| smcginnis_2 | Coming in for a landing. Cya later. | 00:01 |
|---|---|---|
| *** smcginnis_2 has quit IRC | 00:01 | |
| *** akshai has quit IRC | 00:03 | |
| *** zhangjn has joined #openstack-cinder | 00:04 | |
| *** zhangjn has quit IRC | 00:04 | |
| *** alonma has quit IRC | 00:04 | |
| *** zhangjn has joined #openstack-cinder | 00:04 | |
| *** [1]Thelo has joined #openstack-cinder | 00:04 | |
| *** yhayashi has quit IRC | 00:05 | |
| *** mudassirlatif has quit IRC | 00:05 | |
| *** alonma has joined #openstack-cinder | 00:05 | |
| *** Thelo has quit IRC | 00:07 | |
| *** [1]Thelo is now known as Thelo | 00:07 | |
| *** yhayashi has joined #openstack-cinder | 00:07 | |
| *** mudassirlatif has joined #openstack-cinder | 00:10 | |
| *** alonma has quit IRC | 00:10 | |
| *** alonma has joined #openstack-cinder | 00:11 | |
| *** mudassirlatif has quit IRC | 00:12 | |
| *** diogogmt has joined #openstack-cinder | 00:14 | |
| *** alonma has quit IRC | 00:16 | |
| *** zhangjn has quit IRC | 00:16 | |
| *** zhangjn has joined #openstack-cinder | 00:17 | |
| *** alonma has joined #openstack-cinder | 00:18 | |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/cinder: CONF add suppress_request_ssl_warnings https://review.openstack.org/267870 | 00:18 |
| hemna | I'm sure that'll be controversial | 00:18 |
| *** ociuhandu has quit IRC | 00:21 | |
| *** mudassirlatif has joined #openstack-cinder | 00:21 | |
| *** alonma has quit IRC | 00:24 | |
| *** zhangjn has quit IRC | 00:25 | |
| *** alonma has joined #openstack-cinder | 00:26 | |
| *** diablo_rojo has joined #openstack-cinder | 00:26 | |
| *** apoorvad has joined #openstack-cinder | 00:27 | |
| *** alonma has quit IRC | 00:30 | |
| *** alonma has joined #openstack-cinder | 00:32 | |
| *** apoorvad has quit IRC | 00:33 | |
| *** hemna is now known as hemnafk | 00:35 | |
| *** martyturner has quit IRC | 00:35 | |
| *** alonma has quit IRC | 00:36 | |
| *** alonma has joined #openstack-cinder | 00:38 | |
| *** mudassirlatif has quit IRC | 00:40 | |
| *** alonma has quit IRC | 00:43 | |
| *** alonma has joined #openstack-cinder | 00:44 | |
| *** alonma has quit IRC | 00:49 | |
| *** merooney has joined #openstack-cinder | 00:49 | |
| *** mriedem has joined #openstack-cinder | 00:50 | |
| *** apoorvad has joined #openstack-cinder | 00:52 | |
| *** sombrafam has joined #openstack-cinder | 00:52 | |
| *** zhangjn has joined #openstack-cinder | 00:53 | |
| *** leeantho has quit IRC | 00:58 | |
| *** zhangjn has quit IRC | 00:58 | |
| *** cheneydc has joined #openstack-cinder | 01:00 | |
| *** zhangjn has joined #openstack-cinder | 01:02 | |
| *** crose has joined #openstack-cinder | 01:05 | |
| *** rajinir has quit IRC | 01:05 | |
| *** diogogmt has quit IRC | 01:05 | |
| *** garthb__ has quit IRC | 01:06 | |
| *** mudassirlatif has joined #openstack-cinder | 01:06 | |
| *** mudassirlatif has quit IRC | 01:07 | |
| *** EinstCrazy has joined #openstack-cinder | 01:07 | |
| *** bardia has quit IRC | 01:09 | |
| *** jacky-zhang has joined #openstack-cinder | 01:10 | |
| *** zhangjn has quit IRC | 01:14 | |
| *** gouthamr has quit IRC | 01:15 | |
| *** mudassirlatif has joined #openstack-cinder | 01:24 | |
| *** cknight1 has joined #openstack-cinder | 01:28 | |
| *** Lee1092 has joined #openstack-cinder | 01:29 | |
| *** davechen has joined #openstack-cinder | 01:30 | |
| *** cknight has quit IRC | 01:30 | |
| *** leeantho has joined #openstack-cinder | 01:34 | |
| *** cheneydc has quit IRC | 01:36 | |
| *** zhangjn has joined #openstack-cinder | 01:36 | |
| *** gouthamr has joined #openstack-cinder | 01:37 | |
| *** haomaiwang has joined #openstack-cinder | 01:37 | |
| *** cheneydc has joined #openstack-cinder | 01:38 | |
| *** alonma has joined #openstack-cinder | 01:38 | |
| *** leeantho has quit IRC | 01:40 | |
| *** alonma has quit IRC | 01:43 | |
| *** zhangjn has quit IRC | 01:43 | |
| *** alonma has joined #openstack-cinder | 01:44 | |
| *** zhangjn has joined #openstack-cinder | 01:46 | |
| *** alonma has quit IRC | 01:49 | |
| *** zhangjn has quit IRC | 01:50 | |
| *** wanghao has joined #openstack-cinder | 01:51 | |
| *** alonma has joined #openstack-cinder | 01:51 | |
| *** zhangjn has joined #openstack-cinder | 01:53 | |
| *** alonma has quit IRC | 01:56 | |
| *** jacky-zhang_ has joined #openstack-cinder | 01:59 | |
| *** jacky-zhang has quit IRC | 02:00 | |
| *** jacky-zhang_ is now known as jacky-zhang | 02:00 | |
| *** haomaiwang has quit IRC | 02:01 | |
| *** haomaiwang has joined #openstack-cinder | 02:01 | |
| *** abhi has quit IRC | 02:01 | |
| *** zhangjn has quit IRC | 02:02 | |
| *** wenjuan has joined #openstack-cinder | 02:03 | |
| *** wenjuan has quit IRC | 02:03 | |
| *** wenjuan has joined #openstack-cinder | 02:04 | |
| *** wer_ has joined #openstack-cinder | 02:04 | |
| *** wenjuan has quit IRC | 02:05 | |
| *** zhangjn has joined #openstack-cinder | 02:08 | |
| *** laughterwym has joined #openstack-cinder | 02:08 | |
| *** gouthamr has quit IRC | 02:12 | |
| *** alonma has joined #openstack-cinder | 02:13 | |
| *** alonma has quit IRC | 02:17 | |
| *** cknight1 has quit IRC | 02:18 | |
| *** alonma has joined #openstack-cinder | 02:19 | |
| *** chenying has joined #openstack-cinder | 02:23 | |
| *** houming has joined #openstack-cinder | 02:23 | |
| *** alonma has quit IRC | 02:23 | |
| *** alonma has joined #openstack-cinder | 02:25 | |
| *** alonma has quit IRC | 02:30 | |
| *** alonma has joined #openstack-cinder | 02:32 | |
| *** crose has quit IRC | 02:33 | |
| *** diogogmt has joined #openstack-cinder | 02:33 | |
| *** mriedem has quit IRC | 02:34 | |
| *** alonma has quit IRC | 02:37 | |
| *** zhangjn has quit IRC | 02:37 | |
| *** sombrafam has quit IRC | 02:40 | |
| *** julim has joined #openstack-cinder | 02:42 | |
| *** gcb has joined #openstack-cinder | 02:44 | |
| *** wer_ has quit IRC | 02:46 | |
| *** edmondsw has quit IRC | 02:49 | |
| *** diablo_rojo has quit IRC | 02:50 | |
| *** diablo_rojo has joined #openstack-cinder | 02:50 | |
| openstackgerrit | Peter Wang proposed openstack/cinder: Fix inconsistency and typos in replication v2 DOC https://review.openstack.org/265678 | 02:58 |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-cinder | 03:01 | |
| lixiaoy1 | wanghao: ping | 03:02 |
| wanghao | lixiaoy1: hi | 03:03 |
| lixiaoy1 | wanghao: do you have time? i am working on a bug about quota, it seems that you have lots of experiences in this area, may we can work together to fix it? | 03:04 |
| *** mtanino has quit IRC | 03:04 | |
| wanghao | lixiaoy1: sure, I'd glad to. | 03:05 |
| lixiaoy1 | wanghao: the bug is https://bugs.launchpad.net/cinder/+bug/1473183 | 03:05 |
| openstack | Launchpad bug 1473183 in Cinder "quota usage is wrong after volume type rename" [High,Confirmed] - Assigned to Lisa Li (lisali) | 03:05 |
| wanghao | lixiaoy1: got it, check it now | 03:06 |
| lixiaoy1 | wanghao: :) this bug is about quota when volume type is renamed. as you konw, it uses volume_type_name in quota_usage tables | 03:07 |
| openstackgerrit | Mudassir Latif proposed openstack/cinder: Replication V2 for Pure Storage https://review.openstack.org/214825 | 03:07 |
| lixiaoy1 | wanghao: as a result, when volume type is renamed, the relationship between quota and volume types are corrupted | 03:07 |
| *** dims has quit IRC | 03:11 | |
| lixiaoy1 | wanghao: let's talk after you read the bug. I have some solutions... | 03:11 |
| *** mudassirlatif has quit IRC | 03:11 | |
| wanghao | lixiaoy1: great, I need some tests as you did. | 03:12 |
| *** diablo_rojo has quit IRC | 03:13 | |
| *** diablo_rojo has joined #openstack-cinder | 03:14 | |
| *** alkhodos has quit IRC | 03:22 | |
| *** itzsunkum has quit IRC | 03:22 | |
| *** smoriya_ has quit IRC | 03:23 | |
| *** thambi has quit IRC | 03:23 | |
| openstackgerrit | chenzongliang proposed openstack/cinder: Huawei: Refactor driver for the second time https://review.openstack.org/256920 | 03:24 |
| *** apoorvad has quit IRC | 03:27 | |
| *** smoriya_ has joined #openstack-cinder | 03:28 | |
| *** jacky-zhang has quit IRC | 03:30 | |
| wanghao | lixiaoy1: okay, I see this, because cinder just update the volume type record in DB, but didn't update the quota_usage table at same time. | 03:31 |
| lixiaoy1 | wanghao: yes | 03:31 |
| lixiaoy1 | wanghao: what I am concerned is when to update quota_usage? | 03:31 |
| lixiaoy1 | wanghao: as there may be still reversations during the time | 03:32 |
| *** alonma has joined #openstack-cinder | 03:36 | |
| lixiaoy1 | wanghao: lunch now, and back in one hour. talk to you later | 03:39 |
| *** alonma has quit IRC | 03:41 | |
| wanghao | lixiaoy1: emm, if there is reversations, we need to update the resource in reversations too before commit. Otherwise, it still be error. | 03:42 |
| wanghao | lixiaoy1: sure. | 03:42 |
| *** alonma has joined #openstack-cinder | 03:44 | |
| *** cknight has joined #openstack-cinder | 03:47 | |
| *** alonma has quit IRC | 03:48 | |
| *** links has joined #openstack-cinder | 03:49 | |
| *** alonma has joined #openstack-cinder | 03:50 | |
| *** hyakuhei has joined #openstack-cinder | 03:51 | |
| openstackgerrit | Liucheng Jiang proposed openstack/cinder: Huawei: Implement v2 replication (managed) https://review.openstack.org/261246 | 03:54 |
| *** mragupat has joined #openstack-cinder | 03:55 | |
| *** alonma has quit IRC | 03:55 | |
| *** mragupat has quit IRC | 03:55 | |
| *** mragupat has joined #openstack-cinder | 03:56 | |
| *** links has quit IRC | 03:56 | |
| *** haomaiwang has quit IRC | 04:01 | |
| *** alonma has joined #openstack-cinder | 04:01 | |
| *** haomaiwa_ has joined #openstack-cinder | 04:01 | |
| *** alonma has quit IRC | 04:05 | |
| *** diablo_rojo has quit IRC | 04:06 | |
| *** alonma has joined #openstack-cinder | 04:07 | |
| *** salv-orlando has joined #openstack-cinder | 04:11 | |
| *** alonma has quit IRC | 04:11 | |
| *** salv-orl_ has quit IRC | 04:13 | |
| *** alonma has joined #openstack-cinder | 04:14 | |
| *** zhangjn has joined #openstack-cinder | 04:17 | |
| *** alonma has quit IRC | 04:18 | |
| *** alonma has joined #openstack-cinder | 04:21 | |
| *** alonma has quit IRC | 04:25 | |
| *** zhangjn has quit IRC | 04:25 | |
| *** jacky-zhang has joined #openstack-cinder | 04:26 | |
| *** alonma has joined #openstack-cinder | 04:27 | |
| *** apoorvad has joined #openstack-cinder | 04:28 | |
| *** coolsvap|away is now known as coolsvap | 04:28 | |
| *** cheneydc has quit IRC | 04:31 | |
| *** alonma has quit IRC | 04:31 | |
| *** jacky-zhang_ has joined #openstack-cinder | 04:31 | |
| *** vgridnev has joined #openstack-cinder | 04:33 | |
| *** jacky-zhang has quit IRC | 04:33 | |
| *** jacky-zhang_ is now known as jacky-zhang | 04:33 | |
| *** zhangjn has joined #openstack-cinder | 04:35 | |
| *** itlinux has joined #openstack-cinder | 04:35 | |
| *** alonma has joined #openstack-cinder | 04:38 | |
| *** cknight has quit IRC | 04:38 | |
| *** links has joined #openstack-cinder | 04:40 | |
| *** apoorvad has quit IRC | 04:42 | |
| *** alonma has quit IRC | 04:42 | |
| *** alonma has joined #openstack-cinder | 04:44 | |
| *** cheneydc has joined #openstack-cinder | 04:45 | |
| *** lcurtis has quit IRC | 04:46 | |
| *** alonma has quit IRC | 04:49 | |
| *** alonma has joined #openstack-cinder | 04:50 | |
| *** zhangjn has quit IRC | 04:53 | |
| *** haomaiwa_ has quit IRC | 04:53 | |
| *** sheel has joined #openstack-cinder | 04:53 | |
| *** laughterwym has quit IRC | 04:53 | |
| *** hyakuhei has quit IRC | 04:53 | |
| *** 77CAAJ1C4 has joined #openstack-cinder | 04:54 | |
| *** alonma has quit IRC | 04:54 | |
| lixiaoy1 | wanghao: my concerns about this bug: when renaming volume type, if we updated "resource" field in quota_usage and reservation table, this just prevent the bug in future. | 04:56 |
| lixiaoy1 | wanghao: another concern is to clear current invalid rows in quota_usages. | 04:56 |
| lixiaoy1 | wanghao: as there are non-deleted and invalid rows in quota_usages if customers upgrade cinder from liberty. these invalid rows may lead problems | 04:57 |
| lixiaoy1 | wanghao: or we need to check whether existed resources when creating a volume type | 04:58 |
| *** jacky-zhang has quit IRC | 05:00 | |
| *** 77CAAJ1C4 has quit IRC | 05:01 | |
| *** alonma has joined #openstack-cinder | 05:01 | |
| *** haomaiwa_ has joined #openstack-cinder | 05:01 | |
| *** alonma has quit IRC | 05:05 | |
| *** itlinux has quit IRC | 05:06 | |
| *** alonma has joined #openstack-cinder | 05:07 | |
| *** alonma has quit IRC | 05:11 | |
| *** merooney has quit IRC | 05:13 | |
| *** alonma has joined #openstack-cinder | 05:13 | |
| *** vivekd has joined #openstack-cinder | 05:16 | |
| *** daneyon has joined #openstack-cinder | 05:16 | |
| *** alonma has quit IRC | 05:18 | |
| *** asselin has joined #openstack-cinder | 05:18 | |
| *** daneyon__ has quit IRC | 05:19 | |
| *** asselin__ has quit IRC | 05:19 | |
| *** alonma has joined #openstack-cinder | 05:20 | |
| *** daneyon has quit IRC | 05:21 | |
| *** chlong has joined #openstack-cinder | 05:24 | |
| *** alonma has quit IRC | 05:25 | |
