Friday, 2024-07-19

opendevreviewJonathan Koerber proposed openstack/manila master: Adding OpenAPi Shemas to User Messages  https://review.opendev.org/c/openstack/manila/+/92256600:09
opendevreviewVolodymyr Boiko proposed openstack/manila master: Add share driver for VastData storage  https://review.opendev.org/c/openstack/manila/+/91538003:40
opendevreviewOpenStack Proposal Bot proposed openstack/manila master: Imported Translations from Zanata  https://review.opendev.org/c/openstack/manila/+/92446604:31
opendevreviewJonathan Koerber proposed openstack/manila master: Adding JSON Schema to Quota Sets  https://review.opendev.org/c/openstack/manila/+/92446705:46
opendevreviewMerged openstack/manila master: Imported Translations from Zanata  https://review.opendev.org/c/openstack/manila/+/92446606:13
opendevreviewJonathan Koerber proposed openstack/manila master: Adding JSON Schema to Quota Class Sets  https://review.opendev.org/c/openstack/manila/+/92446806:46
opendevreviewchristian okeke proposed openstack/manila-ui master: Add metadata create textbox to Create Snapshot Form  https://review.opendev.org/c/openstack/manila-ui/+/92359709:26
gansojayaanand gireesh hi! I have a customer that is having difficulties balancing the shares across pools. The scheduler keeps creating shares only in the same pool. We found out that for some reason it is not calculating the total_capacity and therefore it cannot calculate the "utilization" property to use in the default goodness function10:32
ganso'total_capacity_gb': 'unknown', 'free_capacity_gb': 72041.21,10:32
gansowe have customized the goodness function as a workaround but we would like to know what is wrong that is preventing the total_capacity_gb from being properly detected, sounds like an issue within the ontap box10:33
gansoI checked some upstream logs and they all show the total_capacity: 'total_capacity_gb': 1360.77, 'free_capacity_gb': 1354.75,10:34
gansois there a known reason or a setting in the ontab box that causes this?10:34
gansoor maybe it is related to the account being svm-scoped and therefore cannot see the total_capacity of the pool?10:34
opendevreviewkiran pawar proposed openstack/manila master: Pass share metadata updates to backend drivers  https://review.opendev.org/c/openstack/manila/+/92406410:35
gireesh@ganso, is openstack share pool list --detail is showing the total_capacity_gb on customer setup ..?10:39
gireeshwhich ONTAP system they are using ..? version of ONTAP ..?10:39
gansoI am trying to find that, hold on a min10:41
gireesh maybe it is related to the account being svm-scoped and therefore cannot see the total_capacity of the pool? : if it is showing the free capacity then there should not be any issue with vserver scoping 10:41
gansogireesh: the customer hasn't provided the output of manila pool list --detail but I believe it would come from the same place as it is being retrieved from the logs: https://paste.openstack.org/show/bPzFHd4atZCI8MkqQSp1/10:43
gansogireesh: I am trying to find the ontap version from the logs, hold on10:44
gireeshIs there any exception in logs, yes pool itself is not showing the total_capacity10:53
gansogireesh: they are running 9.13.1P310:53
gansogireesh: I don't see any exception in the logs related to this10:54
gansobut there are others10:55
gansoI'm gathering them into a paste10:55
gansogireesh: https://paste.openstack.org/show/bXvWl2SdcKO1ErajDJoK/11:02
gireeshthanks @ganso for sharing the details, let me try this in my lab with SVM scoping 11:17
opendevreviewkiran pawar proposed openstack/manila master: Update share/snapshot instance deferred deletion  https://review.opendev.org/c/openstack/manila/+/92131311:35
gireesh@ganso, I check this in my local lab and I am also hitting this issue. we will be hit this issue only for vserver scoping. this is working fire for cluster scoping 12:12
gansogireesh: thanks! do you think this has a possible fix? like a way to obtain the total_capacity even with SVM scoping? or is it considered a definitive limitation?12:12
