| *** dviroel has quit IRC | 00:02 | |
| *** baojg has joined #openstack-manila | 01:13 | |
| *** Kuirong has joined #openstack-manila | 01:23 | |
| *** whoami-rajat has joined #openstack-manila | 01:31 | |
| *** s0ru has quit IRC | 04:36 | |
| *** ganso has quit IRC | 05:05 | |
| *** pcaruana has joined #openstack-manila | 05:44 | |
| *** logan- has quit IRC | 06:00 | |
| *** logan_ has joined #openstack-manila | 06:01 | |
| *** logan_ is now known as logan- | 06:01 | |
| openstackgerrit | Xuan Yandong proposed openstack/python-manilaclient master: Cap sphinx for py2 to match global requirements https://review.opendev.org/671409 | 06:35 |
|---|---|---|
| *** kopecmartin|off is now known as kopecmartin | 06:59 | |
| openstackgerrit | KuiRong proposed openstack/manila master: Add Infortrend Manila Driver https://review.opendev.org/657775 | 07:06 |
| *** tosky has joined #openstack-manila | 08:31 | |
| *** petitbois has joined #openstack-manila | 08:37 | |
| *** petitbois has quit IRC | 09:13 | |
| *** petitbois has joined #openstack-manila | 09:59 | |
| *** petitbois has quit IRC | 09:59 | |
| *** petitbois has joined #openstack-manila | 10:00 | |
| petitbois | hello all! I'd like to define my own share protocol on manila. is this possible? | 10:00 |
| *** Kuirong has quit IRC | 10:03 | |
| *** whoami-rajat has quit IRC | 10:30 | |
| *** takamatsu has joined #openstack-manila | 10:37 | |
| tbarron | petitbois: sure, the last share protocol added was MAPRFS - git show 306c79fd02 will give you an idea | 10:52 |
| petitbois | tbarron: awesome thanks -- I'll check it out! | 10:53 |
| *** carloss has joined #openstack-manila | 11:22 | |
| *** ganso has joined #openstack-manila | 11:39 | |
| *** dviroel has joined #openstack-manila | 11:39 | |
| *** eharney has quit IRC | 11:40 | |
| openstackgerrit | Tom Barron proposed openstack/manila master: Run tempest jobs under python3 https://review.opendev.org/671031 | 11:45 |
| *** eharney has joined #openstack-manila | 12:26 | |
| openstackgerrit | Helen Walsh proposed openstack/manila master: Manila PowerMax - rebrand from VMAX to PowerMax https://review.opendev.org/663089 | 13:25 |
| *** hoonetorg has quit IRC | 13:28 | |
| openstackgerrit | Tom Barron proposed openstack/python-manilaclient master: Cap sphinx for py2 to match global requirements https://review.opendev.org/659439 | 13:31 |
| *** enriquetaso has joined #openstack-manila | 13:31 | |
| openstackgerrit | Merged openstack/manila-ui master: Remove the get_item method https://review.opendev.org/667554 | 13:41 |
| openstackgerrit | Merged openstack/manila-ui master: Modify the url of upper_constraints_file https://review.opendev.org/666212 | 13:46 |
| *** whoami-rajat has joined #openstack-manila | 13:55 | |
| openstackgerrit | Tom Barron proposed openstack/manila master: Run tempest jobs under python3 https://review.opendev.org/671031 | 13:59 |
| openstackgerrit | Merged openstack/python-manilaclient master: Add Python 3 Train unit tests https://review.opendev.org/669383 | 14:26 |
| *** s0ru has joined #openstack-manila | 14:47 | |
| *** jgrosso has joined #openstack-manila | 15:00 | |
| *** kopecmartin is now known as kopecmartin|off | 15:03 | |
| *** eharney_ has joined #openstack-manila | 15:12 | |
| *** eharney has quit IRC | 15:15 | |
| *** eharney_ is now known as eharney | 15:17 | |
| *** hoonetorg has joined #openstack-manila | 15:26 | |
| *** enriquetaso has quit IRC | 15:52 | |
| *** vhariria has joined #openstack-manila | 15:58 | |
| tbarron | gouthamr: you are focused on the deferred cutover but I don't read the bug as complaining about that. | 16:01 |
| tbarron | gouthamr: title of the bug is "space savings from efficiency could be lost ..."P | 16:01 |
