| openstackgerrit | Tom Barron proposed openstack/manila master: Why are m-api logs missing? https://review.openstack.org/352606 | 00:11 |
|---|---|---|
| *** vivsoni has quit IRC | 00:33 | |
| *** hongbin has quit IRC | 00:37 | |
| *** vivsoni has joined #openstack-manila | 00:39 | |
| *** harlowja has quit IRC | 01:08 | |
| *** YuYangWang has joined #openstack-manila | 01:25 | |
| *** YuYangWang has quit IRC | 01:28 | |
| *** YuYangWang has joined #openstack-manila | 01:28 | |
| *** gcb has joined #openstack-manila | 01:43 | |
| *** jiaohaolin has left #openstack-manila | 02:30 | |
| *** YuYangWang has quit IRC | 02:45 | |
| *** YuYangWang has joined #openstack-manila | 02:45 | |
| *** Manor has joined #openstack-manila | 02:49 | |
| *** Manor has quit IRC | 02:54 | |
| *** ociuhandu has joined #openstack-manila | 03:08 | |
| *** ociuhandu has quit IRC | 03:10 | |
| *** ociuhandu has joined #openstack-manila | 03:15 | |
| openstackgerrit | Tom Barron proposed openstack/manila master: coordination: use tooz builtin heartbeat feature https://review.openstack.org/460522 | 03:39 |
| openstackgerrit | Tom Barron proposed openstack/manila master: coordination: use tooz builtin heartbeat feature https://review.openstack.org/460522 | 03:45 |
| tbarron | ^^^ unit tests are failing w/o this due to tooz lib upgrade | 03:46 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/manila master: Updated from global requirements https://review.openstack.org/465261 | 03:49 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/manila-image-elements master: Updated from global requirements https://review.openstack.org/465262 | 03:49 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/manila-ui master: Updated from global requirements https://review.openstack.org/465263 | 03:49 |
| *** gouthamr has quit IRC | 03:57 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-manilaclient master: Updated from global requirements https://review.openstack.org/465366 | 03:57 |
| *** arnewiebalck__ has joined #openstack-manila | 04:11 | |
| *** harlowja has joined #openstack-manila | 04:33 | |
| *** markstur has quit IRC | 04:38 | |
| *** markstur has joined #openstack-manila | 04:39 | |
| *** markstur has quit IRC | 04:44 | |
| *** ociuhandu has quit IRC | 05:00 | |
| *** arnewiebalck__ has quit IRC | 05:17 | |
| *** zhugx has joined #openstack-manila | 05:45 | |
| *** pcaruana has joined #openstack-manila | 05:48 | |
| *** harlowja has quit IRC | 05:54 | |
| *** zhugx_ has joined #openstack-manila | 05:55 | |
| *** dsariel has joined #openstack-manila | 05:57 | |
| *** zhugx has quit IRC | 05:58 | |
| *** a-pugachev has joined #openstack-manila | 06:18 | |
| *** jprovazn has joined #openstack-manila | 06:19 | |
| *** nirg has joined #openstack-manila | 06:24 | |
| *** a-pugachev has quit IRC | 06:33 | |
| *** makowals has joined #openstack-manila | 06:47 | |
| *** Manor has joined #openstack-manila | 06:50 | |
| *** belmoreira has joined #openstack-manila | 06:51 | |
| *** Manor has quit IRC | 06:54 | |
| *** shausy has joined #openstack-manila | 06:59 | |
| *** nirg has quit IRC | 07:00 | |
| *** TD has joined #openstack-manila | 07:10 | |
| *** TD has quit IRC | 07:11 | |
| openstackgerrit | zhongjun proposed openstack/manila-specs master: Do not merge https://review.openstack.org/465436 | 07:40 |
| *** arnewiebalck__ has joined #openstack-manila | 07:45 | |
| *** arnewiebalck__ has quit IRC | 07:46 | |
| *** arnewiebalck__ has joined #openstack-manila | 07:47 | |
| *** arnewiebalck__ has quit IRC | 07:47 | |
| *** markstur has joined #openstack-manila | 07:48 | |
| zhongjun | tbarron: ping | 07:48 |
| *** markstur has quit IRC | 07:53 | |
