstack.org/291287 | 17:56 |
| openstackgerrit | Petrut Lucian proposed openstack/manila: Windows driver: fix share access actions https://review.openstack.org/291837 | 17:56 |
| *** absubram has joined #openstack-manila | 18:00 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Fix handling of share server details after error https://review.openstack.org/291842 | 18:01 |
| *** a_ta has joined #openstack-manila | 18:07 | |
| openstackgerrit | Ben Swartzlander proposed openstack/manila: Collapse common os_region_name option https://review.openstack.org/291844 | 18:07 |
| *** Michele_ has joined #openstack-manila | 18:07 | |
| *** Michele_ has left #openstack-manila | 18:08 | |
| *** lpetrut has quit IRC | 18:18 | |
| *** netapp has quit IRC | 18:18 | |
| *** netapp has joined #openstack-manila | 18:19 | |
| *** jasonsb has quit IRC | 18:21 | |
| *** a_ta has quit IRC | 18:22 | |
| openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Fix HNAS snapshot creation on deleted shares https://review.openstack.org/291806 | 18:23 |
| *** sgotliv has joined #openstack-manila | 18:23 | |
| *** resker has quit IRC | 18:24 | |
| *** mlima has joined #openstack-manila | 18:39 | |
| openstackgerrit | Merged openstack/manila: Fix compatibility with Tempest https://review.openstack.org/290698 | 18:40 |
| openstackgerrit | Merged openstack/manila: Set TCP keepalive options https://review.openstack.org/289296 | 18:43 |
| *** sgotliv has quit IRC | 18:44 | |
| *** merooney has joined #openstack-manila | 18:46 | |
| *** lpetrut has joined #openstack-manila | 18:49 | |
| openstackgerrit | Merged openstack/manila: Fix manila devstack plugin for keystone v3 usage https://review.openstack.org/289281 | 18:53 |
| *** adrianofr_ has quit IRC | 19:00 | |
| *** dencaval has quit IRC | 19:02 | |
| *** testestsetse has joined #openstack-manila | 19:02 | |
| *** testestsetse has quit IRC | 19:02 | |
| *** testestsetse has joined #openstack-manila | 19:02 | |
| *** lpetrut has quit IRC | 19:07 | |
| *** eharney has quit IRC | 19:09 | |
| *** martyturner has quit IRC | 19:13 | |
| *** merooney has quit IRC | 19:14 | |
| *** akerr has quit IRC | 19:17 | |
| *** akerr_ has joined #openstack-manila | 19:17 | |
| *** mlima_ has joined #openstack-manila | 19:20 | |
| *** akerr has joined #openstack-manila | 19:21 | |
| *** lpetrut has joined #openstack-manila | 19:21 | |
| *** akerr_ has quit IRC | 19:21 | |
| *** akerr_ has joined #openstack-manila | 19:23 | |
| *** mlima has quit IRC | 19:24 | |
| *** akerr has quit IRC | 19:24 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Fix delete when share not found in update_access https://review.openstack.org/291279 | 19:25 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: HDS-HNAS: Fix exception in update_access not found https://review.openstack.org/291358 | 19:25 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: HDS-HNAS: Fix exception in update_access not found https://review.openstack.org/291358 | 19:28 |
| *** ociuhandu has quit IRC | 19:36 | |
| *** amrith is now known as _amrith_ | 19:38 | |
| *** fthiagogv has joined #openstack-manila | 19:47 | |
| openstackgerrit | Petrut Lucian proposed openstack/manila: Windows driver: fix share access actions https://review.openstack.org/291837 | 19:49 |
| *** akerr has joined #openstack-manila | 19:57 | |
| *** akerr_ has quit IRC | 19:58 | |
| *** pgbridge has quit IRC | 19:58 | |
| *** Yogi1 has left #openstack-manila | 20:02 | |
| *** Yogi1 has joined #openstack-manila | 20:02 | |
| *** testestsetse has quit IRC | 20:03 | |
| *** akerr has quit IRC | 20:07 | |
| *** akerr has joined #openstack-manila | 20:07 | |
| *** a_ta has joined #openstack-manila | 20:13 | |
| *** akerr_ has joined #openstack-manila | 20:18 | |
| *** akerr has quit IRC | 20:19 | |
| *** ociuhandu has joined #openstack-manila | 20:19 | |
| *** dustins_ has quit IRC | 20:23 | |