| *** harlowja_at_home has joined #openstack-cinder | 05:26 | |
| *** alonma has joined #openstack-cinder | 05:27 | |
| *** gcb has quit IRC | 05:30 | |
| *** anshul has joined #openstack-cinder | 05:31 | |
| *** alonma has quit IRC | 05:32 | |
| *** alonma has joined #openstack-cinder | 05:35 | |
| *** zhangjn has joined #openstack-cinder | 05:35 | |
| *** sgotliv has joined #openstack-cinder | 05:36 | |
| *** zhangjn has quit IRC | 05:36 | |
| *** itlinux has joined #openstack-cinder | 05:37 | |
| *** zhangjn has joined #openstack-cinder | 05:37 | |
| *** pots has quit IRC | 05:37 | |
| *** alonma has quit IRC | 05:39 | |
| *** zhangjn_ has joined #openstack-cinder | 05:39 | |
| *** houming has quit IRC | 05:40 | |
| *** zhangjn_ has quit IRC | 05:40 | |
| *** zhangjn_ has joined #openstack-cinder | 05:41 | |
| *** gcb has joined #openstack-cinder | 05:42 | |
| *** alonma has joined #openstack-cinder | 05:42 | |
| *** zhangjn_ has quit IRC | 05:42 | |
| *** zhangjn has quit IRC | 05:43 | |
| *** zhangjn has joined #openstack-cinder | 05:43 | |
| *** alonma has quit IRC | 05:47 | |
| *** sgotliv has quit IRC | 05:47 | |
| *** zhangjn has quit IRC | 05:47 | |
| *** zhangjn has joined #openstack-cinder | 05:47 | |
| *** laughterwym has joined #openstack-cinder | 05:54 | |
| *** laughterwym has quit IRC | 05:59 | |
| *** vivekd has quit IRC | 06:01 | |
| *** haomaiwa_ has quit IRC | 06:01 | |
| *** haomaiwang has joined #openstack-cinder | 06:01 | |
| *** harlowja_at_home has quit IRC | 06:02 | |
| *** boris-42 has quit IRC | 06:03 | |
| *** nkrinner has joined #openstack-cinder | 06:05 | |
| openstackgerrit | Angus Lees proposed openstack/os-brick: Migrate remotefs from rootwrap to privsep https://review.openstack.org/258252 | 06:09 |
| *** vgridnev has quit IRC | 06:12 | |
| *** kaisers has joined #openstack-cinder | 06:26 | |
| *** ChubYann has quit IRC | 06:28 | |
| yhayashi | smcginnis,xyang: Hi, I updated the patch about Fujits ETERNUS DX FC Driver(https://review.openstack.org/263941 ), and CI is Happy. Could you review it? | 06:32 |
| yhayashi | I'm really sorry for tight schedule of this driver review. | 06:33 |
| *** laughterwym has joined #openstack-cinder | 06:36 | |
| *** chlong has quit IRC | 06:40 | |
| *** alonma has joined #openstack-cinder | 06:42 | |
| *** sheel has quit IRC | 06:43 | |
| *** cfriesen has quit IRC | 06:44 | |
| *** sheel has joined #openstack-cinder | 06:46 | |
| *** alonma has quit IRC | 06:47 | |
| *** alonma has joined #openstack-cinder | 06:48 | |
| *** shyama has quit IRC | 06:51 | |
| *** vgridnev has joined #openstack-cinder | 06:51 | |
| *** shyama has joined #openstack-cinder | 06:51 | |
| *** alonma has quit IRC | 06:53 | |
| *** alonma has joined #openstack-cinder | 06:55 | |
| *** shyama has quit IRC | 06:55 | |
| *** shyama has joined #openstack-cinder | 06:55 | |
| *** teju has joined #openstack-cinder | 06:58 | |
| *** alonma has quit IRC | 06:59 | |
| *** haomaiwang has quit IRC | 07:01 | |
| *** 17WABJ0T0 has joined #openstack-cinder | 07:01 | |
| *** alonma has joined #openstack-cinder | 07:02 | |
| openstackgerrit | huanan proposed openstack/cinder: Huawei: Add manage/unmanage snapshot support https://review.openstack.org/261156 | 07:03 |
| *** alonma has quit IRC | 07:06 | |
| openstackgerrit | dongwenjuan proposed openstack/cinder-specs: Add ServiceGroup using Tooz driver https://review.openstack.org/258968 | 07:08 |
| *** alonma has joined #openstack-cinder | 07:09 | |
| *** shyama_ has joined #openstack-cinder | 07:13 | |
| *** alonma has quit IRC | 07:14 | |
| *** shyama has quit IRC | 07:15 | |
| *** shyama_ is now known as shyama | 07:15 | |
| *** alonma has joined #openstack-cinder | 07:16 | |
| *** links has quit IRC | 07:17 | |
| *** shyama has quit IRC | 07:17 | |
| *** shyama has joined #openstack-cinder | 07:17 | |
| *** vivekd has joined #openstack-cinder | 07:19 | |
| *** alonma has quit IRC | 07:21 | |
| *** alonma has joined #openstack-cinder | 07:24 | |
| *** coolsvap is now known as coolsvap|away | 07:25 | |
| *** alonma has quit IRC | 07:27 | |
| *** alonma has joined #openstack-cinder | 07:27 | |
| openstackgerrit | huanan proposed openstack/cinder: Huawei: Add manage/unmanage snapshot support https://review.openstack.org/261156 | 07:28 |
| *** zwei_ has joined #openstack-cinder | 07:29 | |
| *** lpetrut has joined #openstack-cinder | 07:29 | |
| zwei_ | hi all . join cinder | 07:30 |
| *** links has joined #openstack-cinder | 07:30 | |
| *** savihou has joined #openstack-cinder | 07:34 | |
| *** ociuhandu has joined #openstack-cinder | 07:35 | |
| *** sgotliv has joined #openstack-cinder | 07:37 | |
| *** links has quit IRC | 07:39 | |
| *** Lallana has quit IRC | 07:40 | |
| *** vivekd has quit IRC | 07:41 | |
| *** links has joined #openstack-cinder | 07:42 | |
| *** zwei_ has quit IRC | 07:44 | |
| *** vivekd has joined #openstack-cinder | 07:45 | |
| *** salv-orlando has quit IRC | 07:47 | |
| *** alonma has quit IRC | 07:48 | |
| *** salv-orlando has joined #openstack-cinder | 07:48 | |
| *** alonma has joined #openstack-cinder | 07:48 | |
| *** alonma has quit IRC | 07:53 | |
| *** alonma has joined #openstack-cinder | 07:54 | |
| *** Lallana has joined #openstack-cinder | 07:55 | |
| *** mragupat has quit IRC | 07:58 | |
| *** alonma has quit IRC | 07:59 | |
| *** 17WABJ0T0 has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-cinder | 08:01 | |
| wanghao | lixiaoy1: sorry, have a meeting afternoon. | 08:05 |
| wanghao | lixiaoy1: You're right, in the upgrade case, we should consider how to remove those invalid rows in DB. | 08:07 |
| lixiaoy1 | wanghao: I wrote some codes, you may have a look at first. let's decide how to do it | 08:08 |
| lixiaoy1 | wanghao: it doesnot matter | 08:08 |
| *** alonma has joined #openstack-cinder | 08:09 | |
| wanghao | lixiaoy1: Sure, In my mind, I prefer to use cinder-manage to let users can fix those invalid data. | 08:09 |
| wanghao | lixiaoy1: I will check the code first. | 08:09 |
| wanghao | lixiaoy1: If you commit your code, please let me know, thx :) | 08:11 |
| lixiaoy1 | wanghao: test codes are not ready, :) let's agree on the solution at first | 08:12 |
| *** ociuhandu has quit IRC | 08:13 | |
| *** alonma has quit IRC | 08:13 | |
| wanghao | lixiaoy1: First step, update resource in quota_usage and reservation table when updating type's name. | 08:13 |
| wanghao | lixiaoy1: Second step: Using cinder-manage to fix invalid rows if we upgrade cinder from liberty. | 08:14 |
| wanghao | lixiay2: This's my thought, how do you think? | 08:14 |
| openstackgerrit | LisaLi proposed openstack/cinder: Update quota when volume type renames. https://review.openstack.org/267999 | 08:14 |
| *** ociuhandu has joined #openstack-cinder | 08:15 | |
| wanghao | lixiaoy1: okay, check the code. | 08:15 |
| lixiaoy1 | wanghao: please | 08:15 |
| *** vivekd has quit IRC | 08:16 | |
| *** vgridnev has quit IRC | 08:16 | |
| lixiaoy1 | wanghao: in first step, no need to update reservations, as we can mapping reservation to quota_usage by id (id in quota_usage) which will not change | 08:16 |
| lixiaoy1 | wanghao: feel free to update | 08:17 |
| lixiaoy1 | wanghao: and add your name as co-author there | 08:18 |
| wanghao | lixiaoy1: okay, thx. | 08:19 |
| *** zhangjn has quit IRC | 08:19 | |
| openstackgerrit | Merged openstack/cinder: LeftHand: Implement v2 replication (managed) https://review.openstack.org/255002 | 08:22 |
| wanghao | lixiaoy1: I think we should update reservations since the reservations have resource name too, and when committing reservation, cinder use the resource name to find usage: usage = usages[reservation.resource] | 08:23 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements https://review.openstack.org/266148 | 08:23 |
| openstackgerrit | Merged openstack/cinder: Fix ChunkedBackupDriver _create_container https://review.openstack.org/267608 | 08:24 |
| wanghao | lixiaoy1: So if the reservation.resource is old name, and usage.resource is new name, that's will be mistake. | 08:24 |
| lixiaoy1 | wanghao: this will not lead to problems. as when commit/rollback, we use quota_usage id to find quota_usage | 08:26 |
| *** lpetrut has quit IRC | 08:27 | |
| *** houming has joined #openstack-cinder | 08:32 | |
| *** mudassirlatif has joined #openstack-cinder | 08:41 | |
| *** yhayashi has quit IRC | 08:44 | |
| *** manous has joined #openstack-cinder | 08:45 | |
| *** itlinux has quit IRC | 08:49 | |
| *** zongliang has joined #openstack-cinder | 08:49 | |
| *** salv-orlando has quit IRC | 08:55 | |
| *** salv-orlando has joined #openstack-cinder | 08:56 | |
| openstackgerrit | yifan403 proposed openstack/cinder: Add ZTE Block Storage Driver https://review.openstack.org/258880 | 08:56 |
| *** ankit_ag has joined #openstack-cinder | 08:58 | |
| *** geguileo_mobile has quit IRC | 08:59 | |
| *** haomaiwang has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-cinder | 09:01 | |
| *** geguileo_phone has joined #openstack-cinder | 09:03 | |
| *** zhangjn has joined #openstack-cinder | 09:04 | |
| *** belmoreira has joined #openstack-cinder | 09:05 | |
| *** alonma has joined #openstack-cinder | 09:09 | |
| *** nkrinner has quit IRC | 09:14 | |
| *** alonma has quit IRC | 09:14 | |
| *** nkrinner has joined #openstack-cinder | 09:14 | |
| *** vgridnev has joined #openstack-cinder | 09:15 | |
| *** savihou has quit IRC | 09:15 | |
| *** chhavi has joined #openstack-cinder | 09:15 | |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Add missing RPC calls version to rpcapi's https://review.openstack.org/268024 | 09:15 |
| openstackgerrit | Michal Dulko proposed openstack/cinder: WIP: Report RPC and objects versions https://review.openstack.org/268025 | 09:15 |
| *** alonma has joined #openstack-cinder | 09:15 | |
| openstackgerrit | Michal Dulko proposed openstack/cinder: WIP: Pin RPC and object version to lowest running https://review.openstack.org/268026 | 09:15 |
| *** savihou has joined #openstack-cinder | 09:17 | |
| *** alonma has quit IRC | 09:20 | |
| *** mudassirlatif has quit IRC | 09:21 | |
| *** jordanP has joined #openstack-cinder | 09:22 | |
| wanghao | lixiaoy1: I'm not sure this, you can see reservation_commit in db api. We get the quota_usage filtering by project_id, and then use usages[reservation.resource] to get the usage. Anyway, I will test the code. Thx :) | 09:25 |
| *** savihou has quit IRC | 09:25 | |
| *** houming has quit IRC | 09:26 | |
| openstackgerrit | Alexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver. https://review.openstack.org/190273 | 09:26 |
| openstackgerrit | huanan proposed openstack/cinder: Huawei: Add manage/unmanage snapshot support https://review.openstack.org/261156 | 09:26 |
| *** jistr has joined #openstack-cinder | 09:27 | |
| *** zhangjn has quit IRC | 09:34 | |
| *** manous has quit IRC | 09:35 | |
| *** manous has joined #openstack-cinder | 09:36 | |
| *** teju has quit IRC | 09:36 | |
| *** shyama has quit IRC | 09:43 | |
| openstackgerrit | melissaml proposed openstack/python-cinderclient: Put py34 first in the env order of tox https://review.openstack.org/266761 | 09:46 |
| *** zongliang has quit IRC | 09:46 | |
| *** zongliang has joined #openstack-cinder | 09:47 | |
| *** Lallana has quit IRC | 09:47 | |
| *** Lallana has joined #openstack-cinder | 09:47 | |
| *** vivekd has joined #openstack-cinder | 09:51 | |
| *** salv-orlando has quit IRC | 09:52 | |
| *** salv-orlando has joined #openstack-cinder | 09:52 | |
| *** davechen has left #openstack-cinder | 09:55 | |
| *** coolsvap|away is now known as coolsvap | 09:56 | |
| *** cheneydc has quit IRC | 09:59 | |
| *** haomaiwang has quit IRC | 10:01 | |
| *** manous has quit IRC | 10:01 | |
| *** lpetrut has joined #openstack-cinder | 10:01 | |
| *** haomaiwang has joined #openstack-cinder | 10:01 | |
| *** Thelo has quit IRC | 10:07 | |
| *** EinstCrazy has quit IRC | 10:08 | |
| *** e0ne has joined #openstack-cinder | 10:08 | |
| *** EinstCrazy has joined #openstack-cinder | 10:08 | |
| *** Thelo has joined #openstack-cinder | 10:10 | |
| *** salv-orl_ has joined #openstack-cinder | 10:10 | |
| *** ociuhandu has quit IRC | 10:11 | |
| tiagogomes | Hi, is there any relation between nova availability zones and cinder availability zones? I wonder about the best way of creating an LVM volume in a node where a VM using that volume will run | 10:12 |
| *** ociuhandu has joined #openstack-cinder | 10:13 | |
| *** salv-orlando has quit IRC | 10:13 | |
| *** EinstCrazy has quit IRC | 10:14 | |
| dulek | tiagogomes: Nova is sending AZ name along with boot-from-volume request. | 10:14 |
| dulek | tiagogomes: But… Why not use cinder.scheduler.filters.instance_locality_filter ? | 10:14 |
| *** alonma has joined #openstack-cinder | 10:16 | |
| *** jordanP has quit IRC | 10:16 | |
| tiagogomes | dulek so in the boot-from-volume case, there needs to be a matching availability zone name? | 10:17 |
| *** Thelo has quit IRC | 10:19 | |
| dulek | tiagogomes: By default - yes. You may set availability_zone_fallback option in cinder.conf and disable availability_zone_filter if you don't care about Cinder's AZs. | 10:19 |