| zhongjun | tbarron: Manila specs jenkins are failing(https://review.openstack.org/#/c/464964/1). After I change sphinx version back (sphinx>=1.1.2,!=1.2.0,!=1.3b1,<1.3), it is work(https://review.openstack.org/#/c/464964/1). But the project jenkins is success when you change sphinx version to >1.5.1(https://review.openstack.org/#/c/456398/) | 07:53 |
| zhongjun | tbarron: Could you see it? | 07:54 |
| *** lpetrut has joined #openstack-manila | 07:54 | |
| *** zhugx has joined #openstack-manila | 08:07 | |
| openstackgerrit | zhongjun proposed openstack/manila master: Add export-location filter in share and share instance list API https://review.openstack.org/461712 | 08:08 |
| jprovazn | zhongjun: hi, I think Tom has PTO | 08:08 |
| jprovazn | it seems all tests are failing due to https://launchpad.net/bugs/1691323 | 08:08 |
| openstack | Launchpad bug 1691323 in Manila "unit tests fail in gate due to tooz change" [Critical,In progress] - Assigned to Tom Barron (tpb) | 08:08 |
| jprovazn | (which Tom sent a patch earlier this morning | 08:08 |
| *** Manor has joined #openstack-manila | 08:09 | |
| jprovazn | I'm looking at it now, but if someone familiar with tooz is around and could take a look at the comment, it would be great | 08:09 |
| zhongjun | jprovazn: Manila specs project because of tooz change? | 08:09 |
| jprovazn | zhongjun: no, spec is not due to tooz I guess | 08:10 |
| *** zhugx_ has quit IRC | 08:10 | |
| zhongjun | jprovazn: Ok, two different things | 08:11 |
| jprovazn | zhongjun: yes, 2 different things, I considered manila tests as a highest prio so I mentioned it, but you are right it's irrelevant | 08:16 |
| zhongjun | jprovazn: yes, It is definitely as a highest prio. Do you know when he will come back. | 08:19 |
| openstackgerrit | zhongjun proposed openstack/manila-specs master: Do not merge https://review.openstack.org/465436 | 08:20 |
| *** makowals_ has joined #openstack-manila | 08:21 | |
| jprovazn | zhongjun: more than week IIRC, but since he sent a patch tonight, I wonder to what extent he is really resting :) | 08:21 |
| *** makowals has quit IRC | 08:22 | |
| zhongjun | jprovazn: :) Maybe he is missing here. | 08:24 |
| jprovazn | zhongjun: I bet he is ;) | 08:25 |
| zhongjun | jprovazn: haha | 08:25 |
| *** Manor has quit IRC | 08:27 | |
| *** Manor has joined #openstack-manila | 08:28 | |
| openstackgerrit | zhongjun proposed openstack/manila-specs master: Do not merge https://review.openstack.org/465436 | 08:29 |
| openstackgerrit | zhongjun proposed openstack/manila master: Add share usage size tracking https://review.openstack.org/465055 | 08:38 |
| openstackgerrit | Jan Provaznik proposed openstack/manila master: coordination: use tooz builtin heartbeat feature https://review.openstack.org/460522 | 08:39 |
| *** shausy has quit IRC | 08:41 | |
| *** dsariel_ has joined #openstack-manila | 09:06 | |
| *** gcb has quit IRC | 09:27 | |
| *** zhugx has quit IRC | 09:30 | |
| YuYangWang | who know how to config malina share file ,can get from public net user? | 09:32 |
| *** dsariel_ has quit IRC | 09:33 | |
| *** dsariel has quit IRC | 09:33 | |
| toabctl | vponomaryov, more comments on https://review.openstack.org/#/c/464205 ? | 09:52 |
| toabctl | jprovazn, ^^ | 09:52 |
| jprovazn | toabctl: hi, ack, I'll look in a moment, if both options are no-go, I'm fine the existing code | 09:57 |
| *** johnthetubaguy has left #openstack-manila | 10:05 | |
| -openstackstatus- NOTICE: gerrit is being restarted to help stuck git replication issues | 10:54 | |
| *** YuYangWang has quit IRC | 11:01 | |
| *** ganso has joined #openstack-manila | 11:09 | |