| *** eharney has joined #openstack-manila | 20:26 | |
| *** a_ta has quit IRC | 20:38 | |
| *** vbellur has joined #openstack-manila | 20:42 | |
| *** a_ta has joined #openstack-manila | 20:43 | |
| *** mlima_ has quit IRC | 20:45 | |
| *** timcl has quit IRC | 20:54 | |
| openstackgerrit | Doug Hellmann proposed openstack/manila: register the config generator default hook with the right name https://review.openstack.org/291912 | 21:08 |
| openstackgerrit | Scott DAngelo proposed openstack/manila: Allow api_version_request.matches to accept a string or None https://review.openstack.org/285426 | 21:14 |
| bswartz | tbarron: ping | 21:14 |
| tbarron | bswartz: pong | 21:15 |
| bswartz | tbarron: may we abandon https://review.openstack.org/#/c/288613/ assuming we're going with the other patch? | 21:15 |
| tbarron | bswartz: sure, there are two of them - one for the config gen, one for the unit tests. | 21:16 |
| tbarron | bswartz: I'll abandon them right now. | 21:16 |
| bswartz | two bugs you mean | 21:17 |
| bswartz | one change fixes both tough | 21:17 |
| bswartz | though | 21:17 |
| bswartz | I'm testing your fix now | 21:17 |
| tbarron | bswartz: I also had two alternative patches originally. | 21:17 |
| bswartz | tbarron: I'm a bit puzzled -- why do unit tests work fine for me on F22? | 21:18 |
| bswartz | which platform exactly couldn't handle the pylxd dependency | 21:18 |
| tbarron | bswartz: unit tests work fine as long as you are using a virtualenv | 21:18 |
| tbarron | bswartz: remove the requirement and rebuild your virtualenv | 21:19 |
| bswartz | who runs unit tests without a virtualenv? | 21:19 |
| tbarron | bswartz: we do in CI | 21:19 |
| bswartz | interesting | 21:19 |
| bswartz | on what platform | 21:19 |
| tbarron | bswartz: and customers may if we ship | 21:19 |
| tbarron | unit tests, which is being discussed | 21:19 |
| bswartz | customers running unit tests? that sounds completely insane to tme | 21:20 |
| tbarron | and RDO is very likely to start shipping unit tests | 21:20 |
| bswartz | to me | 21:20 |
| bswartz | what purpose could that possibly serve? | 21:20 |
| tbarron | some RDO folks want it | 21:20 |
| bswartz | oh | 21:20 |
| tbarron | and certification distro may need it | 21:20 |
| bswartz | developers are your "customers" | 21:20 |
| tbarron | even if mainline rhos doesn't | 21:20 |
| tbarron | it's not 100% that unit tests will ship | 21:20 |
| bswartz | it makes sense | 21:21 |
| bswartz | but I'm still not clear which platform barfs | 21:21 |
| tbarron | but it is 100% that we need both to build and to test with only packages, not pip droppinigs | 21:21 |
| bswartz | well fedora barf without a virtualenv? | 21:21 |
| bswartz | will* | 21:21 |
| tbarron | bswartz: I only tested f23 | 21:21 |
| tbarron | but internally someone built for rdo and it broke | 21:21 |
| tbarron | I don't know for sure what platform they used | 21:22 |
| tbarron | but it doesn't maytter: it didn't have pylxd and none of rhel, fedora, or centos do | 21:22 |
| tbarron | as they ship | 21:22 |
| tbarron | at any version | 21:22 |
| bswartz | right | 21:22 |
| bswartz | so what happens in a venv would couldn't also be replicated outside the venv | 21:22 |
| bswartz | sudo pip install pylxd fails? | 21:23 |
| tbarron | so if you are on f22 and remove the requirement and re-build your virtualenv unit tests break w/o the patch and work with the patch | 21:23 |
| tbarron | at least with patch set 1 | 21:23 |
| tbarron | honestly, i didn't reteset with PS2 | 21:23 |
| bswartz | well we're merging patchset 3 | 21:24 |
| bswartz | of 288800 | 21:24 |
| tbarron | no, if you *don't* pip install, then you try to run the unit tests and the pylxd driver unit tests fail | 21:24 |
| tbarron | i meant ps3, latest | 21:24 |
| bswartz | so I think it's safe to abandon 288613 | 21:24 |
| bswartz | the whole point of requirements though is that you're supposed to install them | 21:24 |
