| openstackgerrit | Carlos Eduardo proposed openstack/manila-tempest-plugin master: Fix export location sent to manage share operation https://review.opendev.org/753868 | 00:34 |
|---|---|---|
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila master: Fix logic that determines a share exists before manage https://review.opendev.org/749173 | 00:37 |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila master: Fix logic that determines a share exists before manage https://review.opendev.org/749173 | 00:51 |
| openstackgerrit | Carlos Eduardo proposed openstack/manila-tempest-plugin master: Fix export location sent to manage share operation https://review.opendev.org/753868 | 02:26 |
| openstackgerrit | zhongshengping proposed openstack/puppet-manila master: Prepare Victoria RC1 https://review.opendev.org/753904 | 02:49 |
| openstackgerrit | MaAoyu proposed openstack/python-manilaclient master: Remove install unnecessary packages https://review.opendev.org/753923 | 03:05 |
| *** andrebeltrami has quit IRC | 03:12 | |
| *** vhar has quit IRC | 06:06 | |
| *** lpetrut has joined #openstack-manila | 06:10 | |
| *** vishalmanchanda has joined #openstack-manila | 06:30 | |
| openstackgerrit | haixin proposed openstack/manila master: Add generic fuzzy matching logic to the database layer https://review.opendev.org/734041 | 07:19 |
| openstackgerrit | MaAoyu proposed openstack/manila master: Fix wrong totalcount returned by share listing query https://review.opendev.org/703025 | 07:25 |
| *** baojg has quit IRC | 07:29 | |
| *** baojg has joined #openstack-manila | 07:30 | |
| *** tosky has joined #openstack-manila | 08:04 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/manila-ui stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/753975 | 08:46 |
| openstackgerrit | OpenStack Release Bot proposed openstack/manila-ui stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/753976 | 08:46 |
| openstackgerrit | OpenStack Release Bot proposed openstack/manila-ui master: Update master for stable/victoria https://review.opendev.org/753977 | 08:46 |
| openstackgerrit | OpenStack Release Bot proposed openstack/manila-ui master: Add Python3 wallaby unit tests https://review.opendev.org/753978 | 08:46 |
| *** baojg has quit IRC | 09:04 | |
| *** LinPeiWen has quit IRC | 09:26 | |
| *** LinPeiWen has joined #openstack-manila | 09:26 | |
| *** baojg has joined #openstack-manila | 09:41 | |
| *** vhari has joined #openstack-manila | 09:48 | |
| openstackgerrit | MaAoyu proposed openstack/manila master: Fix wrong totalcount returned by share listing query https://review.opendev.org/703025 | 10:21 |
| *** baojg has quit IRC | 10:22 | |
| *** baojg has joined #openstack-manila | 10:23 | |
| openstackgerrit | Andre Beltrami proposed openstack/manila master: [DOC] Add admin documentation for share revert to snapshot https://review.opendev.org/753105 | 11:59 |
| *** eharney has joined #openstack-manila | 13:06 | |
| *** baojg has quit IRC | 13:21 | |
| *** jv has quit IRC | 13:50 | |
| *** jv has joined #openstack-manila | 13:54 | |
| *** andrebeltrami has joined #openstack-manila | 13:56 | |
| openstackgerrit | Andre Beltrami proposed openstack/manila master: [DOC] Add admin documentation for share revert to snapshot https://review.opendev.org/753105 | 13:56 |
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: dnf: Correctly configure repos https://review.opendev.org/753577 | 14:01 |
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: zuul: Add devstack-plugin-ceph-tempest-fedora-latest NV job https://review.opendev.org/753581 | 14:01 |
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: zuul: Add devstack-plugin-ceph-tempest-fedora-latest NV job https://review.opendev.org/753581 | 14:06 |
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: Use the py3 el8 RPMs on Fedora https://review.opendev.org/754106 | 14:06 |
| openstackgerrit | Merged openstack/manila-ui stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/753975 | 14:12 |
| *** maaritamm has joined #openstack-manila | 14:13 | |
| openstackgerrit | Merged openstack/manila-ui stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/753976 | 14:13 |
| openstackgerrit | Merged openstack/manila-ui master: Update master for stable/victoria https://review.opendev.org/753977 | 14:13 |
| *** baojg has joined #openstack-manila | 14:21 | |
| dviroel | tbarron: o/ | 14:35 |
| tbarron | dviroel: hey | 14:36 |