| tiagogomes | dulek thanks! | 10:20 |
| *** alonma has quit IRC | 10:20 | |
| *** jordanP has joined #openstack-cinder | 10:21 | |
| *** lpetrut1 has joined #openstack-cinder | 10:22 | |
| *** alonma has joined #openstack-cinder | 10:22 | |
| *** ndipanov has joined #openstack-cinder | 10:22 | |
| *** lpetrut has quit IRC | 10:23 | |
| *** lpetrut1 is now known as lpetrut | 10:23 | |
| yuriy_n17 | dulek: Hi Michal. | 10:25 |
| openstackgerrit | Wilson Liu proposed openstack/cinder: Huawei: Balanced FC port selection when zoning https://review.openstack.org/261151 | 10:25 |
| dulek | yuriy_n17: Hi! | 10:25 |
| yuriy_n17 | dulek: I'd like to discuss https://review.openstack.org/#/c/267656/ | 10:25 |
| yuriy_n17 | dulek: It's only a small patch to fix that bug. | 10:26 |
| *** alonma has quit IRC | 10:26 | |
| dulek | yuriy_n17: So my own impression is that this isn't a bug actually. | 10:27 |
| *** Thelo has joined #openstack-cinder | 10:27 | |
| dulek | yuriy_n17: It doens't block anything as we don't support Python 3 yet. | 10:27 |
| openstackgerrit | Philipp Marek proposed openstack/cinder: Infrastructure to use the DRBD transport for NOVA https://review.openstack.org/259973 | 10:27 |
| dulek | yuriy_n17: We *do* Python 3 compatibility work as a blueprint Victor is contributing to. | 10:27 |
| *** alonma has joined #openstack-cinder | 10:28 | |
| dulek | yuriy_n17: Ah, found it: http://lists.openstack.org/pipermail/openstack-dev/2016-January/083877.html | 10:29 |
| dulek | yuriy_n17: I personally don't like such small global bugs and agree with Julien's opinion there completely. | 10:29 |
| *** ntt has joined #openstack-cinder | 10:30 | |
| dulek | yuriy_n17: So if we want to do something with Python 3 compatibility - let's do it completely. | 10:30 |
| yuriy_n17 | dulek: I agree with you that it is not bug. You offer to port full test? | 10:30 |
| ntt | Hi, it is possible to use openstack-cinder-volume in an active/passive configuration? I'm using the icehouse release | 10:30 |
| *** lpetrut has quit IRC | 10:31 | |
| ntt | Actually I'm trying to use the openstack-resource-agents, but it fails with an error message: "ERROR: Cinder Volume is not connected to the AMQP server: 1" | 10:31 |
| dulek | yuriy_n17: I would be totally supportive if changes would result in having whole test module to be added to test-py3.txt. | 10:31 |
| ntt | It seems that it is not possible to detach openstack-cinder-volume from scheduler and api using the Ha configuration | 10:32 |
| dulek | yuriy_n17: I don't however have any resources to actually contribute to this work in Mitaka. | 10:32 |
| *** lpetrut has joined #openstack-cinder | 10:32 | |
| *** sheel has quit IRC | 10:32 | |
| *** alonma has quit IRC | 10:33 | |
| dulek | ntt: It should be possible, but it needs AMQP access. For pacemaker-related issues I recommend asking at #openstack-ha | 10:33 |
| yuriy_n17 | dulek: Ok, I don't object against it. | 10:33 |
| ntt | dulek: ok, thank you | 10:33 |
| *** ildikov has quit IRC | 10:37 | |
| *** smoriya_ has quit IRC | 10:39 | |
| openstackgerrit | Ankit Agrawal proposed openstack/cinder: Python 3: Replace reduce and xrange with six.moves https://review.openstack.org/268059 | 10:39 |
| *** Thelo has quit IRC | 10:40 | |
| *** Thelo has joined #openstack-cinder | 10:41 | |
| openstackgerrit | ting wang proposed openstack/cinder: Remove run_tests.sh https://review.openstack.org/268065 | 10:47 |
| *** sileht has quit IRC | 10:48 | |
| *** sileht has joined #openstack-cinder | 10:48 | |
| *** haomaiwang has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-cinder | 11:01 | |
| *** dims has joined #openstack-cinder | 11:12 | |
| openstackgerrit | Merged openstack/cinder: Base iSCSI initiator validation https://review.openstack.org/267706 | 11:25 |
| *** tpsilva has joined #openstack-cinder | 11:31 | |
| *** haomaiwang has quit IRC | 11:40 | |
| *** coolsvap is now known as coolsvap|away | 11:43 | |
| openstackgerrit | Michal Dulko proposed openstack/cinder: WIP: Pin RPC and object version to lowest running https://review.openstack.org/268026 | 11:45 |
| openstackgerrit | Michal Dulko proposed openstack/cinder: WIP: Report RPC and objects versions https://review.openstack.org/268025 | 11:45 |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Add missing RPC calls version to rpcapi's https://review.openstack.org/268024 | 11:45 |
| *** merooney has joined #openstack-cinder | 11:51 | |
| *** ociuhandu has quit IRC | 11:55 | |
| *** gouthamr has joined #openstack-cinder | 12:09 | |
| *** salv-orl_ has quit IRC | 12:15 | |
| *** salv-orlando has joined #openstack-cinder | 12:15 | |
| *** alonma has joined #openstack-cinder | 12:17 | |
| *** ildikov has joined #openstack-cinder | 12:18 | |
| *** haomaiwa_ has joined #openstack-cinder | 12:19 | |
| *** ociuhandu has joined #openstack-cinder | 12:20 | |
| *** haomaiwa_ has quit IRC | 12:20 | |
| *** alonma has quit IRC | 12:21 | |
| *** alonma has joined #openstack-cinder | 12:23 | |
| *** alonma has quit IRC | 12:27 | |
| *** EinstCrazy has joined #openstack-cinder | 12:28 | |
| *** alonma has joined #openstack-cinder | 12:29 | |
| *** haomaiwa_ has joined #openstack-cinder | 12:31 | |
| *** alonma has quit IRC | 12:34 | |
| *** cdelatte has joined #openstack-cinder | 12:35 | |
| *** alonma has joined #openstack-cinder | 12:36 | |
| *** jwcroppe has joined #openstack-cinder | 12:39 | |
| *** alonma has quit IRC | 12:40 | |
| *** salv-orlando has quit IRC | 12:40 | |
| *** salv-orlando has joined #openstack-cinder | 12:40 | |
| *** alonma has joined #openstack-cinder | 12:42 | |
| *** ankit_ag has quit IRC | 12:44 | |
| *** chhavi has quit IRC | 12:44 | |
| *** alonma has quit IRC | 12:47 | |
| *** alonma has joined #openstack-cinder | 12:49 | |
| *** crose has joined #openstack-cinder | 12:49 | |
| *** openstackgerrit has quit IRC | 12:50 | |
| *** openstackgerrit has joined #openstack-cinder | 12:51 | |
| *** alonma has quit IRC | 12:53 | |
| openstackgerrit | Alexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver. https://review.openstack.org/190273 | 12:54 |
| *** links has quit IRC | 12:55 | |
| *** alonma has joined #openstack-cinder | 12:56 | |
| *** links has joined #openstack-cinder | 12:58 | |
| *** alonma has quit IRC | 13:00 | |
| *** haomaiwa_ has quit IRC | 13:01 | |
| *** raildo-afk is now known as raildo | 13:01 | |
| *** haomaiwang has joined #openstack-cinder | 13:01 | |
| *** links has quit IRC | 13:03 | |
| *** jwcroppe has quit IRC | 13:08 | |
| *** hyakuhei has joined #openstack-cinder | 13:10 | |
| *** dslevin has quit IRC | 13:22 | |
| *** fthiagogv has quit IRC | 13:24 | |
| *** cdelatte has quit IRC | 13:33 | |
| *** cdelatte has joined #openstack-cinder | 13:34 | |
| *** vivekd has quit IRC | 13:36 | |
| *** hyakuhei has quit IRC | 13:39 | |
| *** cbader_2 has quit IRC | 13:41 | |
| *** Yogi1 has joined #openstack-cinder | 13:48 | |
| *** sombrafam has joined #openstack-cinder | 13:48 | |
| *** timcl has joined #openstack-cinder | 13:49 | |
| geguileo | dulek: ping - HA Job Distribution | 13:50 |
| dulek | geguileo: I'm in a meeting, but go on. :) | 13:51 |
| geguileo | dulek: Regarding your comments on host/cluster, etc | 13:51 |
| dulek | Yeah it's complicated, I know… :( | 13:51 |
| geguileo | dulek: If I understand correctly rolling upgrades process | 13:52 |
| geguileo | We do DB changes, then we upgrade API, SCH, C-vol and c-back | 13:52 |
| geguileo | And while this is going on the cloud is expected to work seamlessly | 13:52 |
| dulek | That's right, c-bak can be upgraded before c-sch, it doesn't matter. | 13:53 |
| dulek | (only c-api talks with c-bak) | 13:53 |
| *** dslev has joined #openstack-cinder | 13:53 | |
| * dulek is always scared with such talks - what if I've missed something…? ;) | 13:53 | |
| geguileo | XD | 13:54 |
| geguileo | dulek: I have replied to your comments, but I'll sumarize it here | 13:54 |
| mkoderer | Hi! One quick question.. how do I set the capability "sparse_copy_volume".. is it defined per driver or configuration? | 13:54 |
| dulek | Oh, let me read then | 13:54 |
| dulek | geguileo: Am I understanding correctly that you're talking about `host` column in `volumes` table? | 13:56 |
| geguileo | dulek: Yes | 13:56 |
| *** alonma has joined #openstack-cinder | 13:56 | |
| dulek | geguileo: That's something I haven't had thought of - you don't mention any changes to `volumes` table. | 13:56 |
| geguileo | And in service and maybe in more | 13:57 |
| geguileo | My mind may be slipping, let me have another look | 13:57 |
| dulek | :) | 13:58 |
| *** cbader has joined #openstack-cinder | 13:58 | |
| dulek | So one case is writing a hostname to `host` field on `volume`. We take `cluster` field from service, if it's empty then we get `host` and write it to `volume`. | 13:59 |
| dulek | `cluster` has priority over `host` so once upgraded we only use `cluster`. | 14:00 |
| *** alonma has quit IRC | 14:01 | |
| *** haomaiwang has quit IRC | 14:01 | |
| dulek | On volume migration host field can change, but that's on `volume`, not `service` table. | 14:01 |
| geguileo | dulek: Yes, but now I'm thinking that maybe I didn't consider the implications of host fields on all the other tables | 14:01 |
| *** haomaiwang has joined #openstack-cinder | 14:01 | |
| dulek | You mean in the spec? I don't see an obvious problem there, but I get the point. | 14:02 |
| geguileo | I mean in *any* of the specs | 14:02 |
| dulek | But we're still scheduling volumes for a "cluster" and not for cluster+host pair, aren't we? | 14:04 |
| *** rlrossit has joined #openstack-cinder | 14:04 | |
| geguileo | dulek: Correct | 14:05 |
| *** alonma has joined #openstack-cinder | 14:05 | |
| geguileo | But in the cleanup spec I talk like we have also changed host in volumes to cluster | 14:06 |
| geguileo | So we really need to change host field everywhere | 14:06 |
| dulek | So the logic here won't change at all… Question is - do we use `host` column for something else than identifying where the resource is placed. | 14:06 |
| dulek | Is it totally required? That is only a rename. | 14:06 |
| geguileo | We only need a rename | 14:07 |
| geguileo | But since that is not possible... | 14:07 |
| *** porrua has joined #openstack-cinder | 14:07 | |
| dulek | It is. ;) Just takes time and planning. :) | 14:07 |
| geguileo | As I'm seeing it it's not possible :-( | 14:08 |
| geguileo | (I hope it's my mind being on Friday mode) | 14:08 |
| *** thangp has joined #openstack-cinder | 14:08 | |
| dulek | geguileo: https://specs.openstack.org/openstack/cinder-specs/specs/mitaka/online-schema-upgrades.html | 14:09 |
| dulek | geguileo: I believe this is the same situation. | 14:09 |
| dulek | For this particular rename we cannot use fast-track I've explained on your spec as it can be changed. And old service instance will write only to old place. | 14:10 |
| *** alonma has quit IRC | 14:10 | |
| smcginnis | Thelo: I forget, but did I tell you to make your Cinder driver depends-on the nova patch? | 14:11 |
| *** alonma has joined #openstack-cinder | 14:12 | |
| geguileo | dulek: Is it me or that is not correct? | 14:13 |
| geguileo | dulek: Because when you start rolling the upgrade to O first you change the DB and remove the old place field | 14:13 |
| geguileo | dulek: And then all N nodes will break when they try to write to "both places" | 14:14 |
| dulek | geguileo: "Therefore we can have a post-upgrade migration script that finally drops the column" | 14:14 |
| dulek | geguileo: Where post-upgrade is the key. | 14:14 |
| dulek | Well, it may work also to drop as first migration in P… Either's way would work. | 14:15 |
| geguileo | dulek: Then you should have corrected my initial message ;-P | 14:15 |
| geguileo | 14:52:28 geguileo | We do DB changes, then we upgrade API, SCH, C-vol and c-back | 14:15 |
| dulek | Oh, yes, yes, initially we should run migrations. | 14:16 |
| geguileo | dulek: I mean you should have told me that there could be post upgrade DB changes | 14:16 |
| geguileo | ;-) | 14:16 |
| dulek | That dropping one in O would be a special case. | 14:16 |
| *** alonma has quit IRC | 14:16 | |
| geguileo | dulek: Well, I'm not so sure it's a special case... | 14:17 |
| geguileo | dulek: As I see it we'll need it for HA as wel... | 14:17 |
| dulek | special case = "db post-upgrade-sync" | 14:17 |
| dulek | That's what I've meant. | 14:17 |
| dulek | Not a special case because it's an essential step of every that complicated migration. | 14:18 |
| geguileo | But the post upgrade is the one that drops it, and that's what HA will most likely need | 14:18 |
| *** alonma has joined #openstack-cinder | 14:18 | |
| dulek | Let's not use "special" then. ;) | 14:18 |
| geguileo | Ok, I'll think of a solution considering we can have an extra step to drop/sync things | 14:19 |
| geguileo | Because when I was thinking I kept running into the case where I needed that step | 14:19 |
| geguileo | XD | 14:19 |
| dulek | You mean in case you decide it's worth it to rename columns on Volume, Snapshot, … | 14:19 |
