- Friday, November 6, 2020

release-team@conference.openafs.org

Friday, November 6, 2020< ^ >

yadayada has set the subject to: openafs release team

Room Configuration

Room Occupants

GMT+0
[11:49:48] meffie joins the room
[12:01:39] mbarbosa joins the room
[13:01:43] mbarbosa leaves the room
[13:01:44] mbarbosa joins the room
[17:00:09] <meffie> greetings.
[17:00:20] cwills joins the room
[17:00:24] <kaduk@jabber.openafs.org/barnowl> Salutations
[17:00:43] <cwills> Hello
[17:00:47] mvita2 leaves the room
[17:00:54] wiesand joins the room
[17:01:04] <wiesand> Hi
[17:02:11] <meffie> we are on standard time now. much better
[17:03:01] <wiesand> Long evenings - I like those too.
[17:03:29] <wiesand> So, after all I did stumble over a change on the 1.8.7 wish list not yet merged on master: 14340
[17:03:49] yadayada joins the room
[17:03:57] <yadayada> Hi All
[17:04:31] <wiesand> In other news, Mike asked for another candidate. The pullup is 14427.
[17:05:07] <meffie> i have 3 other gerrits i'd like to ask about.
[17:05:09] <kaduk@jabber.openafs.org/barnowl> Shucks, and I can't even push the 'submit' button before replying "I
don't know what you mean; it looks to me like it's merged on master".
[17:05:15] <meffie> thank you for the pullup!
[17:05:48] <kaduk@jabber.openafs.org/barnowl> For 14340 it looks like I maybe wanted some tweaks in the commit
message; I don't know if we think that is worth changing or not.
[17:06:14] <kaduk@jabber.openafs.org/barnowl> (If we are okay with the commit message as-is I can push the 'rebase'
button to give buildbot another crack)
[17:06:17] <wiesand> I'm not pressing for it, just mentioning
[17:07:21] <meffie> I think it would be good to have.
[17:08:07] <kaduk@jabber.openafs.org/barnowl> It is something we should have on 1.8.x, yes.  Less clear that it
specifically needs to be 1.8.7
[17:08:16] <meffie> concur
[17:09:05] <wiesand> Not much other news from my side I'm afraid. Pulled up a couple more changes after the last meeting but stopped because buildbot was a bit sick. Meanwhile those have been verified and reviewed, so can probably soon be merged away. But still a few changes on the wish list to do.
[17:09:25] <meffie> should we include the following in 1.8.7: 13383, 14416, 14208 ?
[17:09:26] <wiesand> Noted 14340 could wait for 1.8.8 if needed.
[17:09:36] <kaduk@jabber.openafs.org/barnowl> gerrit has been feeling more sluggish than usual, too, in addition to
the buildbot woes
[17:10:08] <meffie> those can wait for 1.8.8 if needed.
[17:10:28] <wiesand> Haven't noticed yet. But maybe that's because my RTT is way beyond your's anyway.
[17:11:00] <wiesand> 13383, 14416, 14208: please hang on
[17:13:30] <kaduk@jabber.openafs.org/barnowl> BTW, Mark: you can fix the braces in 13801 now since 13800 is merged
[17:16:56] <wiesand> 13383, 14416 look reasonable at first glance - put them on the list
[17:17:06] <meffie> thanks
[17:17:32] <wiesand> 14208, according to my records, should already be merged on 1.8.x (14326)
[17:18:02] <meffie> oh! let me double check :)
[17:19:39] <meffie> gah, it's never easy to find these.
[17:21:10] <wiesand> Any other 1.8.x topics? Cheyenne, any news on the latest Linux rc?
[17:21:15] <kaduk@jabber.openafs.org/barnowl> git log origin/openafs-stable-1_8_x, search for 14208, finds commit
9947625a1d67b4ffdc0582e9081000e34be2b46b
[17:21:45] <meffie> thank you. I added a commit in 14208. "Submitted to openafs-stable-1_8_x as https://gerrit.openafs.org/#/c/14326/"
[17:22:06] <cwills> just did a master build against latest pull from linux repo .. that is clean.  doing 1.8.x now