| dviroel | tbarron: hey, i believe that our gate is broken until https://review.opendev.org/#/c/743134/ got merged | 14:37 |
| dviroel | lvm is failing on Focal | 14:38 |
| tbarron | dviroel: ok, thanks, i'll hold of rechecking other jobs | 14:38 |
| tbarron | reviews | 14:38 |
| dviroel | tbarron: np, tks | 14:39 |
| *** baojg has quit IRC | 14:46 | |
| *** baojg has joined #openstack-manila | 14:47 | |
| openstackgerrit | Merged openstack/manila-ui master: Add Python3 wallaby unit tests https://review.opendev.org/753978 | 14:49 |
| *** lpetrut has quit IRC | 15:00 | |
| *** felipe_rodrigues has joined #openstack-manila | 15:00 | |
| *** lpetrut has joined #openstack-manila | 15:07 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila-tempest-plugin master: Keep few integration jobs running on Bionic https://review.opendev.org/743134 | 15:13 |
| *** baojg has quit IRC | 15:45 | |
| openstackgerrit | Andre Beltrami proposed openstack/manila master: [DOC] Add admin and user documentation for share revert to snapshot https://review.opendev.org/753105 | 15:46 |
| openstackgerrit | Merged openstack/manila-ui master: Remove install unnecessary packages https://review.opendev.org/753506 | 15:53 |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila-test-image master: [DNM] Testing CI https://review.opendev.org/754126 | 15:58 |
| openstackgerrit | Merged openstack/manila master: [DOC] Add admin and user documentation for share revert to snapshot https://review.opendev.org/753105 | 16:19 |
| *** maaritamm has quit IRC | 16:30 | |
| *** maaritamm has joined #openstack-manila | 16:31 | |
| *** maaritamm has quit IRC | 16:33 | |
| gouthamr | vkmc: i can haz your +2 back on https://review.opendev.org/743134/ ? :) | 17:04 |
| vkmc | gouthamr, of course | 17:05 |
| gouthamr | vkmc++ ty | 17:05 |
| *** lpetrut has quit IRC | 17:06 | |
| *** lpetrut has joined #openstack-manila | 17:09 | |
| ganso | carloss and dviroel: please see https://review.opendev.org/#/c/749173/ | 17:32 |
| *** baojg has joined #openstack-manila | 17:32 | |
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: zuul: Add devstack-plugin-ceph-tempest-fedora-latest NV job https://review.opendev.org/753581 | 17:33 |
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: Do not run dnf update -y after configuring ceph repos https://review.opendev.org/754203 | 17:33 |
| ganso | Also, is there anywhere where the logic related to this has arealdy changed? is it only here https://review.opendev.org/#/c/741588/ ? sorry if I missed that behavioral change and it has already merged | 17:38 |
| ganso | also ping gouthamr on https://review.opendev.org/#/c/749173/ as he is the committer | 17:40 |
| dviroel | ganso: o/ | 17:40 |
| dviroel | i believe that the manage issue was only https://review.opendev.org/#/c/749173/15/manila/share/api.py@775 | 17:41 |
| dviroel | replace the key 'export_location' by 'export_location_path' | 17:42 |
| dviroel | then I started the discussion why include the share_type_id in the filter | 17:42 |
| gouthamr | ganso: hey, I agree we can explore the optimization suggested in point #2 | 17:43 |
| dviroel | your point regarding the 'host' make sense to me | 17:43 |
| ganso | dviroel: right, the share_type_id doesn't make sense to be there. I don't remember why it was there in the first place | 17:43 |
| gouthamr | ganso: i don't understand point #1 | 17:44 |
| gouthamr | " if there are 2 backends with shares that were created outside of manila (hence why they are being managed), and they are named the same, isn't this going to fail? | 17:44 |
| ganso | gouthamr: by filtering based on 'host' we don't allow duplicate export locations on the same backend, but we do allow on separate different backend, as they are no relation to each other, so no reason why they cannot be the same. Using virtual IPs and virtual networks it is possible that they can be named the same way | 17:45 |
| gouthamr | ganso: so you have two backends on the same network with a share that is named the same? | 17:45 |
| ganso | gouthamr: not on the "same" network, but with identical CIDRs, with identical IPs, named the same way | 17:46 |
| ganso | thinking of it in a public cloud environment, makes it even more likely to be possible | 17:48 |
| gouthamr | ganso: hmmm, there's a possibility - but the degree of randomness we ensure doesn't allow for this | 17:48 |