| dulek | ? | 14:19 |
| geguileo | dulek: Well, we can't have volume.host when it's not referring to the host but the cluster... | 14:20 |
| *** shyama has joined #openstack-cinder | 14:20 | |
| dulek | Technically we can. It's just confusing for developers. | 14:20 |
| geguileo | And for operators | 14:21 |
| geguileo | And when we talk to each other | 14:21 |
| geguileo | We won't know if we are talking about the cluster or the host | 14:21 |
| geguileo | XD | 14:21 |
| geguileo | That's the road to madness and a lot of angry people | 14:21 |
| dulek | You've managed to convince me that this is important. ;) | 14:22 |
| geguileo | dulek: Anyway, thanks for the help | 14:22 |
| *** alonma has quit IRC | 14:22 | |
| dulek | One more idea - what if we'll leave host field meaning "where the volume is placed" (meaning cluster in your model) and we'll add a new field that will serve as "field" in your model? | 14:23 |
| dulek | s/"field"/"host" | 14:24 |
| *** alonma has joined #openstack-cinder | 14:24 | |
| *** shyama has quit IRC | 14:26 | |
| *** chhavi has joined #openstack-cinder | 14:27 | |
| *** geguileo_phone has quit IRC | 14:27 | |
| *** cknight has joined #openstack-cinder | 14:27 | |
| geguileo | I think that's still confusing | 14:28 |
| openstackgerrit | Sean McGinnis proposed openstack/python-cinderclient: Update HACKING with current information https://review.openstack.org/265594 | 14:28 |
| *** akshai has joined #openstack-cinder | 14:28 | |
| geguileo | cluster and host are well know and specific names | 14:28 |
| *** xyang1 has joined #openstack-cinder | 14:28 | |
| geguileo | And host (meaning cluster) and field (meaning host) not so much | 14:28 |
| *** jgregor has joined #openstack-cinder | 14:29 | |
| *** dustins has joined #openstack-cinder | 14:29 | |
| *** alonma has quit IRC | 14:29 | |
| *** alonma has joined #openstack-cinder | 14:31 | |
| openstackgerrit | yifan403 proposed openstack/cinder: Add ZTE Block Storage Driver https://review.openstack.org/258880 | 14:33 |
| *** akerr has joined #openstack-cinder | 14:35 | |
| *** alonma has quit IRC | 14:35 | |
| *** hyakuhei has joined #openstack-cinder | 14:36 | |
| *** mriedem has joined #openstack-cinder | 14:36 | |
| *** alonma has joined #openstack-cinder | 14:37 | |
| *** krtaylor has quit IRC | 14:39 | |
| *** crose has quit IRC | 14:41 | |
| *** nkrinner has quit IRC | 14:41 | |
| *** alonma has quit IRC | 14:41 | |
| *** esker has quit IRC | 14:41 | |
| *** diablo_rojo has joined #openstack-cinder | 14:42 | |
| *** dansmith is now known as superdan | 14:43 | |
| *** alonma has joined #openstack-cinder | 14:43 | |
| *** hyakuhei has quit IRC | 14:45 | |
| *** cfriesen has joined #openstack-cinder | 14:45 | |
| *** hyakuhei has joined #openstack-cinder | 14:46 | |
| *** eharney has quit IRC | 14:46 | |
| e0ne | dulek: hi! how are you? | 14:46 |
| *** hyakuhei has quit IRC | 14:46 | |
| dulek | e0ne: Hi, I'm fine! What's up? | 14:47 |
| *** alonma has quit IRC | 14:47 | |
| e0ne | dulek: I'm reading your comments for https://etherpad.openstack.org/p/mitaka-cinder-spec-review-tracking | 14:48 |
| e0ne | dulek: especially "I wouldn't ever call this stable for M as we don't have a CI testing that yet." | 14:48 |
| dulek | e0ne: Ah, someone noticed. ;) | 14:48 |
| e0ne | dulek: I'm totally agree with this statement! | 14:48 |
| *** hyakuhei has joined #openstack-cinder | 14:48 | |
| e0ne | dulek: is any plan to introduce such kind of testing on gates? | 14:49 |
| dulek | I want to follow up with that through RC (if it won't be a hot period for me) or in early N. | 14:49 |
| dulek | Partial grenade is the thing, but the problem with it is that no one ever did it in 3-parts. | 14:50 |
| dulek | Nova for example is just upgrading controller services and running tempest against non-upgraded computes. | 14:50 |
| e0ne | dulek: let's add this topic to discuss at midcycle | 14:50 |
| dulek | I agree, we can add it. | 14:50 |
| e0ne | IMO, we have to do it | 14:51 |
| dulek | Although I don't know if we'll have a infra-magician there. ;) | 14:51 |
| e0ne | without test this extremly important feature will be broken sooner or earlier | 14:52 |
| dulek | e0ne: Or if I get hit by a bus and stop doing reviews. :D | 14:52 |
| dulek | Adding it right now. | 14:52 |
| e0ne | dulek: we have to discuss what can we do from the cinder side at least | 14:52 |
| dulek | "Are we going to setup CI for rolling upgrades or grenade job is enough?" | 14:53 |
| e0ne | dulek: don't get out the building before CI will be implemented ;) | 14:53 |
| e0ne | yes, sounds good | 14:53 |
| dulek | Okay, there's that item already in rolling upgrades section, so we're fine. I'll prepare for the discussion before the mid-cycle. | 14:53 |
| * dulek is running for another meeting… | 14:56 | |
| *** knomura has joined #openstack-cinder | 14:56 | |
| *** alonma has joined #openstack-cinder | 14:56 | |
| *** gabriel has quit IRC | 14:57 | |
| *** alonma has quit IRC | 15:00 | |
| *** haomaiwang has quit IRC | 15:01 | |
| *** haomaiwang has joined #openstack-cinder | 15:01 | |
| *** eharney has joined #openstack-cinder | 15:02 | |
| *** alonma has joined #openstack-cinder | 15:06 | |
| *** EinstCrazy has quit IRC | 15:07 | |
| *** alonma has quit IRC | 15:10 | |
| *** krtaylor has joined #openstack-cinder | 15:11 | |
| *** ericksonsantos has quit IRC | 15:11 | |
| *** alonma has joined #openstack-cinder | 15:12 | |
| *** lprice has joined #openstack-cinder | 15:14 | |
| *** jungleboyj has joined #openstack-cinder | 15:14 | |
| *** alonma has quit IRC | 15:16 | |
| *** mtanino has joined #openstack-cinder | 15:18 | |
| *** lprice1 has joined #openstack-cinder | 15:18 | |
| *** alonma has joined #openstack-cinder | 15:19 | |
| *** lprice has quit IRC | 15:20 | |
| *** raildo is now known as raildo-afk | 15:20 | |
| smcginnis | Reminder for cores that I should have brought up in the meeting: | 15:22 |
| smcginnis | New driver deadline is early next week. | 15:22 |
| smcginnis | So if you have time to get them feedback, sooner would be better. | 15:22 |
| *** lcurtis has joined #openstack-cinder | 15:23 | |
| smcginnis | And FYI for everyone - session proposals are open for the summit. | 15:23 |
| smcginnis | If you have anything you want to present, submit your topics soon. | 15:23 |
| *** alonma has quit IRC | 15:23 | |
| *** alonma has joined #openstack-cinder | 15:26 | |
| Thelo | smcginnis: Morning Sean, any news regarding the new os-brick release? | 15:26 |
| smcginnis | Thelo: Still in flight. | 15:26 |
| smcginnis | Thelo: I forget, did I tell you to make the cinder patch depends-on the nova patch? | 15:27 |
| openstackgerrit | Eric Harney proposed openstack/cinder: NetApp eseries: report max_over_subscription_ratio correctly https://review.openstack.org/267726 | 15:27 |
| smcginnis | Thelo: I vaguely remember discussing some kind of dependency, but I don't know why I would have done that. | 15:27 |
| smcginnis | Thelo: It actually seems backwards. You want the driver in first before nova can use it. | 15:27 |
| smcginnis | Thelo: So your driver patch is approved, but because of the depends-on in the commit it is holding. | 15:28 |
| Thelo | smcginnis: So the cinder patch cannot be merged unless the nova patch has merged? | 15:28 |
| smcginnis | Thelo: With that in the commit message, yes. | 15:28 |
| smcginnis | Thelo: So if you could update that to remove the depends-on line, I'll just reapprove it. | 15:28 |
| Thelo | smcginnis: ok | 15:29 |
| smcginnis | Thelo: Then at least your driver will get merged in Cinder and you just have the nova work outstanding. | 15:29 |
| *** martyturner has joined #openstack-cinder | 15:29 | |
| *** crose has joined #openstack-cinder | 15:29 | |
| smcginnis | Thelo: And sorry if I was the one that told you to do it in the first place. :) | 15:29 |
| *** alonma has quit IRC | 15:30 | |
| *** hyakuhei has quit IRC | 15:31 | |
| openstackgerrit | Thelo Gaultier proposed openstack/cinder: ITRI DISCO cinder driver https://review.openstack.org/253356 | 15:31 |
| *** SergeyLukjanov has quit IRC | 15:32 | |
| Thelo | smcginnis: about the deadline do I need the nova patch to be merged too? | 15:33 |
| *** aix has quit IRC | 15:33 | |
| smcginnis | Thelo: Nova deadlines are a whole other thing. | 15:33 |
| *** alonma has joined #openstack-cinder | 15:33 | |
| smcginnis | Thelo: I'm not sure what their policies are at the moment. | 15:33 |
| *** martyturner has quit IRC | 15:34 | |
| smcginnis | Thelo: It would probably be good to bring it up over in #openstack-nova to get the official word. | 15:35 |
| *** sheel has joined #openstack-cinder | 15:35 | |
| *** hyakuhei has joined #openstack-cinder | 15:36 | |
| *** martyturner has joined #openstack-cinder | 15:38 | |
| *** alonma has quit IRC | 15:38 | |
| *** SergeyLukjanov has joined #openstack-cinder | 15:38 | |
| smcginnis | What the heck. All of a sudden I'm having all kinds of DNS issues, including review.openstack.org. Anyone else having issues? | 15:38 |
| *** sgotliv has quit IRC | 15:39 | |
| *** mragupat has joined #openstack-cinder | 15:40 | |
| mtanino | I can open review.openstack.org without any problem. | 15:44 |
| *** lcurtis has quit IRC | 15:47 | |
| smcginnis | mtanino: Thanks. Looks like whatever DNS I'm going through is having issues. | 15:47 |
| *** wilson has joined #openstack-cinder | 15:47 | |
| *** wilson is now known as Guest89334 | 15:47 | |
| smcginnis | And now it's better. Sheesh | 15:47 |
| mtanino | :) | 15:48 |
| smcginnis | Darn computers. | 15:48 |
| *** edtubill has joined #openstack-cinder | 15:49 | |
| *** Guest89334 has quit IRC | 15:51 | |
| *** takedakn has joined #openstack-cinder | 15:51 | |
| *** wilson_liu has joined #openstack-cinder | 15:51 | |
| Thelo | smcginnis: FYI I just asked about the nova deadline and I have until 1/21 to merge the nova patch ( under a nova blueprint), Hence I should abandon the current patch for nova or simply update the blueprint in the commit message? | 15:54 |
| Thelo | smcginnis: For cinder part is it ok if the nova patch has not merged on 1/19? | 15:54 |
| sheel | geguileo:hi | 15:58 |
| *** takedakn has quit IRC | 15:59 | |
| *** baumann has joined #openstack-cinder | 16:00 | |
| *** haomaiwang has quit IRC | 16:01 | |
| *** aix has joined #openstack-cinder | 16:01 | |
| *** haomaiwang has joined #openstack-cinder | 16:01 | |
| smcginnis | Thelo: The cinder part is independent. | 16:03 |
| *** chris_morrell has joined #openstack-cinder | 16:04 | |
| *** belmoreira has quit IRC | 16:05 | |
| Swanson | ls | 16:06 |
| Swanson | damnit | 16:07 |
| smcginnis | :P | 16:07 |
| *** cbader has quit IRC | 16:07 | |
| *** salv-orl_ has joined #openstack-cinder | 16:10 | |
| *** chhavi has quit IRC | 16:12 | |
| *** edtubill has quit IRC | 16:12 | |
| *** lcurtis has joined #openstack-cinder | 16:13 | |
| *** salv-orlando has quit IRC | 16:13 | |
| *** edtubill has joined #openstack-cinder | 16:14 | |
| *** dslev has quit IRC | 16:20 | |
| sheel | smcginnis: hi there | 16:21 |
| sheel | smginnis: need your opinion on https://bugs.launchpad.net/cinder/+bug/1524450 | 16:23 |
| openstack | Launchpad bug 1524450 in Cinder "Filtering bootable volumes are not working for non-admin users" [Undecided,In progress] - Assigned to Sheel Rana (ranasheel2000) | 16:23 |
| sheel | smginnis: Under "launch instance" with non-admin user login, drop down list for bootable volumes shows non bootable volumes as well. This is because in _get_volumes, we remove filter options if they are other than " name, status, metadata, availability_zone" | 16:25 |
| *** garthb__ has joined #openstack-cinder | 16:25 | |
| *** vivekd has joined #openstack-cinder | 16:25 | |
| sheel | smginnis:reference: https://github.com/openstack/cinder/blob/master/cinder/api/v2/volumes.py#L229 | 16:26 |
| sheel | https://github.com/openstack/cinder/blob/master/cinder/api/v2/volumes.py#L39 | 16:26 |
| openstackgerrit | Alexey Khodos proposed openstack/cinder: NexentaStor 5 iSCSI backend driver. https://review.openstack.org/190391 | 16:27 |
| sheel | oh sorry, its smcginnis: | 16:27 |
| openstackgerrit | Alexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver. https://review.openstack.org/190273 | 16:28 |
| *** alkhodos has joined #openstack-cinder | 16:29 | |
| *** ericksonsantos has joined #openstack-cinder | 16:31 | |
| *** vgridnev has quit IRC | 16:32 | |
| *** sombrafam has quit IRC | 16:33 | |
| *** vgridnev has joined #openstack-cinder | 16:33 | |
| *** alonma has joined #openstack-cinder | 16:33 | |
| *** vgridnev has quit IRC | 16:34 | |
| *** cbader has joined #openstack-cinder | 16:35 | |
| openstackgerrit | Helen Walsh proposed openstack/cinder: EMC VMAX - necessary updates for CG changes https://review.openstack.org/268223 | 16:36 |
| *** daneyon has joined #openstack-cinder | 16:36 | |
| *** zhangjn has joined #openstack-cinder | 16:38 | |
| *** alonma has quit IRC | 16:38 | |
| *** alonma has joined #openstack-cinder | 16:40 | |
| *** edtubill has quit IRC | 16:41 | |
| *** hemnafk is now known as hemna | 16:42 | |