gireeshI looked into code also and  in vserver scoping wre using "vserver-get" ZAPI to get the aggregate used capacity and ZAPI also not have any option to get the total capacity of aggregate 12:14
gireeshhttps://www.irccloud.com/pastebin/GTsQJFvM/12:15
gireeshyou can asked customer to file the bug, let see how to fix this issue. Might be we have invoke one more ZAPI in our code to get the aggregate capacity. 12:17
gansogireesh: oh ok, so it will require driver code change to achieve that12:18
gireeshyes, once customer file the bug please share with  us12:18
gireesh@ganso, what is the name of the customer who is hitting this issue?12:27
gansogireesh: responded to you privately, and also logged the bug: https://bugs.launchpad.net/manila/+bug/207359912:38
gansolet me know if you need more info in the bug12:38
gireeshthanks @ganso for sharing the bug. I know how to reproduce this issue, so no additional information is needed on bug 15:25
VolodymyrBoiko[m]Hi gouthamr carloss Thanks for review. Could I ask what is a next step?... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JMzpHxRgqkwLGxXMKhApTHbz>)15:53
gouthamrVolodymyrBoiko[m]: great thank you15:57
gouthamrVolodymyrBoiko[m]: we ensure we have 2 code-review +2s and any number of +1s are helpful.. looks like this patch has had enough reviewers looking; so i just hit the "workflow" vote.. this means that the Zuul CI system will be running through its final tests and if they pass, it'll vote positively, and gerrit will merge your patch15:59
gouthamrVolodymyrBoiko[m]: so, next steps would be for you to enable your CI to vote on all manila changes in the master branch; we totally recommend testing stable branches too.. but, not irrelevant ones.. i.e., the driver's getting added in the "stable/2024.2" branch (we don't have that yet, but we'll have that branch in a few weeks: https://releases.openstack.org/dalmatian/schedule.html) 16:01
gouthamrso when we do, it'll be helpful to run your CI system on changes submitted against this branch too.. 16:01
gouthamrVolodymyrBoiko[m]: if someone needs to report bugs against the driver, they will do so here: https://bugs.launchpad.net/manila 16:02
gouthamrVolodymyrBoiko[m]: with the information you added here https://wiki.openstack.org/wiki/ThirdPartySystems/VASTData_Manila_CI (thank you), we can get in touch with you when we need to16:03
gouthamrbut don't be a stranger: https://docs.openstack.org/manila/latest/contributor/contributing.html#communication :)16:04
gouthamrwe also have a Virtual Project Teams Gathering in Oct: https://ptg.openinfra.dev/  https://etherpad.opendev.org/p/epoxy-ptg-manila-planning16:05
carlossgouthamr++16:16
carlossVolodymyrBoiko[m]: thank you very much for your work on the driver. Goutham added a good summary of what is next16:16
opendevreviewMerged openstack/manila master: Add share driver for VastData storage  https://review.opendev.org/c/openstack/manila/+/91538017:14
opendevreviewJonathan Koerber proposed openstack/manila master: Adding OpenAPi Shemas to User Massages  https://review.opendev.org/c/openstack/manila/+/92256617:33
opendevreviewAshley Rodriguez proposed openstack/manila master: Metadata for Share Export Location  https://review.opendev.org/c/openstack/manila/+/88754318:09
opendevreviewAshley Rodriguez proposed openstack/python-manilaclient master: Metadata for share export location  https://review.opendev.org/c/openstack/python-manilaclient/+/88967718:09
opendevreviewCarlos Eduardo proposed openstack/python-manilaclient master: WIP Add source share server ID to server list  https://review.opendev.org/c/openstack/python-manilaclient/+/92453919:51
opendevreviewAshley Rodriguez proposed openstack/python-manilaclient master: Metadata for share export location  https://review.opendev.org/c/openstack/python-manilaclient/+/88967721:30
opendevreviewAshley Rodriguez proposed openstack/manila master: Metadata for Share Export Location  https://review.opendev.org/c/openstack/manila/+/88754321:35

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!