| *** td has joined #openstack-manila | 11:15 | |
| *** dsariel has joined #openstack-manila | 11:18 | |
| vponomaryov | toabctl: after looking more closely at https://github.com/openstack/manila/blob/16bfc829/manila/utils.py#L406 I think more strongly that we should reuse existing func, because it is very easy to change it to support infiniteness | 11:18 |
| *** dsariel_ has joined #openstack-manila | 11:18 | |
| toabctl | vponomaryov, does retrying support endless loops? | 11:19 |
| vponomaryov | toabctl: just look at source code | 11:19 |
| vponomaryov | toabctl: we should allow "retries" be equal to, for example, -1 and use it for infiniteness | 11:20 |
| vponomaryov | toabctl: in this case func "_print_stop" will always return False | 11:21 |
| toabctl | vponomaryov, yes. agreed. but this is a bug fix. and for fixing bugs, I don't want to add features to other functions | 11:21 |
| vponomaryov | toabctl: you do ADD feature | 11:22 |
| toabctl | vponomaryov, because adding the endless loop is clearly another commit and then fixing the bug in stable/ocata needs 2 commits | 11:22 |
| vponomaryov | toabctl: you add retried | 11:22 |
| vponomaryov | retrier* | 11:22 |
| toabctl | vponomaryov, I don't. I fix a bug during init | 11:22 |
| toabctl | both, you and bswartz also expected that the behavior is like I implemented it now with the bug fix. so that is not a feature imo. | 11:23 |
| vponomaryov | toabctl: this bug is not blocker | 11:23 |
| toabctl | vponomaryov, depends on the view. for one of our customers it really is... | 11:23 |
| vponomaryov | toabctl: so, I do not see any problem making little update to utility func | 11:23 |
| vponomaryov | and do "small" bugfix | 11:23 |
| vponomaryov | toabctl: I mean "technically" | 11:24 |
| vponomaryov | toabctl: we can merge separately fix and utility func update | 11:24 |
| vponomaryov | if you want it be separate | 11:24 |
| toabctl | ok. thanks for looking into that! | 11:26 |
| vponomaryov | toabctl: not persuaded but agreed? )) | 11:27 |
| toabctl | vponomaryov, yes :). it generates more work to dig into retrying (which hopefully is easy), create a second patch, rebase the other one, test it. but it's the right thing todo instead of reimplementing the retry loop over and over again... | 11:29 |
| toabctl | I just in general don't like it that in OpenStack, bug fixes take ages to get in... | 11:29 |
| vponomaryov | toabctl: actually, no need to separate to 2 patches | 11:30 |
| vponomaryov | toabctl: personally, I would do it in single patch | 11:30 |
| vponomaryov | toabctl: because utility func update will not be used separately in previous releases | 11:31 |
| vponomaryov | toabctl: so, it is one single change, AND no need to dig in "retrying" | 11:31 |
| toabctl | yeah. but the bugfix itself is much easier to understand if you do it in an extra commit. | 11:31 |
| toabctl | vponomaryov, ok. | 11:31 |
| vponomaryov | toabctl: only change of our own utility func | 11:31 |
| vponomaryov | you need only to remove this -> https://github.com/openstack/manila/blob/16bfc829/manila/utils.py#L449-L451 and update func doc | 11:33 |
| vponomaryov | then reuse it with "-1" value | 11:33 |
| toabctl | vponomaryov, I'm pretty sure some unittest will fail. let's see :) | 11:34 |
| vponomaryov | toabctl: it is normal. | 11:35 |
| vponomaryov | toabctl: and very fast action too | 11:36 |
| vponomaryov | toabctl: we talk more than it takes )) | 11:36 |
| toabctl | :) | 11:36 |
| toabctl | talking also fun from time to time :) | 11:36 |
| openstackgerrit | Digvijay Ukirde proposed openstack/manila master: GPFS Path: Fix bugs related to initialization of GPFS Driver https://review.openstack.org/464610 | 11:45 |