| *** edtubill has joined #openstack-cinder | 16:42 | |
| *** alonma has quit IRC | 16:44 | |
| *** sombrafam has joined #openstack-cinder | 16:45 | |
| *** alonma has joined #openstack-cinder | 16:46 | |
| *** itlinux has joined #openstack-cinder | 16:48 | |
| hemna | smcginnis, ping | 16:48 |
| *** vivekd has quit IRC | 16:48 | |
| *** zhangjn has quit IRC | 16:51 | |
| *** alonma has quit IRC | 16:52 | |
| *** alonma has joined #openstack-cinder | 16:52 | |
| *** vilobhmm11 has joined #openstack-cinder | 16:53 | |
| vilobhmm11 | rdmcnair : ping | 16:53 |
| *** vivekd has joined #openstack-cinder | 16:54 | |
| *** alonma has quit IRC | 16:57 | |
| *** alonma has joined #openstack-cinder | 16:59 | |
| *** vilobhmm11 has quit IRC | 17:00 | |
| *** diogogmt has quit IRC | 17:00 | |
| *** haomaiwang has quit IRC | 17:01 | |
| *** haomaiwang has joined #openstack-cinder | 17:01 | |
| *** diogogmt has joined #openstack-cinder | 17:02 | |
| *** alonma has quit IRC | 17:03 | |
| smcginnis | sheel: Sorry, was away. Yeah, looks like that's the root cause. | 17:03 |
| smcginnis | hemna: pong | 17:03 |
| *** wilson_liu has quit IRC | 17:03 | |
| *** lpetrut has quit IRC | 17:03 | |
| *** salv-orl_ has quit IRC | 17:04 | |
| *** salv-orlando has joined #openstack-cinder | 17:04 | |
| sheel | smcginnis: its ok...So, its ok to include "bootable" as default filter ? | 17:05 |
| hemna | smcginnis, any updates on the os-brick release ? | 17:05 |
| *** laughterwym has quit IRC | 17:05 | |
| sheel | so that non admin user can also get accurate entry | 17:05 |
| *** alonma has joined #openstack-cinder | 17:07 | |
| *** crose has quit IRC | 17:08 | |
| smcginnis | sheel: I believe so. I don't see any reason why non-admin users shouldn't be able to do that. | 17:08 |
| smcginnis | hemna: Not yet. Let me look at the latest. | 17:09 |
| sheel | smcginnis: yes... | 17:09 |
| sheel | smcginnis:thank you for your prompt support | 17:09 |
| smcginnis | sheel: Thanks for looking into the fix! | 17:09 |
| sheel | smcginnis : :) | 17:09 |
| smcginnis | hemna: Just a +1 from mriedem: https://review.openstack.org/#/c/267722/ | 17:09 |
| smcginnis | hemna: I'll go ping some folks. | 17:10 |
| hemna | ok thanks | 17:10 |
| *** alonma has quit IRC | 17:12 | |
| *** e0ne has quit IRC | 17:13 | |
| smcginnis | hemna: Shot down - no releases on Friday unless absolutely necessary. | 17:13 |
| smcginnis | hemna: Looks like it will have to sit until Monday. | 17:13 |
| hemna | wah wah wahhhhh | 17:13 |
| *** apoorvad has joined #openstack-cinder | 17:13 | |
| *** cbader has quit IRC | 17:14 | |
| *** alonma has joined #openstack-cinder | 17:14 | |
| *** vgridnev has joined #openstack-cinder | 17:16 | |
| knomura | smcginnis: Hi Sean, I open new blueprints for supporting CG and replication in HBSD driver. Could you approve these new blueprints? https://blueprints.launchpad.net/cinder/+spec/hbsd-add-consistency-groups https://blueprints.launchpad.net/cinder/+spec/hbsd-add-v2-replication | 17:18 |
| *** alonma has quit IRC | 17:18 | |
| smcginnis | knomura: Sure thing. | 17:19 |
| knomura | smcginnis: Thanks. | 17:19 |
| Thelo | DuncanT: Would you mind reviewing again? https://review.openstack.org/#/c/253356/ | 17:20 |
| smcginnis | Thelo: Since you just removed the depends-on line I just reapproved it. | 17:21 |
| *** alonma has joined #openstack-cinder | 17:21 | |
| smcginnis | Not that others can't still take a look, but not strictly necessary at this point. | 17:21 |
| asselin_ | Thelo, thanks for the write-up. Any objection to making it public? | 17:23 |
| smcginnis | jungleboyj: Old man. | 17:23 |
| *** leeantho has joined #openstack-cinder | 17:23 | |
| Thelo | asselin_: no problem, there might be typos though | 17:23 |
| *** raildo-afk is now known as raildo | 17:24 | |
| Thelo | smcginnis: I still need two +2 to get in the workflow no? | 17:24 |
| *** raildo is now known as raildo-afk | 17:25 | |
| openstackgerrit | Anthony Lee proposed openstack/cinder: LeftHand: Implement un/manage snapshot support https://review.openstack.org/255015 | 17:25 |
| jungleboyj | smcginnis: Look who's talking! | 17:25 |
| *** alonma has quit IRC | 17:25 | |
| smcginnis | Thelo: No, I gave +A. Nothing will hold it back pending another one. | 17:25 |
| smcginnis | jungleboyj: Hehe | 17:25 |
| openstackgerrit | Anthony Lee proposed openstack/cinder: LeftHand: Updating minimum client version https://review.openstack.org/267780 | 17:25 |
| smcginnis | jungleboyj: Very true though. :] | 17:25 |
| jungleboyj | smcginnis: I am still young at heart. | 17:26 |
| *** mdenny has quit IRC | 17:26 | |
| smcginnis | jungleboyj: Hah! | 17:27 |
| jungleboyj | smcginnis: You know it is true. | 17:27 |
| smcginnis | jungleboyj: Know it and have seen it! Haha | 17:28 |
| jungleboyj | :-) | 17:28 |
| Thelo | smcginnis:I see, thanks. So now we need to wait until monday for the new os-brick release, am I correct? One more question, If this patch is accepted and merged, Do I need to change the blueprint status? | 17:29 |
| *** mudassirlatif has joined #openstack-cinder | 17:30 | |
| smcginnis | Thelo: No, it should be good. With the "Implements blueprint" line there it _should_ automatically update the blueprint when the patch merges. | 17:32 |
| openstackgerrit | Yuriy Nesenenko proposed openstack/cinder: Port drivers/netapp/eseries/test_library.py to Python 3 https://review.openstack.org/267656 | 17:32 |
| smcginnis | Thelo: And then hopefully the os-brick release makes it through on Monday and you can update your CI to not have to pull in brick from github. | 17:33 |
| Swanson | asselin_, I'll take any additional notes right now as I'm CIing it again today. | 17:33 |
| *** yuriy_n17 has quit IRC | 17:34 | |
| Thelo | smcginnis: and I can try to get my nova patch merged.. because having everything except that patch would just ruined everything | 17:35 |
| smcginnis | Thelo: Yes, I agree. At least you have the cinder side in place, but it won't do much good until the nova side makes it too. | 17:35 |
| *** jordanP has quit IRC | 17:35 | |
| smcginnis | Unfortunately that's been a theme lately. | 17:35 |
| openstackgerrit | Alexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver. https://review.openstack.org/190273 | 17:36 |
| diablo_rojo | smcginnis: Might be good to get some eyes on this spec I mentioned the other day during the meeting since it will effect the cinderclient https://review.openstack.org/#/c/236712/2 | 17:38 |
| smcginnis | diablo_rojo: Oh right, thanks for the reminder. | 17:40 |
| diablo_rojo | smcginnis: No problem :) | 17:40 |
| *** crose has joined #openstack-cinder | 17:40 | |
| diablo_rojo | smcginnis: This one too https://review.openstack.org/#/c/226157/ | 17:42 |
| smcginnis | diablo_rojo: Should you be taking jungleboyj out somehere? | 17:42 |
| smcginnis | :P | 17:42 |
| diablo_rojo | smcginnis: Yes. Its a surprise. We are leaving at 1 so we can drink the day away. | 17:42 |
| diablo_rojo | smcginnis: Trying to get the guys to work up in the cities next Wed if you wanted to hang with us :) Would set up camp in some hipster coffee shop and go to Surly for lunch? | 17:43 |
| smcginnis | diablo_rojo: Ooo, Cinder hackathon! | 17:44 |
| *** chhavi has joined #openstack-cinder | 17:44 | |
| jungleboyj | smcginnis: I will still accept birthday gifts next week as well. | 17:44 |
| jungleboyj | ;-) | 17:44 |
| diablo_rojo | smcginnis: So you're interested? | 17:45 |
| smcginnis | jungleboyj: I'll buy you dinner Tuesday night at the meetup. (via Pure) :P | 17:45 |
| smcginnis | diablo_rojo: Yeah, definitely! | 17:45 |
| diablo_rojo | smcginnis: Coolio :) | 17:45 |
| *** shakamunyi has joined #openstack-cinder | 17:46 | |
| *** edtubill has quit IRC | 17:47 | |
| smcginnis | diablo_rojo: Just let me know where you guys end up. | 17:47 |
| jungleboyj | smcginnis: :=p | 17:47 |
| *** knomura has quit IRC | 17:47 | |
| diablo_rojo | smcginnis: Thinking Claddagh Coffee on West 7th, but I know of a good one in Northeast if you don't wanna go to St. Paul. | 17:48 |
| diablo_rojo | smcginnis: Spyhouse on Broadway | 17:48 |
| smcginnis | diablo_rojo: Been to spyhouse a bunch. Never been to claddagh. Either works. I'll ping you Tuesday night or Wednesday morning. | 17:49 |
| diablo_rojo | smcginnis: Sounds good. | 17:49 |
| openstackgerrit | Eric Harney proposed openstack/cinder: Add missing requirements https://review.openstack.org/268254 | 17:51 |
| openstackgerrit | Eric Harney proposed openstack/cinder: Add pip-missing-reqs tox env https://review.openstack.org/268255 | 17:51 |
| *** jamielennox is now known as jamielennox|away | 17:52 | |
| *** ericksonsantos has quit IRC | 17:53 | |
| openstackgerrit | Helen Walsh proposed openstack/cinder: EMC VMAX - VMAX driver failing to remove zones https://review.openstack.org/244933 | 17:53 |
| *** cknight has quit IRC | 17:53 | |
| asselin_ | Swanson, Thelo https://etherpad.openstack.org/p/thirdpartyci | 17:55 |
| *** edtubill has joined #openstack-cinder | 17:57 | |
| *** bardia has joined #openstack-cinder | 17:57 | |
| mc_nair | vilobhmm11 - looks like you were trying to ping me before? | 17:57 |
| openstackgerrit | Alexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver. https://review.openstack.org/190273 | 17:57 |
| *** cknight has joined #openstack-cinder | 17:57 | |
| *** ntpttr has joined #openstack-cinder | 17:57 | |
| *** harlowja_at_home has joined #openstack-cinder | 17:57 | |
| asselin_ | Swanson, Thelo fyi I added that etherpad also to here: https://wiki.openstack.org/wiki/Meetings/ThirdParty#Reference_links | 17:58 |
| *** fthiagogv has joined #openstack-cinder | 18:00 | |
| *** e0ne has joined #openstack-cinder | 18:00 | |
| *** haomaiwang has quit IRC | 18:01 | |
| *** haomaiwang has joined #openstack-cinder | 18:01 | |
| *** ericksonsantos has joined #openstack-cinder | 18:05 | |
| *** mriedem has quit IRC | 18:05 | |
| *** laughterwym has joined #openstack-cinder | 18:06 | |
| *** edtubill has quit IRC | 18:06 | |
| xyang1 | hemna: rhagarty_ hi, do you want to talk about the force flag | 18:06 |
| *** hyakuhei has quit IRC | 18:08 | |
| xyang1 | hemna: rhagarty_ actually create cgsnapshot is already using the force flag. I looked at the wrong place the other day. so no need to change anything now | 18:09 |
| *** laughterwym has quit IRC | 18:10 | |
| *** vivekd has quit IRC | 18:11 | |
| *** leeantho has quit IRC | 18:11 | |
| *** mriedem has joined #openstack-cinder | 18:12 | |
| *** leeantho has joined #openstack-cinder | 18:13 | |
| *** bardia has quit IRC | 18:13 | |
| rhagarty_ | xyang1, I think that will work for the Horizon use case. I may attempt to warn users in the "create snapshot" diaglog if any volumes are attached | 18:18 |
| xyang1 | rhagarty_: ok | 18:19 |
| rhagarty_ | hemna, thoughts? | 18:19 |
| *** alonma has joined #openstack-cinder | 18:21 | |
| rhagarty_ | xyang1, so that leaves 3 issues... 1) exposing the CG create quota, 2) allow update of CG snapshot name and description, and 3) defaulting CG policies to enabled. That sound right to you? | 18:23 |
| xyang1 | rhagarty_: sure. the last one has to be discussed. I can't enable it yet | 18:24 |
| rhagarty_ | xyang1, #1 above is https://bugs.launchpad.net/cinder/+bug/1526923 | 18:24 |
| openstack | Launchpad bug 1526923 in Cinder "Consistency Group quota needs to be accessible" [Undecided,New] | 18:24 |
| rhagarty_ | ok | 18:24 |
| xyang1 | rhagarty_: ok, can you open a bug for #2 as well | 18:25 |
| rhagarty_ | xyang1, will do | 18:25 |
| xyang1 | rhagarty_: I'll bring up the default CG policy issue at the meetup | 18:25 |
| *** alonma has quit IRC | 18:25 | |
| rhagarty_ | xyang1, ok | 18:26 |
| *** alonma has joined #openstack-cinder | 18:27 | |
| *** anshul has quit IRC | 18:27 | |
| *** ntt_ has joined #openstack-cinder | 18:30 | |
| *** martyturner has quit IRC | 18:30 | |
| hemna | back | 18:30 |
| hemna | ok | 18:30 |
| hemna | so the regular snapshot api defaults the force flag to False | 18:30 |
| *** cknight has quit IRC | 18:31 | |
| hemna | and if you want to snapshot a volume that's attached, you must pass in the force flag True. | 18:31 |
| *** vilobhmm11 has joined #openstack-cinder | 18:31 | |
| hemna | we need to be consistent with that mechanism | 18:31 |
| hemna | that's all I'd like to make sure we have | 18:31 |
| *** martyturner has joined #openstack-cinder | 18:31 | |
| xyang1 | hemna: a snapshot in Cgsnapshot cannot be used by itself, so it is different | 18:32 |
| *** alonma has quit IRC | 18:32 | |
| xyang1 | hemna: it does not make sense to only allow cgsnapshot to be created when volume is available | 18:32 |
| hemna | I'm just saying that if we allow snapshotting CG's where the volumes are attached, the force flag should be defaulted off, and required to be passed in True | 18:33 |
| xyang1 | hemna: it is always true | 18:33 |
| *** jgregor has quit IRC | 18:33 | |
| rhagarty_ | and on the Horizon side, if we detect any attached volumes, we warn the user and set the force-flag to true. For CG snapshots, there is no force-flag available | 18:33 |
| *** jgregor has joined #openstack-cinder | 18:33 | |