| ganso | gouthamr: are we enforcing that export locations are renamed after being managed? | 17:48 |
| gouthamr | ganso: that's the contract, but we're not enforcing it | 17:49 |
| gouthamr | ganso: there is a proposed unimplemented test case iirc | 17:49 |
| ganso | gouthamr: oh ok, then if there is a contract and it is expected that they are renamed, ok, they will never collide. So we can pursue only #2 (which is even easier if we don't check the host) | 17:49 |
| gouthamr | https://docs.openstack.org/manila/latest/contributor/driver_requirements.html#managing-unmanaging-shares -- (this needs to be updated) | 17:52 |
| gouthamr | we don't mention that here and "After the previously existing share is registered in manila, it is completely controlled by manila and should not be handled externally anymore." makes sense, but doesn't make a requirement for drivers | 17:52 |
| gouthamr | nor is it mentioned here: https://github.com/openstack/manila/blob/39efc2bde81c2a0c747a491d3a778b822ca263b8/manila/share/driver.py#L961-L1013 | 17:53 |
| ganso | right, so the case where 2 different shares exist in //10.0.0.10/my_share is still possible, not enforced, and seems like we are not adivising that as well. Therefore we may need to handle this situation otherwise it could become a bug, someone will could come and say "hey why can't I managed 2 shares named my_share?" if this is not handled | 17:55 |
| gouthamr | yeah, i can add the host check back | 17:56 |
| ganso | for the NetApp and Container drivers DHSS=True manage/unmanage I don't remember if it was renamed. I think it is possible that the container driver is not renaming, I remember having this discussion in the past | 17:56 |
| gouthamr | it's an esoteric situation, but possible - and i don't mean to make an API change.. | 17:57 |
| ganso | gouthamr: about https://review.opendev.org/#/c/741929 I don't agree with the change (or I don't understanding it completely). We've agreed in the past that manage_error should only be unmanageable, not deletable | 17:58 |
| gouthamr | ganso: https://github.com/openstack/manila/blob/ea7e304e33b2573920820848c844973c733a1120/manila/share/drivers/container/driver.py#L361 | 17:58 |
| ganso | gouthamr: thanks! easier to just update the docs then | 17:59 |
| gouthamr | ganso: yeah i'd think so - but, i'm trying to think - specifically in case of DHSS=True, this different-network, same CIDRs becomes prominent, so i wonder if we must dedupe with share_server_id as well? | 18:01 |
| ganso | gouthamr: that was my thought back then, but we always have UUIDs, we will never need to dedupe | 18:03 |
| ganso | *if we always have UUIDs | 18:03 |
| ganso | (re: delete manage_error) because whatever share we try to delete will invoke the delete path of the driver, and that could potentially delete the actual share. IIRC lseki and I tested this in the lab and we could get real shares deleted, because we could never be sure why the manage_error failed | 18:04 |
| gouthamr | ganso: yes, can't be sure that's what all share drivers are doing - we're removing a guardrail in effect with this change, and it will be back-ported | 18:04 |
| *** maaritamm has joined #openstack-manila | 18:06 | |
| gouthamr | ganso: yeah, re: delete manage_error i agree - haixin's issues were around quota calculations - | 18:06 |
| *** maaritamm has quit IRC | 18:08 | |
| gouthamr | unmanaging does reset the quota, so i think that's a good workflow - i wonder if the confusion arose because it wasn't documented well | 18:08 |
| ganso | gouthamr: thanks! I -1'ed both of his patches. | 18:09 |
| ganso | and sorry again gouthamr and dviroel from popping in and asking all those questions, a lot of stuff has been changing since the last time I knew those parts of the code like the back of my hand. Indeed I could have spent some time searching the docs or re-reading the code. I will try to do that last time. I saw the patch already with +W and went a bit desperate lol | 18:12 |
| ganso | try to do that *next time | 18:12 |
| * ganso gets more dyslexic as he gets older | 18:13 | |
| gouthamr | heh, thank you for reviewing :) all of these were good remarks! | 18:13 |
| gouthamr | ganso: older? you're going the wrong direction - turn back | 18:14 |
| ganso | xD | 18:14 |
| ganso | gouthamr: I miss those architectural discussions of ours =) | 18:15 |
| dviroel | ganso: hey, was everything intentional, just to get your attention :) | 18:16 |