| jprovazn | toabctl: hah, I've commented the patch but I see vponomaryov already suggested quite similar at the beginning of your chat | 11:45 |
| toabctl | thx | 11:45 |
| vponomaryov | brother in soul ^_^ | 11:47 |
| jprovazn | :) | 11:52 |
| openstackgerrit | Digvijay Ukirde proposed openstack/manila master: GPFS Path: Fix bugs related to initialization of GPFS Driver https://review.openstack.org/464610 | 11:54 |
| *** chlong has quit IRC | 11:57 | |
| toabctl | jprovazn, max wait time is still missing. I get the point that I need to fix that as well. I know that this is all simple... | 11:57 |
| jprovazn | toabctl: well, I would consider using backoff_rate=1 and long enough interval in this case | 11:59 |
| jprovazn | toabctl: or do you see an issue with it? | 12:00 |
| toabctl | jprovazn, retrying more often in the beginning and increase the time is what we discussed some days ago. | 12:01 |
| jprovazn | toabctl: then it would require updating the _backoff_sleep func | 12:03 |
| toabctl | jprovazn, I know. it's easy. it's just no longer a bugfix imo. anyway. I'll just do it... | 12:04 |
| toabctl | and btw. I wonder why we have that wrapper fucntion at all.... | 12:04 |
| *** td has quit IRC | 12:05 | |
| jprovazn | toabctl: sorry for inconvenience, personally I think it's better to keep code DRY in this case, I can imagine that each component should use something like this | 12:08 |
| jprovazn | toabctl: which brings me to idea whether it would be better to handle these failrues on upper layer in manila/service.py | 12:10 |
| jprovazn | but out-of-scope of this patch (/me puts this to TODO list to take look at) | 12:10 |
| toabctl | jprovazn, hm. actually it seems not so easy (at least I did not found a solution in the last 5 minutes) to add a max() time for the backoff_sleep | 12:11 |
| toabctl | a configurable max() time I mean | 12:11 |
| jprovazn | toabctl: something liek this wouldn't work? | 12:14 |
| jprovazn | sec, network issues :) | 12:14 |
| jprovazn | https://pastebin.com/gpNzLpwk | 12:15 |
| toabctl | jprovazn, maybe. but retrying already has wait_exponential_max . shouldn't we use that? DRY :) | 12:17 |
| toabctl | anyway - thanks for helping with that. I'll update the patch later. | 12:18 |
| jprovazn | toabctl: it depends how DRY you wanna be :), briefly checking the retrying module, it seems we don't use the multiplier at all, so personally I think it's just fine to update _backoff_sleep func | 12:22 |
| *** rpittau has joined #openstack-manila | 12:26 | |
| *** jprovazn has quit IRC | 12:37 | |
| *** gouthamr has joined #openstack-manila | 12:39 | |
| *** xyang1 has joined #openstack-manila | 12:58 | |
| *** eharney has joined #openstack-manila | 13:03 | |
| *** dsariel_ has quit IRC | 13:04 | |
| *** dsariel has quit IRC | 13:04 | |
| *** dsariel has joined #openstack-manila | 13:06 | |
| openstackgerrit | Merged openstack/manila-image-elements master: Updated from global requirements https://review.openstack.org/465262 | 13:13 |
| *** chlong has joined #openstack-manila | 13:16 | |
| *** mkoderer has joined #openstack-manila | 13:17 | |
| *** jprovazn has joined #openstack-manila | 13:32 | |
| *** dustins has joined #openstack-manila | 13:34 | |
| openstackgerrit | Thomas Bechtold proposed openstack/manila master: Retry backend initialization https://review.openstack.org/464205 | 13:57 |
| openstackgerrit | Thomas Bechtold proposed openstack/manila master: Allow endless retry loops in the utility function https://review.openstack.org/465584 | 13:57 |
| toabctl | jprovazn, vponomaryov updated (and splitted into 2 commits) | 14:18 |