| hemna | xyang1, that doesn't seem right. | 18:33 |
| hemna | you always allow snapshotting attached volumes in the CG ? | 18:33 |
| xyang1 | hemna: yes | 18:33 |
| hemna | that defintely can cause problems | 18:33 |
| openstackgerrit | Szymon Wróblewski proposed openstack/cinder: Conditional update support for model fields https://review.openstack.org/268281 | 18:33 |
| hemna | IO not flushed out to the array | 18:34 |
| hemna | not being there when the snapshot was taken. | 18:34 |
| *** alonma has joined #openstack-cinder | 18:34 | |
| hemna | that's bad mmmkay | 18:34 |
| xyang1 | hemna: it is not useful to create cgsnapshot for available volumes only | 18:34 |
| *** cbader has joined #openstack-cinder | 18:34 | |
| *** dslev has joined #openstack-cinder | 18:34 | |
| *** mdenny has joined #openstack-cinder | 18:34 | |
| hemna | that's fine | 18:34 |
| xyang1 | hemna: io will be frozen on the array when creating cgsnapshot | 18:34 |
| hemna | but don't default the force flag to True. | 18:35 |
| hemna | users need to know that it's a dangerous operation | 18:35 |
| *** ociuhandu has quit IRC | 18:35 | |
| xyang1 | hemna: it has to be true | 18:35 |
| hemna | I understand that you need it to be true for the call to work | 18:35 |
| hemna | that's not my problem | 18:35 |
| xyang1 | hemna: otherwise snapshot cannot be taken when inuse | 18:35 |
| *** cknight has joined #openstack-cinder | 18:35 | |
| hemna | yes I understand that | 18:35 |
| hemna | the problem is the assumption it makes to the user if they don't have to set the force flag to True | 18:36 |
| hemna | there will be IO that hasn't been flushed to the backend yet at the time they snapshot | 18:36 |
| xyang1 | hemna: snapshot is not taken individually | 18:36 |
| patrickeast | isn't the whole point of it being in a cg so that you can take in-use snapshots and have them not be borked? | 18:36 |
| xyang1 | patrickeast: yes | 18:36 |
| hemna | patrickeast, yes it is | 18:36 |
| hemna | the entire point of the force flag | 18:36 |
| patrickeast | yea so it should always force | 18:36 |
| hemna | is a warning to people snapshotting attached volumes | 18:36 |
| *** salv-orlando has quit IRC | 18:37 | |
| openstackgerrit | Ryan McNair proposed openstack/cinder: Make IBM drivers return snapshot object for cg ops https://review.openstack.org/268283 | 18:37 |
| hemna | because IO might not be flushed to the array from the host doing IO | 18:37 |
| xyang1 | hemna: so I suggest you give a warning for cgsnapshot | 18:37 |
| hemna | I'm saying our API should be consistent | 18:37 |
| *** salv-orlando has joined #openstack-cinder | 18:37 | |
| xyang1 | hemna: whenever you create cgsnapshot, it is always force | 18:37 |
| hemna | the reason we have a force flag at all is because of this IO flush issue and potential inconsistency with the point in time the snapshot is taken. | 18:37 |
| xyang1 | hemna: we are not using snapshot api | 18:37 |
| hemna | that same problem holds true for volumes in a cg | 18:38 |
| hemna | you are missing my point entirely. | 18:38 |
| hemna | *sigh* | 18:38 |
| xyang1 | hemna: cgsnapshot api is different from snapshot api | 18:38 |
| *** alonma has quit IRC | 18:38 | |
| hemna | nevermind | 18:38 |
| hemna | I give up | 18:38 |
| patrickeast | hemna: so i get what you are saying, but i think maybe its then a difference in what is expected of the cgsnapshot api | 18:38 |
| hemna | the expectation is that you want to snapshot a cg | 18:39 |
| hemna | for volumes in use | 18:39 |
| patrickeast | yea | 18:39 |
| hemna | otherwise a cg is kinda useless. | 18:39 |
| patrickeast | which implies there *might* be data that hasn't been flushed on the initiator | 18:39 |
| hemna | but that is completely irrelevant to the IO flush and consistency problem of snapshotting volumes that are in use. | 18:40 |
| hemna | regardless of cinder and it's api | 18:40 |
| *** ericksonsantos has quit IRC | 18:40 | |
| hemna | this is a problem regardless of cg or not | 18:40 |
| patrickeast | its a block level snapshot of the cinder volume, not like an app level thing where it can stop io from the db or whatever it is | 18:40 |
| hemna | volumes in a cg and out of a cg have the same exact problem. | 18:40 |
| *** ericksonsantos has joined #openstack-cinder | 18:40 | |
| rhagarty_ | I agree, to be consistent, both API's should have a force-flag. | 18:40 |
| hemna | patrickeast, exactly | 18:40 |
| hemna | which is why our API should be consistent | 18:41 |
| hemna | the force flag makes that consistent. | 18:41 |
| rhagarty_ | and both should fail for attached volumes unless the flag is true | 18:41 |
| xyang1 | no, cgsnapshot is useless if we do not allow inuse volume | 18:41 |
| hemna | yes | 18:41 |
| hemna | *sigh* | 18:41 |
| *** chhavi has quit IRC | 18:41 | |
| patrickeast | i get your point | 18:41 |
| hemna | I'm not sure xyang1 does | 18:41 |
| patrickeast | but i'm also concerned that we would add a flag that must always be used | 18:41 |
| patrickeast | like | 18:41 |
| patrickeast | its seems kind of wasteful | 18:42 |
| hemna | we have to do it for normal volumes | 18:42 |
| xyang1 | hemna: the flag has to set to true always | 18:42 |
| rhagarty_ | on the Horizon side, we warn users, then set it to true for them. | 18:42 |
| hemna | *smh* | 18:42 |
| hemna | seriously, I give up | 18:42 |
| patrickeast | lol | 18:42 |
| xyang1 | rhagarty_: for cgsnapshot, always warn users | 18:43 |
| xyang1 | rhagarty_: because force is always true | 18:43 |
| rhagarty_ | yes, that would match behavior with normal snapshots. At least for Horizon users | 18:43 |
| rhagarty_ | hemna, would a user ever decide not to do the snapshot in this case? | 18:44 |
| hemna | rhagarty_, a user could decide to pause the io, before doing the snapshot to make sure the data is flushed to the volumes. | 18:45 |
| *** hyakuhei has joined #openstack-cinder | 18:46 | |
| hemna | ideally, we could eventually add a call to nova to pause IO on the vms to coordinate the snapshotting. | 18:46 |
| patrickeast | wasn't there a spec or something for that? hypervisor assisted snapshots or somethin | 18:47 |
| patrickeast | hard to find them without knowing the name in launchpad :( | 18:47 |
| xyang1 | patrickeast: I think that is different | 18:47 |
| hemna | I think so, I'm not sure if it works though without having a guest agent | 18:47 |
| *** alonma has joined #openstack-cinder | 18:47 | |
| patrickeast | hmm yea thats true | 18:47 |
| hemna | :( | 18:47 |
| xyang1 | patrickeast, hemna nova already has api for that | 18:48 |
| xyang1 | patrickeast: hemna but not external yet | 18:48 |
| rhagarty_ | hemna, so this is a CLI issue. I think the Horizon use case is acceptable. | 18:48 |
| hemna | xyang1, but without a guest agent, it doesn't really work as expected from what I understand. | 18:48 |
| *** lpetrut has joined #openstack-cinder | 18:48 | |
| xyang1 | hemna: you are right | 18:48 |
| rhagarty_ | xyang1, here is the defect you requested.. https://bugs.launchpad.net/cinder/+bug/1534781 | 18:50 |
| openstack | Launchpad bug 1534781 in Cinder "Consistency Group snapshot name and description should be editable" [Undecided,New] | 18:50 |
| xyang1 | rhagarty_: thanks | 18:50 |
| *** bardia has joined #openstack-cinder | 18:51 | |
| *** alonma has quit IRC | 18:52 | |
| *** ericksonsantos has quit IRC | 18:53 | |
| *** alonma has joined #openstack-cinder | 18:53 | |
| *** lpetrut has quit IRC | 18:56 | |
| kmartin | Happy Birthday jungleboyj !!! | 18:57 |
| jungleboyj | kmartin: Thank you! | 18:57 |
| kmartin | jungleboyj, I owe you a beer in NC | 18:58 |
| *** alonma has quit IRC | 18:58 | |
| *** martyturner has quit IRC | 18:59 | |
| *** haomaiwang has quit IRC | 19:01 | |
| *** haomaiwa_ has joined #openstack-cinder | 19:01 | |
| *** hyakuhei has quit IRC | 19:01 | |
| *** hyakuhei has joined #openstack-cinder | 19:02 | |
| *** hyakuhei has quit IRC | 19:02 | |
| *** savihou has joined #openstack-cinder | 19:02 | |
| *** rlrossit has quit IRC | 19:02 | |
| *** sgotliv has joined #openstack-cinder | 19:04 | |
| *** crose has quit IRC | 19:05 | |
| ntt_ | Hi, I'm trying to realize a configuration for cinder high availability and I need some advises. Actually I'm using a san that exposes one large disk to openstack-cinder-volume. On that disk I create a pv and then a vg named cinder-volumes. My san is already high available, but how can I manage the case when the openstack-cinder-volume node goes down? Is there a canonical solution for this? | 19:07 |
| *** rlrossi__ has joined #openstack-cinder | 19:07 | |
| *** rlrossi__ has quit IRC | 19:07 | |
| ntt_ | My idea is to start openstack-cinder-volume inside the san. In this way I can create an high available cinder-volumes VG, but failover seems not working correctly. | 19:08 |
| vilobhmm11 | jungleboyj : Happy Birthday ! | 19:09 |
| vilobhmm11 | jungleboyj : qq on https://bugs.launchpad.net/cinder/+bug/1531502 is there a WIP or patch out there or someone working on this one | 19:10 |
| openstack | Launchpad bug 1531502 in Cinder "Child project's default quota not enforced" [High,New] - Assigned to Ryan McNair (rdmcnair) | 19:10 |
| vilobhmm11 | smcginnis : ^^ | 19:10 |
| *** fthiagogv has quit IRC | 19:10 | |
| alkhodos | Hi. Is there anything I can do with this neutron test failure? https://jenkins07.openstack.org/job/gate-tempest-dsvm-neutron-full/21178/console Not even sure how to debug this | 19:11 |
| mc_nair | vilobhmm11: hey - I'm working on that currently | 19:13 |
| *** ntpttr has left #openstack-cinder | 19:13 | |
| mc_nair | vilobhmm11: basically the best options would be adding parent_project_id to the Keystone token, or allowing a project to be retrieve by a user of the project instead of just admin. Actually about to talk to bring that up in the Keystone channel (and we added it to their next meeting as well) | 19:14 |
| *** thangp has left #openstack-cinder | 19:16 | |
| openstackgerrit | Michal Dulko proposed openstack/cinder: OVO request_spec in migrate/manage/retype calls https://review.openstack.org/252458 | 19:16 |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Switch request_spec in create volume calls to ovo https://review.openstack.org/246600 | 19:16 |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Switch request_spec_list in create CG to OVO https://review.openstack.org/253033 | 19:16 |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Convert filter_properties to OVO (create, retype) https://review.openstack.org/257305 | 19:16 |
| mtanino | ntt_: I think DRBD is one of a solution for your situation. http://www.slideshare.net/mail4visu/disaster-recovery-of-openstack-cinder-using-drbd | 19:18 |
| vilobhmm11 | mc_nair : why can't this be done 1. change to admin context and try fetching the heirachy in create vol request. | 19:20 |
| mtanino | ntt_: Also C-Vol Active/Active HA might help too. But this is ongoing feature. https://etherpad.openstack.org/p/mitaka-cinder-spec-review-tracking | 19:20 |
| *** martyturner has joined #openstack-cinder | 19:21 | |
| *** rlrossit has joined #openstack-cinder | 19:21 | |
| vilobhmm11 | mc_nair : you there ? | 19:23 |
| mc_nair | vilobhmm11: switch to admin context as in do something like context.get_admin_context? I tried that initially, don't think that will help us be able to grab things from Keystone, just flips the is_admin flag true | 19:23 |
| vilobhmm11 | mc_nair : the requirement here is the user should be able to get quota values for a child project right ? | 19:24 |
| vilobhmm11 | a child project should be able to perform get | 19:25 |
| vilobhmm11 | of the quota values | 19:25 |
| mc_nair | vilobhmm11: the requirement would that a user be able to do a get of their current project from Keystone | 19:25 |
| *** crose has joined #openstack-cinder | 19:26 | |
| mc_nair | currently there's no way to get to the "parent_project_id" reliably, so in the case of child prject default enforcement we don't know whether to enforce child project defaults or non-child defaults | 19:26 |
| vilobhmm11 | not only info about the project but the quota values as well IMHO because lets say (B) is a child of A and intially quota of B is zero then we should prohitbit creating volumes inside B right ? | 19:27 |
| vilobhmm11 | mc_nair : i understand that | 19:27 |
| mc_nair | vilobhmm11: correct, but I don't think there's any issue with the case that the quota is set. Because the child project can get it's quota *if it's explicitly* been set | 19:28 |
| mc_nair | because it can retrieve that from Cinder | 19:28 |
| mc_nair | the only issue that I'm seeing here is with the default because that requires knowing about the current project heirarchy | 19:28 |
| ntt_ | mtanino: I'm interested in a active/passive solution, but first link is difficult to understand .... have you more details about? | 19:28 |
| vilobhmm11 | mc_nair : do you see quota values for child project in cinder.quotas as 0 | 19:29 |
| vilobhmm11 | cinder.quotas should have values for resources for each project | 19:29 |
| mc_nair | vilobhmm11: are you talking about in the case the child project doesn't have quota set (i.e. defaults)? | 19:30 |