| dviroel | ganso: lol, thanks for your review man | 18:17 |
| gouthamr | ganso: hey come to the PTG - i'm still fun to argue with, i make a point and derail a discussion, and dviroel tries to tell me i'm wrong in a very gentleman-ly way | 18:17 |
| openstackgerrit | Merged openstack/manila-tempest-plugin master: Keep few integration jobs running on Bionic https://review.opendev.org/743134 | 18:21 |
| *** tosky has quit IRC | 18:22 | |
| dviroel | gouthamr: ^ | 18:27 |
| gouthamr | good news, lets get the qos rechecked, and hold on the manage one | 18:28 |
| gouthamr | heh, jinx dviroel | 18:29 |
| dviroel | haha | 18:30 |
| *** maaritamm has joined #openstack-manila | 18:43 | |
| *** maaritamm has quit IRC | 18:44 | |
| *** openstackgerrit has quit IRC | 19:14 | |
| *** openstackgerrit has joined #openstack-manila | 19:37 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila master: Fix logic that determines a share exists before manage https://review.opendev.org/749173 | 19:37 |
| gouthamr | tbarron: o/ so the errors in manila-test-image are identical - the build's currently broken, and the new job fails at exactly the same place: compare https://review.opendev.org/#/c/754126/ with https://review.opendev.org/#/c/746565/ | 19:48 |
| gouthamr | tbarron: looks like we need to debug the image building (if we care) | 19:48 |
| gouthamr | tbarron: but the job can be fixed up prior to that.. :) | 19:49 |
| tbarron | gouthamr: ack | 19:50 |
| *** maaritamm has joined #openstack-manila | 19:58 | |
| *** maaritamm has quit IRC | 20:00 | |
| *** maaritamm has joined #openstack-manila | 20:15 | |
| *** maaritamm has left #openstack-manila | 20:17 | |
| tbarron | gouthamr: manila-image-elements is independent of release cycles | 20:28 |
| tbarron | gouthamr: are you aiming to get the job fixed in this time frame anyways? | 20:29 |
| gouthamr | tbarron: this is for manila-test-image | 20:29 |
| tbarron | d'oh | 20:29 |
| gouthamr | tbarron: yes, i think so; don't expect anything to break if we don't transition, except its a bit cleaner/leaner code to maintain | 20:30 |
| gouthamr | tbarron: tosky puts it as low-prio because it doesn't involve devstack-gate :) | 20:30 |
| tbarron | so that's even *more* independent, nopt in deliverables at all | 20:30 |
| tbarron | but | 20:30 |
| tbarron | iiuc you are proposing that we merge https://review.opendev.org/#/c/746565 and visit the (independent) build failures later | 20:32 |
| tbarron | I'm ok with that, just want to make sure I understand. | 20:32 |
| tbarron | Added my +2, you can workflow or ping me. | 20:33 |
| tbarron | Good cleanup. | 20:34 |
| openstackgerrit | Tom Barron proposed openstack/manila-test-image master: DNM - testing with current buildroot https://review.opendev.org/754215 | 21:00 |
| *** lpetrut has quit IRC | 21:30 | |
| *** raildo has quit IRC | 21:42 | |
| *** baojg has quit IRC | 21:53 | |
| gouthamr | great, ty tbarron | 21:53 |
| *** baojg has joined #openstack-manila | 21:54 | |
| *** baojg has quit IRC | 21:58 | |
| *** baojg has joined #openstack-manila | 21:59 | |
| gouthamr | "test_manage_invalid_param_manage_error" - has a weird test case | 22:00 |
| openstackgerrit | Merged openstack/manila-test-image master: Drop legacy-base from manila-test-image-build https://review.opendev.org/746565 | 22:03 |
| *** vishalmanchanda has quit IRC | 22:39 | |
| openstackgerrit | Tom Barron proposed openstack/manila-test-image master: DNM - testing with current buildroot https://review.opendev.org/754215 | 22:53 |
| gouthamr | dviroel: o/ | 23:01 |
| gouthamr | dviroel: if you're still around, it looks like https://review.opendev.org/749173 needs some focus; we have a failing test which, imo could be improved - but i don't want to delay RC for this issue | 23:02 |
| gouthamr | dviroel: i'm interested in your view however.. lets keep talking on the review | 23:02 |
| gouthamr | with the rest of the bugfixes deferred, i'm waiting on https://review.opendev.org/753315/ to merge before requesting the RC | 23:03 |
| openstackgerrit | Tom Barron proposed openstack/manila-test-image master: DNM - testing with current buildroot https://review.opendev.org/754215 | 23:13 |
| *** andrebeltrami has quit IRC | 23:45 | |
| openstackgerrit | Merged openstack/manila master: [NetApp] Improve Adaptive QoS functionality https://review.opendev.org/753315 | 23:54 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!