| jprovazn | toabctl, cool, thanks | 14:21 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-manila master: Update tox configuration https://review.openstack.org/465614 | 14:27 |
| *** jprovazn is now known as jprovazn_afk | 14:27 | |
| *** jprovazn_afk has quit IRC | 14:28 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-ui master: Fix compatibility with Horizon https://review.openstack.org/465628 | 14:47 |
| *** ociuhandu has joined #openstack-manila | 14:50 | |
| *** markstur has joined #openstack-manila | 14:52 | |
| *** ociuhandu has quit IRC | 15:01 | |
| *** dsariel has quit IRC | 15:04 | |
| *** dsariel has joined #openstack-manila | 15:20 | |
| *** hongbin has joined #openstack-manila | 15:22 | |
| openstackgerrit | Merged openstack/manila-ui master: Fix compatibility with Horizon https://review.openstack.org/465628 | 15:26 |
| *** rpittau has quit IRC | 15:28 | |
| *** a-pugachev has joined #openstack-manila | 15:41 | |
| *** dsariel has quit IRC | 15:42 | |
| *** gouthamr_ has joined #openstack-manila | 15:58 | |
| *** gouthamr has quit IRC | 15:59 | |
| *** lpetrut has quit IRC | 16:01 | |
| *** gouthamr2 has joined #openstack-manila | 16:09 | |
| *** pcaruana has quit IRC | 16:11 | |
| *** ociuhandu has joined #openstack-manila | 16:16 | |
| *** belmoreira has quit IRC | 16:20 | |
| *** ociuhandu has quit IRC | 16:25 | |
| *** gouthamr_ has quit IRC | 16:28 | |
| *** harlowja has joined #openstack-manila | 16:32 | |
| *** eharney has quit IRC | 16:44 | |
| *** eharney has joined #openstack-manila | 16:48 | |
| *** gouthamr has joined #openstack-manila | 17:01 | |
| *** markstur has quit IRC | 17:03 | |
| *** markstur has joined #openstack-manila | 17:04 | |
| *** harlowja has quit IRC | 17:26 | |
| bswartz | http://lists.openstack.org/pipermail/openstack-dev/2017-May/116921.html | 17:28 |
| bswartz | ^ everyone interested in PTG please read and respond on thread or in tomorrow's meeting | 17:28 |
| *** gaurangt has joined #openstack-manila | 17:29 | |
| *** arnewiebalck__ has joined #openstack-manila | 17:38 | |
| *** kaisers has joined #openstack-manila | 17:39 | |
| *** arnewiebalck__ has quit IRC | 18:04 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-ui master: Refactor admin dashboard https://review.openstack.org/464716 | 18:20 |
| openstackgerrit | Merged openstack/puppet-manila master: Update tox configuration https://review.openstack.org/465614 | 18:42 |
| *** MVenesio has joined #openstack-manila | 18:47 | |
| *** a-pugachev has quit IRC | 18:48 | |
| MVenesio | HI guys, i'm going to start testing manila with ceph, and a wanted to know if the generic driver is the best approach to provide NFS or maybe the nfs-ganesha driver works well for production environments | 18:49 |
| *** catintheroof has joined #openstack-manila | 18:59 | |
| *** lpetrut has joined #openstack-manila | 19:03 | |
| *** mkoderer has quit IRC | 19:27 | |
| *** harlowja has joined #openstack-manila | 19:39 | |
| *** xyang1 has quit IRC | 19:43 | |
| *** xyang1 has joined #openstack-manila | 19:45 | |
| *** lpetrut has quit IRC | 19:47 | |
| *** dsariel has joined #openstack-manila | 19:50 | |
| *** harlowja has quit IRC | 19:53 | |
| *** harlowja has joined #openstack-manila | 19:57 | |
| *** harlowja has quit IRC | 20:00 | |
| bswartz | MVenesio: if you're just testing, the simplest driver to configure is LVM, and it will give you NFS | 20:26 |
| bswartz | the Ceph driver works well but AFAIK it only gives you CephFS, not NFS yet | 20:27 |
| bswartz | Generic can turn any cinder storage into NFS storage, but it comes with it's own set of limitations -- I recommend LVM | 20:27 |