| vilobhmm11 | mc_nair : yes | 19:30 |
| vilobhmm11 | default case | 19:30 |
| *** ericksonsantos has joined #openstack-cinder | 19:32 | |
| vilobhmm11 | mc_nair : ^^ | 19:32 |
| mtanino | ntt_: I haven't tried that configuration. I guess some engineers from DRBD might be in this channel. It's better to talk with them if you are interested in that configuratoin. | 19:32 |
| mc_nair | vilobhmm11: for the default case, I am getting 0's on the quota show. But not when it tries to grab the quota as part of the enforcement. | 19:32 |
| mc_nair | vilobhmm11: here running a few commands again as refresher.... | 19:33 |
| vilobhmm11 | mc_nair : what do you mean by "But not when it tries to grab the quota as part of the enforcement." | 19:33 |
| ntt_ | ok.... my san already uses drbd (and works well). With pacemaker/corosync it easy to make an HA volume group... so I think that cinder should not be so difficult | 19:33 |
| vilobhmm11 | so if the quota for default case is 0 then why can't we add a validation in the volume create path to avoid running into this case | 19:33 |
| vilobhmm11 | mc_nair : ^^ | 19:34 |
| mc_nair | vilobhmm11: so I'm double checking that "quota show" will always 0 for the default cause I don't remember anymore offhand. It may have been that only returned 0 when you were using the admin context. | 19:35 |
| vilobhmm11 | mc_nair : can you please confirm ? in either case whether admin or non-admin we can get the quota values and then do a validation check that if zero don't let the volume creation proceed till valid quota is set for the project ? | 19:36 |
| vilobhmm11 | thoughts ? | 19:37 |
| *** jistr has quit IRC | 19:37 | |
| *** erlon has joined #openstack-cinder | 19:37 | |
| mc_nair | vilobhmm11: give me like 15 minutes. I'm trying to get Keystone happy again cause I was messing with it, and then I'll let you know the exact behavior and we can decide how to move forward | 19:39 |
| *** merooney has quit IRC | 19:39 | |
| vilobhmm11 | mc_nair : sounds good | 19:39 |
| mtanino | ntt_: ah, you already have drbd configuration in your san. nice | 19:40 |
| ntt_ | yes..... actually the san has drbd resource -> LVM RA with pacemaker -> iscsi target RA -> iscsilun RA | 19:41 |
| *** ebalduf has joined #openstack-cinder | 19:41 | |
| ntt_ | My idea is to create another LVM RA with a volume called cinder-volumes. In this way the cinder-volumes VG is HA | 19:41 |
| ntt_ | but the problem is that I need to make openstack-cinder-volume HA too | 19:42 |
| *** merooney has joined #openstack-cinder | 19:43 | |
| openstackgerrit | Vincent Hou proposed openstack/cinder: Storwize: Implement v2 replication (split IO) https://review.openstack.org/237387 | 19:45 |
| *** merooney has quit IRC | 19:45 | |
| *** ntt_ has quit IRC | 19:45 | |
| openstackgerrit | Andrew Kerr proposed openstack/cinder: Convert osapi_volume_ext_list to ListOpt https://review.openstack.org/268306 | 19:47 |
| *** martyturner has quit IRC | 19:49 | |
| *** merooney has joined #openstack-cinder | 19:50 | |
| *** sombrafam has quit IRC | 19:52 | |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Convert filter_properties to OVO (create, retype) https://review.openstack.org/257305 | 19:53 |
| *** mudassirlatif has quit IRC | 19:53 | |
| mc_nair | vilobhmm11: ok - just confirmed the issue. Even quota show will give different things for child default value depending on whether you use an admin or non-admin user | 19:53 |
| *** sgotliv has quit IRC | 19:53 | |
| *** alonma has joined #openstack-cinder | 19:54 | |
| vilobhmm11 | mc_nair : can you please paste the output here http://paste.openstack.org/ | 19:55 |
| *** mudassirlatif has joined #openstack-cinder | 19:55 | |
| vilobhmm11 | child default values for admin vs non-admin | 19:55 |
| mc_nair | which makes sense because "quota-show" is an admin-or-owner action, but Keystone project get is an admin only action | 19:55 |
| mc_nair | sure | 19:55 |
| *** mudassirlatif has quit IRC | 19:55 | |
| *** hyakuhei has joined #openstack-cinder | 19:56 | |
| patrickeast | hemna: did your package mirror ever finish downloading? | 19:56 |
| mc_nair | vilobhmm11: http://paste.openstack.org/show/484038/ | 19:57 |
| *** alonma has quit IRC | 19:58 | |
| *** crose has quit IRC | 20:00 | |
| hemna | patrickeast, yup | 20:00 |
| hemna | then the network on my blade died | 20:00 |
| hemna | :P | 20:00 |
| hemna | but I think I have it dialed | 20:00 |
| *** haomaiwa_ has quit IRC | 20:01 | |
| *** alonma has joined #openstack-cinder | 20:01 | |
| hemna | I was able to do an apt-get update against it | 20:01 |
| hemna | and it worked | 20:01 |
| hemna | once I get it finished, I can share my setup | 20:01 |
| *** 5EXAAQR7M has joined #openstack-cinder | 20:01 | |
| patrickeast | hemna: sweet | 20:01 |
| *** lcurtis_ has joined #openstack-cinder | 20:02 | |
| patrickeast | hemna: i was inspired by that and started playing around with some stuff too, stood up a cache server to see how much if helps | 20:02 |
| patrickeast | hemna: was going to wait to hear how well the mirror worked before downloading my own ;) | 20:02 |
| hemna | so debmirror is garbage | 20:02 |
| patrickeast | oh yea? | 20:02 |
| hemna | don't even bother | 20:02 |
| hemna | it's awful | 20:02 |
| patrickeast | i did apt-cacher-ng behind a squid proxy/cache | 20:03 |
| hemna | single threaded and sequential | 20:03 |
| patrickeast | and devpi also behind squid | 20:03 |
| hemna | takes for-effing-ever | 20:03 |
| patrickeast | ouch | 20:03 |
| hemna | apt-mirror is what I've been playing with | 20:03 |
| hemna | it fires up 20 threads and saturates the ports | 20:03 |
| hemna | :) | 20:03 |
| patrickeast | haha perfect | 20:04 |
| mc_nair | vilobhmm11: look, this right here basically explains why child project default quotas are busted currently - https://github.com/openstack/cinder/blob/master/cinder/api/contrib/quotas.py#L245-L259 | 20:04 |
| vilobhmm11 | mc_nair : non-admin should not be allowed to do any quota-* ops check here http://docs.openstack.org/user-guide-admin/cli_set_compute_quotas.html they are meant only for *admins* | 20:05 |
| *** alonma has quit IRC | 20:05 | |
| vilobhmm11 | mc_nair : so we should only have values returned by admin user | 20:06 |
| vilobhmm11 | which http://paste.openstack.org/show/484038/ seem to be correct as default values are 0 | 20:06 |
| mc_nair | vilobhmm11: https://github.com/openstack/cinder/blob/master/etc/cinder/policy.json#L37 - by default we let admin or member of the current project do them | 20:06 |
| vilobhmm11 | mc_nair : ok | 20:07 |
| *** laughterwym has joined #openstack-cinder | 20:07 | |
| vilobhmm11 | https://github.com/openstack/cinder/blob/master/cinder/api/contrib/quotas.py#L245-L259 what was your point again ? | 20:07 |
| vilobhmm11 | mc_nair : ^^ | 20:07 |
| mc_nair | vilobhmm11: regardless, the fact that we only let the explicit "quota.*" commands be admin is just masking the real issue.... on other flows, e.g. create_volume, we need to enforce quotas. Those are not admin only commands | 20:07 |
| *** aix has quit IRC | 20:08 | |
| *** alonma has joined #openstack-cinder | 20:08 | |
| *** esker has joined #openstack-cinder | 20:09 | |
| *** esker has quit IRC | 20:09 | |
| *** martyturner has joined #openstack-cinder | 20:09 | |
| akerr | patrickeast: I use apt-cacher behind our corp firewall | 20:09 |
| mc_nair | Keystone by only lets get_project for admin users. So we don't currently have a way to get parent_project_id for any non-admin actions | 20:09 |
| akerr | patrickeast: also use devpi for pip caching | 20:09 |
| patrickeast | akerr: do they work pretty well? | 20:09 |
| akerr | patrickeast: haven't had any issues with them | 20:09 |
| *** mudassirlatif has joined #openstack-cinder | 20:09 | |
| patrickeast | i've just got them working, waiting for a nodepool image to build with confs baked in | 20:10 |
| akerr | patrickeast: but preinstalling the pip requirements for the main projects really sped up my devstack builds | 20:10 |
| *** mudassirlatif has quit IRC | 20:10 | |
| patrickeast | akerr: yea, iirc the upstream infra scripts do that for the images too | 20:11 |
| patrickeast | i should check to make sure thats actually working in mine... | 20:11 |
| *** laughterwym has quit IRC | 20:12 | |
| *** esker has joined #openstack-cinder | 20:12 | |
| akerr | patrickeast: i know they preload the repos but I don't think they install the requirements.txt files. I limited it to just nova, cinder, glance, neutron, and manila on our nodes | 20:12 |
| patrickeast | akerr: oo interesting | 20:13 |
| *** alonma has quit IRC | 20:13 | |
| patrickeast | akerr: don't suppose you've got a script for that around somewhere i can look at? | 20:13 |
| akerr | patrickeast: let me see if i can find it | 20:13 |
| *** alonma has joined #openstack-cinder | 20:14 | |
| openstackgerrit | Gorka Eguileor proposed openstack/cinder: Remote unused iscsi_targets table https://review.openstack.org/268320 | 20:15 |
| mc_nair | vilobhmm11: that make any sense? | 20:17 |
| *** alonma has quit IRC | 20:19 | |
| *** alonma has joined #openstack-cinder | 20:21 | |
| akerr | patrickeast: so i'm not sure how out of sync my config project is from the current ones, but I modified the cache_gi_repos.py file under the nodepool scripts: http://paste.openstack.org/show/484042/ | 20:22 |
| mc_nair | so I mean the easiest fix would just be to change https://github.com/openstack/keystone/blob/master/etc/policy.json#L37 to "admin_or_owner". But I don't know if that'd make people sad, going to ask keystone | 20:22 |
| akerr | patrickeast: primarily I changed what file I was reading to know which projects to cache and then I added the "install_deps" function | 20:23 |
| vilobhmm11 | mc_nair : sure..sorry someone stopped by | 20:23 |
| *** e0ne has quit IRC | 20:23 | |
| patrickeast | akerr: awesome, thanks! | 20:23 |
| vilobhmm11 | mc_nair : how about create_volume => quota_call => change user context admin context => get_projects | 20:23 |
| *** mudassirlatif has joined #openstack-cinder | 20:24 | |
| *** dslev has quit IRC | 20:24 | |
| akerr | patrickeast: that shaved about 10-15 minutes off our runs | 20:25 |
| patrickeast | akerr: woa thats some serious improvements | 20:25 |
| *** alonma has quit IRC | 20:26 | |
| *** [1]Thelo has joined #openstack-cinder | 20:26 | |
| *** mudassirlatif has quit IRC | 20:26 | |
| *** e0ne has joined #openstack-cinder | 20:26 | |
| openstackgerrit | Sheel Rana proposed openstack/cinder: Added 'bootable volume' filter for non-admin user https://review.openstack.org/268322 | 20:27 |
| *** alonma has joined #openstack-cinder | 20:27 | |
| mc_nair | vilobhmm11: so when you're saying change to admin context, we need to be a user with admin priveledges I believe. So DuncanT had mentioned that it may be possible to fix this using the Cinder service role (though not ideal). I hacked up a WIP along those lines here - https://review.openstack.org/#/c/267848/ | 20:28 |
| *** Thelo has quit IRC | 20:28 | |
| *** [1]Thelo is now known as Thelo | 20:28 | |
| vilobhmm11 | mc_nair : why can't we do context_elevated () | 20:29 |
| mc_nair | which works so long as the Cinder user is added to "service" project as an admin role (which it wasn't in devstack for some reason) | 20:29 |
| mc_nair | vilobhmm11: so I'm still reasonably unfamiliar but my understanding is that just flips some the "is_admin" flag in the Cinder RequestContext, so it lets you do *Cinder* admin only actions but admin actions in Keystone | 20:30 |
| mc_nair | but I'm sure others could give better perspective on that. My experience with that was "I tried that and it didn't work, maybe this is why..." | 20:31 |
| *** alonma has quit IRC | 20:33 | |
| vilobhmm11 | mc_nair : IMHO we need admin priviledge to get quota values (default) for the child project and this should help us | 20:33 |
| vilobhmm11 | smcginnis : ^^ | 20:33 |
| *** ericksonsantos has quit IRC | 20:33 | |
| vilobhmm11 | jungleboyj :^^ | 20:33 |
| *** alonma has joined #openstack-cinder | 20:33 | |
| mc_nair | vilobhmm11: well we never needed admin priveledges before to get quota defaults (certainly not internally with flows like create volume). This is all coming out of the fact we need to ask keystone to fetch us the project | 20:35 |
| vilobhmm11 | mc_nair : yes ..if its was not needed before doesn;t mean we won't need it in future :) its just for getting quota details for rest operation we do as cinder user IMHO | 20:36 |
| *** ericksonsantos has joined #openstack-cinder | 20:36 | |
| *** alonma has quit IRC | 20:38 | |
| *** mudassirlatif has joined #openstack-cinder | 20:39 | |
| *** mudassirlatif has quit IRC | 20:40 | |
| mc_nair | vilobhmm11: but why wouldn't someone be able to see their current quota? | 20:41 |
| mc_nair | vilobhmm11: we would just say "You can't create a volume, not enough quota".... "Can't look at your quota to see what your maxes are"? | 20:41 |
| sheel | eharney: hi, you there? | 20:42 |
| eharney | sheel: hi | 20:43 |
| sheel | eharney: I wanted to talk about https://review.openstack.org/#/c/266036/... | 20:43 |