| gouthamr | tbarron: yes, but ONTAP says that the deferring is going to cost with "space efficiency loss" | 16:01 |
| tbarron | gouthamr: isn't that true whether the cutover is deferred or not? | 16:01 |
| tbarron | gouthamr: the volume will be in a new aggregate and | 16:02 |
| *** eharney has quit IRC | 16:02 | |
| tbarron | gouthamr: maybe my memory fails but | 16:02 |
| tbarron | gouthamr: won't dedupe be done for blocks in the whole aggregate? | 16:02 |
| * tbarron might be confused about that | 16:02 | |
| tbarron | gouthamr: if so, then whether there are matching blocks to dedupe or not will be a total crap shoot | 16:03 |
| tbarron | gouthamr: so space efficiency on the destination won't match that on the source | 16:03 |
| tbarron | gouthamr: but I see that this warning is being quoted in the context of the delayed cutover | 16:04 |
| tbarron | gouthamr: so probably you are right that there is some other space-efficiency effect going on that I'm not aware of | 16:04 |
| tbarron | gouthamr: "Deduplication is not ready for volmove cutover" | 16:05 |
| tbarron | gouthamr: so ack, but I don't get it. | 16:05 |
| gouthamr | tbarron: bswartz may know - the way this works right now is; when "migration-start" is issued, we begin a data copy to the destination and we wait on that initial data copy - you know how ONTAP does it - internal snapshots and xcopy and all that magic... When the data copy is complete, ONTAP is ready to cutover to the destination volume -> what this means is writes will be directed to the new volume, and reads may be temporarily done to the | 16:05 |
| gouthamr | old volume while any remaining data that wasn't part of the original snapshot is going to be copied over | 16:05 |
| gouthamr | we have a choice to allow ONTAP to do a cutover when we initiate a migration - that is really the best way to use this feature, but, that isn't teh way migration is designed in manila | 16:06 |
| gouthamr | we allow administrators a two phased approach, for consistency with the fallback path -> this means, they must issue a migration-complete to begin using the destination volume and delete the source and all that.. | 16:07 |
| gouthamr | unlike SnapMirror that keeps running on a schedule, volume moves in ONTAP don't copy any new data while cutover is deferred | 16:08 |
| gouthamr | so if users continue writing to their shares, new data won't be available in the destination share (not like they care) | 16:09 |
| gouthamr | until the administrator issues migration-complete | 16:09 |
| ganso | tbarron, gouthamr: sorry about the delay, I added my thoughts in https://bugs.launchpad.net/manila/+bug/1492211 w.r.t https://review.opendev.org/#/c/224102/ | 16:09 |
| openstack | Launchpad bug 1492211 in Manila "Generic driver: error deleting shares if service_instance is not created" [Low,In progress] - Assigned to Lucian Petrut (petrutlucian94) | 16:09 |
| gouthamr | ganso: Good idea, he's added some notes on the code patch: "The issue being fixed occurs only if the server is not created at all. In that case, you won't have an instance_id. This can easily be reproduced by trying to use a wrong flavor id, for example, and then attempting to delete the share. If the instance is created, you will have the instance id set in the exception "detail_data" field, used to update the DB. | 16:11 |
| gouthamr | https://github.com/openstack/manila/blob/master/manila/share/drivers/service_instance.py#L523" | 16:11 |
| ganso | gouthamr: hmm I missed that. Either way, would be a good idea to have this retested | 16:13 |
| tbarron | ganso: gouthamr: thanks! | 16:13 |
| gouthamr | tbarron: here's a warning each time we try this from the command line: "When the transfer is complete, trigger the cutover to avoid loss of space savings from efficiency." | 16:13 |