| Reepicheep | so thoughts regarding limitations of the generic driver with a ceph backed glance. are the limitations worth it when you need to not have a single point of failer that normal LVM provides? | 20:34 |
| bswartz | Reepicheep: not sure why the single point of failure is what concerns you -- generic would have the same limitation | 20:39 |
| bswartz | all of the "first party drivers" in manila are more focused on features and functionality than maintaining high availability in the face of unavoidable failures | 20:40 |
| Reepicheep | you are correct from the instance stand point there is still a single point that can fail. but my data volume exists within the Ceph pool | 20:42 |
| Reepicheep | where as if I have an LVM that volume is tied to a single node that could burst into flames and a user could loose data | 20:43 |
| *** gouthamr has quit IRC | 20:44 | |
| Reepicheep | is that accurate to the current state of manila drivers? | 20:44 |
| bswartz | that's a good point | 20:45 |
| bswartz | at least the generic driver outsources the basic data reliability to cinder which has lots of solutions for that | 20:46 |
| bswartz | the Cinder LVM driver has that limitation, but you have a bunch of other choices | 20:46 |
| bswartz | the Manila LVM driver has the same kind of downsides | 20:47 |
| bswartz | from that perspective if you want to run Manila for NFS and have that level of data reliability, your options are: use a proprietary solution, wait for the ceph nfs driver, use the generic driver | 20:48 |
| *** chlong has quit IRC | 20:52 | |
| *** catintheroof has quit IRC | 21:01 | |
| Reepicheep | bswartz: yeah.. really I'm waiting for the ceph nfs driver and for manilla to have some form of nova intance HA | 21:03 |
| Reepicheep | which if I remember correctly I saw as a possibility in an upcomming release Pike or maybe beyond.. I don't remember atm | 21:04 |
| Reepicheep | then there will be at least a solution to provide reliability across manila without a vendor driven solution | 21:05 |
| Reepicheep | I would love to help where I can, if I can come up with some time :| | 21:06 |
| *** hongbin has quit IRC | 21:08 | |
| *** Manor has quit IRC | 21:09 | |
| *** vivsoni has quit IRC | 21:11 | |
| *** vivsoni has joined #openstack-manila | 21:11 | |
| *** MVenesio has quit IRC | 21:11 | |
| *** gouthamr has joined #openstack-manila | 21:14 | |
| *** hongbin has joined #openstack-manila | 21:16 | |
| *** dustins has quit IRC | 21:17 | |
| *** a-pugachev has joined #openstack-manila | 21:44 | |
| *** a-pugachev has quit IRC | 21:49 | |
| *** xyang1 has quit IRC | 21:58 | |
| *** harlowja has joined #openstack-manila | 22:02 | |
| *** Manor has joined #openstack-manila | 22:10 | |
| *** markstur has quit IRC | 22:17 | |
| *** harlowja has quit IRC | 22:18 | |
| *** ganso has quit IRC | 22:18 | |
| *** markstur has joined #openstack-manila | 22:27 | |
| *** markstur has quit IRC | 22:32 | |
| *** dsariel has quit IRC | 22:39 | |
| bswartz | Reepicheep: I don't know the likelihood of that happening in Pike tbh | 22:42 |
| bswartz | we're all pretty busy, and if I get some time I plan to focus more on a containers-based solution for NFS | 22:42 |
| Reepicheep | bswartz: understood to me the data resiliance is priority one and I can get that today with ceph and generic | 22:43 |
| Reepicheep | as for containers i'm using kolla and I plan on continuing that ;) so I'm sold on that model | 22:44 |
| *** gouthamr2 has quit IRC | 22:49 | |
| *** eharney has quit IRC | 22:52 | |
| *** gouthamr has quit IRC | 23:12 | |
| *** gouthamr has joined #openstack-manila | 23:20 | |
| *** hongbin has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!