09:02:01 <masahito> #startmeeting blazar
09:02:01 <openstack> Meeting started Tue Jul 31 09:02:01 2018 UTC and is due to finish in 60 minutes.  The chair is masahito. Information about MeetBot at http://wiki.debian.org/MeetBot.
09:02:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
09:02:04 <openstack> The meeting name has been set to 'blazar'
09:02:16 <masahito> sorry, being late.
09:02:24 <masahito> time to blazar team meeting :-)
09:02:29 <masahito> #topic RollCall
09:02:34 <priteau> o/
09:02:40 <bertys> o/
09:03:12 <masahito> priteau, bertys: o/
09:03:18 <masahito> Today's agenda is
09:03:19 <masahito> 1.
09:03:28 <masahito> 1. R-3 milestone
09:03:32 <masahito> 2. PTG planning
09:03:35 <masahito> 3. AOB
09:03:38 <masahito> anything else
09:03:39 <masahito> ?
09:03:49 <bertys> PTL election maybe?
09:04:12 <bertys> Few hours left http://lists.openstack.org/pipermail/openstack-dev/2018-July/132587.html
09:04:36 <masahito> bertys: right.
09:04:42 * tetsuro was sorry tobe late
09:04:50 <masahito> thanks.
09:05:06 <masahito> All right. Let's start.
09:05:23 <masahito> #topic R-3 milestone
09:05:53 <masahito> This is an announcement for the team.
09:06:15 <masahito> I putted tags 2.0.0.0b3 for the blazar repo and 2.0.0 for the blazarclient.
09:06:30 <masahito> https://git.openstack.org/cgit/openstack/blazar/log/
09:06:41 <masahito> https://git.openstack.org/cgit/openstack/python-blazarclient/log/
09:07:32 <masahito> The next milestone is RC1 which happens in next week.
09:07:40 <masahito> https://releases.openstack.org/rocky/schedule.html
09:08:17 <masahito> any comments?
09:08:29 <priteau> masahito: Did you see that my last merged patch to python-blazarclient was buggy, and unfortunately it made the 2.0.0 release.
09:08:38 <priteau> I have uploaded a fix: https://review.openstack.org/#/c/586631/
09:08:46 <priteau> We will need to release 2.0.1
09:10:15 <masahito> I see. I check it and will release 2.0.1
09:10:46 <priteau> I am very sorry for the trouble. I was working on other patches at the same time and overlooked testing this one.
09:11:03 <masahito> I'm checking whether some CLI and API works well now. If I find another bug I try to include the fix to 2.0.1.
09:11:24 <priteau> Thanks!
09:11:43 <masahito> priteau: np, we eventually fix it :-)
09:12:42 <masahito> #topic PTG planning
09:13:21 <masahito> I've already see some topic proposal in the ptg etherpad page. Thanks bertys and tetsuro !
09:13:23 <masahito> https://etherpad.openstack.org/p/blazar-ptg-stein
09:14:22 <priteau> bertys: Are you travelling on the Monday?
09:14:44 <masahito> sorry, I forgot to announce the page to openstack-dev ML. I'll do that after this meeting.
09:15:52 <bertys> priteau: yes, I may be available on Monday evening, e.g. for dinner
09:17:56 <masahito> bertys: Will you back to home on Wed. or Thu.?
09:18:18 <bertys> masahito: I am flying back on Fri.
09:19:06 <masahito> Of course, if we have another topics we can talk in another room on Wednesday or Thursday.
09:19:10 <masahito> bertys: got it.
09:19:45 <masahito> any comments?
09:20:03 <priteau> I will soon confirm whether or not I am attending the PTG
09:20:26 <tetsuro> I'm there from Mon.-Fri.
09:20:42 <masahito> priteau: Good to hear. I hope you'll be there :-)
09:20:58 <masahito> I
09:21:24 <masahito> I'll be there Mon. to Thu. or Fri., too.
09:22:08 <masahito> #topic PTL election
09:22:46 <masahito> The Stein cycle PTL nomination ends in few hours.
09:23:06 <masahito> https://governance.openstack.org/election/
09:24:22 <masahito> I'm thinking the T cycle or the S cycle is one of good time to change the PTL.