| tbarron | removing the requirement just simulates the packaged situation | 21:25 |
| bswartz | I must be missing something subtle about what redhat does different from the rest of us | 21:25 |
| *** akerr_ has quit IRC | 21:26 | |
| tbarron | unit tests work fine for redhat developers doing regular tox or run_tests.sh | 21:26 |
| bswartz | if redhat doesn't bother to install packages in requirements.txt, then I'm surprised that anything can work | 21:26 |
| tbarron | we install the packages in requirements.txt | 21:26 |
| tbarron | but we patch-out pylxd | 21:26 |
| *** akerr has joined #openstack-manila | 21:26 | |
| bswartz | okay that's what I'm missing | 21:26 |
| tbarron | apevec would have the whole story on that | 21:26 |
| tbarron | so in my env I simulate that by removing that line from requirements.txt | 21:27 |
| bswartz | you've intentionally taken it out -- for presumably nontechnical reasons -- because I just did sudo pip install pylxd and it works fine | 21:27 |
| tbarron | and see that even with it removed and the patch it works | 21:27 |
| tbarron | yes | 21:27 |
| tbarron | it will install fine | 21:27 |
| tbarron | pip works on rhel, fedora, etc. | 21:28 |
| tbarron | but we don't distribute using pip, we use rpms | 21:28 |
| tbarron | and we don't include pylxd rpm | 21:28 |
| bswartz | so in the future, what is the correct way to add a requirement such that it won't blow up on redhat? | 21:28 |
| tbarron | that breaks the build because of opts.py | 21:28 |
| tbarron | and it breaks our CI running unit tests | 21:28 |
| tbarron | bswartz: I don't know, but perhaps toabctl's suggestion in the review would work | 21:29 |
| tbarron | (looking) | 21:29 |
| bswartz | I just know we'll end up at this place again | 21:30 |
| tbarron | See http://docs.openstack.org/developer/pbr/#extra-requirements | 21:30 |
| tbarron | What about moving the pylxd from requirements.txt to setup.cfg in the [extra] section? | 21:30 |
| tbarron | ^^^ quoting toabctl ^^^ | 21:30 |
| bswartz | I don't see how that helps you | 21:30 |
| openstackgerrit | Merged openstack/manila: Disallow scheduling multiple replicas on a given pool https://review.openstack.org/291287 | 21:30 |
| bswartz | it gets you out of having to patch requirements.txt downstream but you still blow up if the driver had a dependency | 21:30 |
| tbarron | well, we may want to talk to infra about the right way to do soft dependencies. | 21:31 |
| bswartz | soft dependencies are one thing, but I'm talking about future hard dependencies | 21:31 |
| tbarron | or manila may be pioneering here | 21:31 |
| bswartz | before I had assumed that all of pypi.python.org was fair game | 21:32 |
| bswartz | clearly that's not the case | 21:32 |
| bswartz | so now how do we determine what's safe to depend on and what isn't? | 21:32 |
| tbarron | well, I think talking about it will likely avoid this going forwards. Honestly, I didn't know to think this might be an issue or | 21:33 |
| tbarron | I would have brought it up much earlier in the development/design cycle. | 21:33 |
| tbarron | I feel bad about bringing it up late. | 21:33 |
| bswartz | is it as simple as asking redhat, suse, canonical, and debian to review every new requirement? | 21:34 |
| bswartz | that's a very heavy requirement but if there's no alternative then we need to make sure everyone understands | 21:34 |
| tbarron | well, I don't know that it has to be formal, and I think good judgment will help determine the probability that there could be a problem. | 21:35 |
| tbarron | most pypi packages are unproblematic. | 21:35 |
| tbarron | s/modules/packages/ | 21:36 |
| tbarron | I'm going away for a couple minutes to clean up those other reviews. | 21:36 |
| bswartz | k | 21:36 |
| tbarron | 288613 and 288721 have been abandoned | 21:38 |
| *** porrua has quit IRC | 21:39 | |