| sheel | eharney: it seems partially fixed... | 20:43 |
| *** alonma has joined #openstack-cinder | 20:43 | |
| vilobhmm11 | mc_nair : even the quota-show cli needs to change to make sure admin gets those details since _get_projects is admin onyl api | 20:43 |
| eharney | sheel: sure | 20:43 |
| sheel | eharney:please let me know if you are busy and need help to fix other APIs context | 20:44 |
| sheel | eharney:i tested your fix, it was working fine.. | 20:44 |
| mc_nair | vilobhmm11: but I disagree - I think that get_projects should change in Keystone to let you grab your current project. Then both problems are solved (don't have to make quota-show admin and the create_volume enforcement works) | 20:45 |
| *** dslevin has joined #openstack-cinder | 20:45 | |
| eharney | sheel: i don't intend to apply it to all the APIs myself as of now | 20:45 |
| vilobhmm11 | mc_nair : thats a long process but we should even fix it for people seeing it now | 20:46 |
| sheel | eharney:so should i go ahead with left APIs? | 20:46 |
| eharney | sheel: sure | 20:46 |
| eharney | sheel: i was trying to come up with a nice decorator or something to make it simple to apply to many of them, but didn't really get there | 20:46 |
| vilobhmm11 | mc_nair : and it will happen one day..I am not denying that its not a valid solution | 20:46 |
| sheel | eharney:ok, thank you... | 20:46 |
| sheel | eharney:i think fix is clean | 20:46 |
| eharney | sheel: it's not bad, just a little more verbose than i was hoping | 20:47 |
| sheel | eharney: right... | 20:47 |
| *** alonma has quit IRC | 20:47 | |
| sheel | eharney: so, if you have any suggestion or may be you can let me know what addtional things you want in fix.. | 20:48 |
| sheel | eharney: I will try accordingly | 20:49 |
| eharney | sheel: i don't really have anything as of now | 20:49 |
| mc_nair | vilobhmm11: so changing to admin only for quota-show is easy enough to fix. But the enforcement fix is tougher. We *could* do something with the service user role I think (see that WIP), but couldn't the easier fix for existing customers be change "policy.json" for Cinder to let "get_projects" be admin_or_owner (assuming Keystone folks don't think that's an issue to do)? | 20:49 |
| sheel | eharney: Ok, I think for now we can go ahead with same solution for all APIs | 20:50 |
| eharney | sheel: sounds good | 20:50 |
| sheel | eharney: Ok, thank you | 20:50 |
| vilobhmm11 | mc_nair : ok | 20:51 |
| vilobhmm11 | lemme check | 20:51 |
| sheel | eharney: I will be drafting BP-https://blueprints.launchpad.net/cinder/+spec/summarymessage | 20:52 |
| *** lcurtis_ has quit IRC | 20:52 | |
| sheel | eharney: will discuss with you on design around 25th.. | 20:52 |
| *** timcl has quit IRC | 20:53 | |
| *** sombrafam has joined #openstack-cinder | 20:53 | |
| sheel | eharney: please find some spare time, will seek some bandwidth of yours on design discussion | 20:53 |
| mc_nair | vilobhmm11: that's just my current take on things. It's been changing every day :) | 20:53 |
| eharney | sheel: ok, i'll be sure to be involved there. i suspect a handful of us already have ideas about what we want to see there | 20:54 |
| *** mudassirlatif has joined #openstack-cinder | 20:54 | |
| *** hyakuhei has quit IRC | 20:55 | |
| *** mudassirlatif has quit IRC | 20:55 | |
| sheel | eharney: yes, this is important improvement..so many of us would have though about... lets go ahead with this BP with combined ideas... | 20:55 |
| sheel | eharney: I have already given html tagging in BP itself, do have a look when you get some time.. | 20:56 |
| sheel | eharney: this will sync our ideas.. | 20:56 |
| eharney | sheel: ok | 20:56 |
| sheel | eharney: :) | 20:57 |
| *** apoorvad has quit IRC | 20:57 | |
| *** savihou has quit IRC | 20:58 | |
| *** savihou has joined #openstack-cinder | 20:58 | |
| openstackgerrit | Eric Harney proposed openstack/cinder: Set LVM driver default overprovisioning ratio to 1.0 https://review.openstack.org/266986 | 21:00 |
| *** 5EXAAQR7M has quit IRC | 21:01 | |
| *** haomaiwang has joined #openstack-cinder | 21:01 | |
| mc_nair | vilobhmm11: even if we changed that piece in Cinder, we'd have to update Cinder so that we're actually retrieving the project in the create_volume flow (currently missing) which I'll do once we hash the other parts of it out. Ideally we'd move parent_project_id to Keystone token so we don't have to query Keystone on each create_volume, but I think that'll get some pushback as there's a limit to how much you can put in there | 21:05 |
| *** hyakuhei has joined #openstack-cinder | 21:06 | |
| openstackgerrit | Sheel Rana proposed openstack/cinder: Added 'bootable volume' filter for non-admin user https://review.openstack.org/268322 | 21:07 |
| *** merooney has quit IRC | 21:08 | |
| openstackgerrit | Vincent Hou proposed openstack/cinder: Storwize: Implement v2 replication (mirror) https://review.openstack.org/249042 | 21:09 |
| *** mudassirlatif has joined #openstack-cinder | 21:09 | |
| openstackgerrit | Gorka Eguileor proposed openstack/cinder-specs: Job Distribution to support HA A/A https://review.openstack.org/232595 | 21:10 |
| *** cdelatte has quit IRC | 21:10 | |
| *** mudassirlatif has quit IRC | 21:10 | |
| *** martyturner1 has joined #openstack-cinder | 21:13 | |
| *** alkhodos has quit IRC | 21:13 | |
| *** sheel has quit IRC | 21:13 | |
| *** sheel has joined #openstack-cinder | 21:14 | |
| *** porrua has quit IRC | 21:15 | |
| *** gouthamr has quit IRC | 21:15 | |
| *** martyturner has quit IRC | 21:16 | |
| *** dslevin has quit IRC | 21:19 | |
| *** adrianofr has quit IRC | 21:20 | |
| *** edmondsw has joined #openstack-cinder | 21:20 | |
| *** edmondsw has quit IRC | 21:21 | |
| *** mudassirlatif has joined #openstack-cinder | 21:24 | |
| *** mudassirlatif has quit IRC | 21:25 | |
| *** martyturner1 has quit IRC | 21:26 | |
| *** apoorvad has joined #openstack-cinder | 21:29 | |
| *** laughterwym has joined #openstack-cinder | 21:29 | |
| *** salv-orlando has quit IRC | 21:29 | |
| *** salv-orlando has joined #openstack-cinder | 21:29 | |
| openstackgerrit | xing-yang proposed openstack/cinder: ScaleIO QoS Support https://review.openstack.org/260719 | 21:31 |
| *** Lee1092 has quit IRC | 21:32 | |
| *** gouthamr has joined #openstack-cinder | 21:33 | |
| *** daneyon has quit IRC | 21:34 | |
| *** laughterwym has quit IRC | 21:35 | |
| *** gouthamr_ has joined #openstack-cinder | 21:35 | |
| openstackgerrit | Sheel Rana proposed openstack/cinder: Added 'bootable volume' filter for non-admin user https://review.openstack.org/268322 | 21:36 |
| *** jungleboyj has quit IRC | 21:36 | |
| *** gouthamr has quit IRC | 21:39 | |
| *** mudassirlatif has joined #openstack-cinder | 21:39 | |
| *** baumann has quit IRC | 21:39 | |
| *** mudassirlatif has quit IRC | 21:40 | |
| *** cknight has quit IRC | 21:40 | |
| *** akerr has quit IRC | 21:41 | |
| *** alonma has joined #openstack-cinder | 21:44 | |
| *** savihou has quit IRC | 21:45 | |
| *** jgregor has quit IRC | 21:46 | |
| *** alonma has quit IRC | 21:48 | |
| *** diablo_rojo has quit IRC | 21:48 | |
| *** baumann has joined #openstack-cinder | 21:48 | |
| openstackgerrit | xing-yang proposed openstack/cinder: Manage/unmanage volume in ScaleIO driver https://review.openstack.org/268344 | 21:49 |
| openstackgerrit | xing-yang proposed openstack/cinder: Manage/unmanage volume in ScaleIO driver https://review.openstack.org/268344 | 21:50 |
| *** alonma has joined #openstack-cinder | 21:50 | |
| *** mragupat has quit IRC | 21:52 | |
| *** jamielennox|away is now known as jamielennox | 21:54 | |
| *** mudassirlatif has joined #openstack-cinder | 21:54 | |
| *** alonma has quit IRC | 21:55 | |
| *** mudassirlatif has quit IRC | 21:55 | |
| openstackgerrit | Eric Harney proposed openstack/cinder: Reduce use of eval() https://review.openstack.org/268345 | 21:55 |
| *** dustins has quit IRC | 21:55 | |
| openstackgerrit | Alex O'Rourke proposed openstack/cinder: Remove old client version checks from 3PAR driver https://review.openstack.org/268346 | 21:56 |
| *** alonma has joined #openstack-cinder | 21:56 | |
| *** tpsilva has quit IRC | 21:56 | |
| *** cdelatte has joined #openstack-cinder | 21:59 | |
| *** dims_ has joined #openstack-cinder | 21:59 | |
| *** cdelatte has quit IRC | 22:00 | |
| *** cdelatte has joined #openstack-cinder | 22:00 | |
| *** alonma has quit IRC | 22:00 | |
| *** haomaiwang has quit IRC | 22:01 | |
| *** haomaiwang has joined #openstack-cinder | 22:01 | |
| *** baumann has quit IRC | 22:01 | |
| *** dims has quit IRC | 22:01 | |
| *** alonma has joined #openstack-cinder | 22:03 | |
| *** sombrafam has quit IRC | 22:04 | |
| openstackgerrit | xing-yang proposed openstack/cinder: Manage/unmanage volume in ScaleIO driver https://review.openstack.org/268344 | 22:05 |
| *** e0ne has quit IRC | 22:08 | |
| *** dslevin has joined #openstack-cinder | 22:08 | |
| *** dslevin has quit IRC | 22:08 | |
| *** mudassirlatif has joined #openstack-cinder | 22:09 | |
| *** alonma has quit IRC | 22:09 | |
| *** mudassirlatif has quit IRC | 22:10 | |
| *** salv-orl_ has joined #openstack-cinder | 22:10 | |
| *** baumann has joined #openstack-cinder | 22:11 | |
| *** alonma has joined #openstack-cinder | 22:11 | |
| *** sheel has quit IRC | 22:12 | |
| *** salv-orlando has quit IRC | 22:13 | |
| *** alonma has quit IRC | 22:16 | |
| *** baumann has quit IRC | 22:16 | |
| *** alonma has joined #openstack-cinder | 22:18 | |
| *** eharney has quit IRC | 22:19 | |
| *** alonma has quit IRC | 22:22 | |
| *** mudassirlatif has joined #openstack-cinder | 22:24 | |
| *** alonma has joined #openstack-cinder | 22:24 | |
| *** mudassirlatif has quit IRC | 22:25 | |
| *** vgridnev has quit IRC | 22:26 | |
| *** vilobhmm11 has left #openstack-cinder | 22:27 | |
| *** alonma has quit IRC | 22:29 | |
| *** eharney has joined #openstack-cinder | 22:33 | |
| *** mudassirlatif has joined #openstack-cinder | 22:39 | |
| *** mudassirlatif has quit IRC | 22:40 | |
| *** eharney has quit IRC | 22:42 | |
| *** rlrossit has quit IRC | 22:44 | |
| *** julim has quit IRC | 22:45 | |
| *** mriedem has quit IRC | 22:47 | |
| *** mudassirlatif has joined #openstack-cinder | 22:54 | |
| *** IlyaG has joined #openstack-cinder | 22:54 | |
| *** mudassirlatif has quit IRC | 22:55 | |
| *** jiangliucheng has quit IRC | 23:00 | |
| *** haomaiwang has quit IRC | 23:01 | |
| *** haomaiwang has joined #openstack-cinder | 23:01 | |
| *** mudassirlatif has joined #openstack-cinder | 23:03 | |
| *** mudassirlatif has quit IRC | 23:10 | |
| *** hyakuhei has quit IRC | 23:11 | |
| *** IlyaG has quit IRC | 23:16 | |
| openstackgerrit | Apoorva Deshpande proposed openstack/cinder: Tintri image cache cleanup https://review.openstack.org/268370 | 23:19 |
| *** sheel has joined #openstack-cinder | 23:19 | |
| *** mdenny has quit IRC | 23:20 | |
| *** mudassirlatif has joined #openstack-cinder | 23:24 | |
| sheel | grenade test failed: tempest_lib.exceptions.SSHTimeout: Connection to the 10.1.0.3 via SSH timed out. | 23:24 |
| sheel | tempest.scenario.test_volume_boot_pattern.TestVolumeBootPattern.test_volume_boot_patter | 23:24 |
| *** alonma has joined #openstack-cinder | 23:25 | |
| *** mudassirlatif has quit IRC | 23:25 | |
| sheel | anyone have any idea what these are all about - ssh timedout? | 23:25 |
| sheel | http://logs.openstack.org/22/268322/3/check/gate-grenade-dsvm/de9b29b/logs/testr_results.html.gz | 23:25 |
| *** alonma has quit IRC | 23:29 | |
| *** dims_ has quit IRC | 23:29 | |
| *** xyang1 has quit IRC | 23:29 | |
| Swanson | It created an instance and tried to log in to it. ssh timed out. | 23:30 |
| sheel | Swanson: right, but how could i get to know if this is due to my changes or not | 23:31 |
| sheel | its a bit confusing to derive | 23:31 |
| sheel | :( | 23:31 |
| *** alonma has joined #openstack-cinder | 23:31 | |
| *** lcurtis has quit IRC | 23:31 | |
| sheel | Swanson: is this seems some general error ? | 23:32 |
| *** alonma has quit IRC | 23:35 | |
| *** dims has joined #openstack-cinder | 23:37 | |
| *** alonma has joined #openstack-cinder | 23:37 | |
| sheel | Swanson: By the way thanks for your prompt reply, I think best way for now to check is rerunning tests..(recheck) | 23:38 |
| sheel | running recheck | 23:38 |
| *** alonma has quit IRC | 23:42 | |
| *** alonma has joined #openstack-cinder | 23:43 | |
| *** IlyaG has joined #openstack-cinder | 23:44 | |
| *** garthb__ has quit IRC | 23:47 | |
| *** alonma has quit IRC | 23:48 | |
| *** alonma has joined #openstack-cinder | 23:50 | |
| *** alkhodos has joined #openstack-cinder | 23:51 | |
| *** mudassirlatif has joined #openstack-cinder | 23:52 | |
| *** alonma has quit IRC | 23:55 | |
| *** alonma has joined #openstack-cinder | 23:56 | |
| alkhodos | Hi guys, could you please review this patch https://review.openstack.org/#/c/190273 when you have time. I've addressed all previous comments and CI is running stable. Thanks in advance. | 23:58 |
| *** dims has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!