09:25:56 <priteau> You've done a lot of great work as PTL masahito :-)
09:26:01 <masahito> Of course, we're a small team. It's hard to change the PTL in every cycle.
09:26:22 <masahito> priteau: my pleasure.
09:28:33 <masahito> If you have a passion to do the position don't forget to submit your nomination :-)
09:30:06 <masahito> any comment?
09:31:16 <masahito> #topic AOB
09:31:38 <masahito> Does someone have something to share or discuss?
09:32:02 <priteau> I do.
09:32:18 <priteau> I have finished my patch series to provide soft-delete capabilities.
09:32:58 <priteau> This is extracted (and cleaned up and improved) from our Chameleon fork that we've been running in production for three years.
09:33:34 <priteau> I don't think it can be considered for Rocky, but if reviewed positively it could be merged in early Stein cycle.
09:33:48 <priteau> https://review.openstack.org/#/c/585807/
09:34:09 <priteau> I would like to request feedback on this feature. I didn't implement soft delete for compute hosts and extra capabilities as that would require deep changes to the model to make each row unique even when deleted, as described in the commit message.
09:34:53 <priteau> The patch linked above depends on four others which include small fixes.
09:35:10 <masahito> priteau: Great! Thank you for the nice patches.
09:35:33 <priteau> In addition to the discussion about soft-delete for some models, things still remaining:
09:35:46 <priteau> - a CLI tool for archiving rows, like Nova and Glance have
09:35:56 <priteau> (archiving or pruning)
09:36:02 <priteau> - release note and docs
09:36:44 <masahito> I agree the change should be in early S release. And we'll be able to discuss the soft-delete support for compute host in the PTG.
09:37:57 <priteau> Sounds good. If I cannot join physically I will participate remotely.
09:38:29 <masahito> Nice.
09:39:21 <masahito> I'm also thinking about the soft-delete feature because it's one of required feature for production deployments.
09:39:50 <masahito> For audit, troubleshooting, and etc.
09:41:04 <masahito> Speaking of soft-delete, the allocation API with terminated query parameter required the deleted data, but the blazar doesn't keep the data now.
09:41:14 <masahito> https://review.openstack.org/#/c/585698/
09:42:25 <priteau> It should keep them as long as the user doesn't delete the lease.
09:42:47 <priteau> But yes, once deleted you can't find allocations anymore.
09:43:38 * tetsuro is asking to google what "soft-delete" is, and getting to understand a little bit
09:44:20 <masahito> In current implementation, Blazar deletes the allocation data once a lease is end (meaning terminated status).
09:45:17 <masahito> tetsuro: soft-delete means don't remove real DB records from its table. Just adding a DLETED flag for the record.
09:45:47 <tetsuro> Cool explanation. You're a better teacher than google. Thanks.
09:46:25 <priteau> masahito: You're right, I see the allocations are destroyed during on_end().
09:46:35 <priteau> I have worked too much with our forked code ;-)
09:47:14 <masahito> haha :-)
09:47:35 <priteau> Since the PTG is still 1.5 months away, I would appreciate if we could start review and early discussion via Gerrit.
09:47:49 <masahito> Anyway, the soft-delete is an important feature.
09:47:51 <masahito> right.
09:49:31 <masahito> The points we should discuss in the PTG are 1. whether we should support soft-delete for compute host tables and 2. what tools are required for cloud admins.
09:50:37 <priteau> +1
09:50:52 <masahito> anything else?
09:51:09 <bertys> This is just fyi. I may be able to attend the Summit in Berlin if [1] Blazar project update session is confirmed [2] You need me as co-speaker
09:52:41 <masahito> bertys: Got it.
09:52:45 <masahito> I
09:53:55 <masahito> I've not received the announcement for the update session yet. I try to replay and share it once I've got any information.
09:54:54 <bertys> masahito: thanks
09:54:57 <masahito> last 5 mins. anything else?
09:56:23 <masahito> All right. Thanks all!
09:56:40 <priteau> Thanks everyone.
09:56:41 <tetsuro> Thanks, masa and all.
09:56:43 <priteau> Bye!
09:56:52 <masahito> See you in next week. bye
09:56:56 <masahito> #endmeeting