| tbarron | bswartz: going forwards, i will commit to keeping an eye out and raising my hand as early as reasonable if I see a dependency that seems problematic. | 21:42 |
| bswartz | tbarron: I'm not blaming you for causing this problem | 21:42 |
| bswartz | tbarron: and I don't see why you should be responsible for solving it single handedly | 21:42 |
| bswartz | there must be an answer that works for the whole community | 21:42 |
| tbarron | on this one, when I became aware, I checked and we were the only non-ubuntu distro trying to build and test mitaka with rpms/debs yet. | 21:43 |
| tbarron | actually, i guess zigo contradicts me | 21:43 |
| tbarron | b/c I didn't check with him | 21:43 |
| tbarron | but I did talk to SUSE and HELION folks | 21:43 |
| zigo | tbarron: o/ | 21:43 |
| zigo | May I help? | 21:43 |
| tbarron | zigo: of course! | 21:43 |
| zigo | tbarron: How? | 21:44 |
| tbarron | so in the future, collecting data on this stuff, we'll talk to zigo and treat pure debian as a distro | 21:44 |
| zigo | (sorry, the backlog is huge...) | 21:44 |
| bswartz | pure debian is a distro as far as the install guide folks are concerned | 21:44 |
| tbarron | zigo: nothing immediate. I'm just saying that we'll all communicate about dependencies like pylxd, which are problematic for some distros and not others | 21:45 |
| zigo | Well, the install-guide folks aren't being super helpful with the Debian stuff. | 21:45 |
| zigo | Most of it, it's actually *me* who contributed it. | 21:45 |
| zigo | And they currently refuse to publish it like the other guides. | 21:45 |
| bswartz | sorry to hear that zigo | 21:45 |
| tbarron | zigo: :-( | 21:45 |
| bswartz | the install guide is a bit understaffed, as we have discovered | 21:45 |
| zigo | Mostly, they don't like the fact that I'm using debconf and I do things a bit differently. | 21:45 |
| zigo | that's correct | 21:46 |
| zigo | Though they should still accept my contribution and publish it, IMO, on a best effort basis. | 21:46 |
| zigo | Having something is better than nothing. | 21:46 |
| tbarron | zigo: the immediate issue here/atm is that one can't assume 100% that because a package is in pypi that it will be unproblematic to package it as a deb or rpm and ship it | 21:46 |
| zigo | Did they refused to have Manila in the install guide ? | 21:46 |
| zigo | I've read such a thread. | 21:47 |
| tbarron | zigo: i don't know about that one way or the other | 21:48 |
| *** cknight has joined #openstack-manila | 21:50 | |
| *** tpsilva has quit IRC | 21:57 | |
| *** scottda is now known as scottda_on_vacat | 21:59 | |
| *** scottda_on_vacat is now known as scottda_vacation | 22:00 | |
| *** cknight has quit IRC | 22:01 | |
| *** alyson_ has quit IRC | 22:02 | |
| *** fthiagogv has quit IRC | 22:08 | |
| openstackgerrit | Merged openstack/manila: Fix pylxd hard dependencies https://review.openstack.org/288800 | 22:09 |
| *** merooney has joined #openstack-manila | 22:29 | |
| *** merooney has quit IRC | 22:29 | |
| *** a_ta has quit IRC | 22:32 | |
| *** esker has joined #openstack-manila | 22:32 | |
| *** david-lyle has quit IRC | 22:37 | |
| *** sgotliv has joined #openstack-manila | 22:44 | |
| *** JoseMello has quit IRC | 22:48 | |
| *** sgotliv has quit IRC | 22:51 | |
| *** ayma has quit IRC | 22:52 | |
| *** xyang1 has quit IRC | 23:01 | |
| *** lpetrut has quit IRC | 23:03 | |
| openstackgerrit | Yogesh proposed openstack/manila: Add tempest tests for Share Replication https://review.openstack.org/244843 | 23:05 |
| *** gouthamr has quit IRC | 23:05 | |
| *** Yogi1 has quit IRC | 23:07 | |
| *** akshai has quit IRC | 23:16 | |
| *** ganso has quit IRC | 23:19 | |
| *** sgotliv has joined #openstack-manila | 23:37 | |
| *** akerr is now known as akerr_away | 23:40 | |
| *** erlon has quit IRC | 23:48 | |
| *** yangyapeng has quit IRC | 23:54 | |
| *** yangyapeng has joined #openstack-manila | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!