| tbarron | gouthamr: ack and probably worth noting in the bug | 16:14 |
| gouthamr | and some more unofficial docs here: https://www.storagefreak.net/2017/07/netapp-cdot-volume-move | 16:14 |
| gouthamr | "DataMotion for Volume (volume move operation) are retaining deduplication savings on a FlexVol. If deduplication operation is in progress when the volume move is active, then during the volume move cutover phase the current deduplication ops will be aborted." | 16:14 |
| gouthamr | "Once the volume move is completed, deduplication operations can be restarted (however it will not pick-up and the previous checkpoint, just start from the beginning)." | 16:14 |
| gouthamr | so it's quite bad to leave volume moves in that state of deferred cutovers :) iirc we do call this out in the docs | 16:15 |
| tbarron | gouthamr: thanks for 'splainin ... | 16:23 |
| openstackgerrit | Tom Barron proposed openstack/manila master: Run tempest jobs under python3 https://review.opendev.org/671031 | 17:05 |
| *** eharney has joined #openstack-manila | 17:13 | |
| *** tosky has quit IRC | 17:17 | |
| *** irclogbot_2 has quit IRC | 18:07 | |
| *** altlogbot_1 has quit IRC | 18:07 | |
| *** petitbois has quit IRC | 18:08 | |
| *** irclogbot_3 has joined #openstack-manila | 18:08 | |
| *** altlogbot_1 has joined #openstack-manila | 18:09 | |
| *** dsariel has joined #openstack-manila | 18:42 | |
| *** dsariel has quit IRC | 18:43 | |
| *** enriquetaso has joined #openstack-manila | 19:18 | |
| *** _erlon_ has joined #openstack-manila | 19:23 | |
| openstackgerrit | Tom Barron proposed openstack/manila master: Run tempest jobs under python3 https://review.opendev.org/671031 | 19:27 |
| *** e0ne has joined #openstack-manila | 19:38 | |
| *** whoami-rajat has quit IRC | 19:44 | |
| *** vhariria has left #openstack-manila | 19:48 | |
| *** pcaruana has quit IRC | 20:01 | |
| *** jgrosso has quit IRC | 20:07 | |
| *** eharney has quit IRC | 20:32 | |
| enriquetaso | gouthamr, o/ how are u? what are your thoughts about my review https://review.opendev.org/#/c/667744/ ? :D | 21:01 |
| enriquetaso | s0ru ^ | 21:03 |
| *** raissa has joined #openstack-manila | 21:06 | |
| *** raissa has quit IRC | 21:06 | |
| gouthamr | enriquetaso: agree with vkmc.. :) - partially also thought we can give s0ru exposure to unit tests this way | 21:07 |
| *** raissa has joined #openstack-manila | 21:07 | |
| *** raissa has quit IRC | 21:07 | |
| enriquetaso | vkmc, how can do https://opendev.org/openstack/python-manilaclient/src/branch/master/manilaclient/tests/unit/v2/test_shell.py#L407 receive a list of shares empty ? | 21:07 |
| vkmc | enriquetaso, I was trying to figure out that so we can help out s0ru | 21:07 |
| enriquetaso | gouthamr++ | 21:07 |
| enriquetaso | vkmc++ | 21:07 |
| enriquetaso | thanks! | 21:07 |
| vkmc | but I cannot finish my devstack deploy on rdocloud :/ | 21:07 |
| *** raissa has joined #openstack-manila | 21:08 | |
| vkmc | I'll spin up a vm so we can get her unblocked | 21:08 |
| *** raissa has quit IRC | 21:08 | |
| *** eharney has joined #openstack-manila | 21:08 | |
| s0ru | \o/ | 21:12 |
| gouthamr | vkmc: she should still have access to mine? | 21:12 |
| s0ru | yeah! | 21:17 |
| *** e0ne has quit IRC | 21:17 | |
| s0ru | I've been using it :) | 21:18 |
| enriquetaso | :O | 21:22 |
| enriquetaso | oki | 21:28 |
| s0ru | I don't know if vkmc can access to the vm too, devstack it's up and running | 21:29 |
| s0ru | gouthamr: like u did with my public key | 21:31 |
| *** enriquetaso has quit IRC | 21:46 | |
| *** eharney has quit IRC | 22:16 | |
| *** carloss has quit IRC | 22:21 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!