[17:23:02] <wiesand> Thanks!
[17:23:04] <meffie> s/commit/comment/
[17:23:09] <wiesand> NB git log --grep=/14208 origin/openafs-stable-1_8_x
[17:23:33] <kaduk@jabber.openafs.org/barnowl> ooh, fancy
[17:23:41] <meffie> nice, thank you.
[17:23:58] <yadayada> 14416 will surely help.
[17:25:01] <meffie> (gerrit needs a better way to track changes between branches in my opinion:)
[17:25:06] <wiesand> Yes, and no path conflicts :) Low hanging fuit
[17:25:41] <kaduk@jabber.openafs.org/barnowl> I think maybe newer versions of gerrit let you use the same change-id
on multiple branches, which presumably comes with better tracking for
that.
[17:25:47] <kaduk@jabber.openafs.org/barnowl> But I forget which version(s) are relevant for that
[17:26:22] <wiesand> It's a real pity that early gerrit had issues with duplicate change-ids between branches and our rules thus still require a different one per branch.
[17:26:43] <kaduk@jabber.openafs.org/barnowl> We could presumably change our rules if there's not a technical reason
anymore
[17:27:03] <wiesand> If I just were 100% sure...
[17:27:50] <meffie> i feel we'd want a "parent id"
[17:28:10] <wiesand> but the "—grep=/<gerrit-number>" is fairly convenient I think
[17:28:34] <kaduk@jabber.openafs.org/barnowl> yes (I think I just learned about it today, so thank you)
[17:28:36] <meffie> only if it is already merged tho, right?
[17:28:59] <kaduk@jabber.openafs.org/barnowl> yes
[17:29:01] <wiesand> right
[17:29:21] <cwills> report: 1.8.x built cleanly as well
[17:29:26] <wiesand> but I fail to see how a common change-id or a parent-id would help here
[17:29:34] <wiesand> Thanks Cheyenne!
[17:30:29] <kaduk@jabber.openafs.org/barnowl> I think the right way to get cross-branch tracking is in gerrit's
database, but we are currently stymieing that by changing change-id
when pulling up to the stable branch
[17:31:44] <meffie> i see. thanks
[17:32:32] <wiesand> If we're really sure it's safe, we can forego generating a new change-id when pulling up. Actually it's one less step to do (or forget).
[17:33:11] <kaduk@jabber.openafs.org/barnowl> I expect that the gerrit release notes would talk about this, but
haven't had a chance to look through them all.  (hint hint)
[17:34:34] <wiesand> (and yay, expanded my vocabulary again - "to stymie" was completely unknown to me :)
[17:35:06] <meffie> ha
[17:35:49] <wiesand> On to 1.9/master then? Saw a flurry of activity there lately.
[17:35:50] <kaduk@jabber.openafs.org/barnowl> (I think I spelled the "-ing" version wrong, of course.)
[17:36:08] <kaduk@jabber.openafs.org/barnowl> Activity, yes, though maybe not anything terribly exciting
[17:36:24] <kaduk@jabber.openafs.org/barnowl> Mostly just trying to finish up the stuff I talked about last week
[17:37:20] <kaduk@jabber.openafs.org/barnowl> But yeah, a bunch of stuff got merged
[17:38:18] <meffie> excellent thank you!
[17:38:31] <kaduk@jabber.openafs.org/barnowl> For, I guess, Cheyenne and Mark I'll call out that I left a comment on
14417 (return errors in d_revalidate) since I am not sure if we're
changing behavior in one case
[17:38:41] <kaduk@jabber.openafs.org/barnowl> (or whether such a change would be desired)
[17:40:54] <meffie> 14417 is probably for 1.8.7 since it was hit on current rhel kernels
[17:43:40] <cwills> will review comment..
[17:44:36] <wiesand> if it does change behavior, that would be a problem IMO
[17:44:53] <kaduk@jabber.openafs.org/barnowl> Anyway, my status for master/1.9 is otherwise just "keep plugging
away"; were there other topics people wanted to discuss?
[17:45:14] <cwills> thanks for the feedback on the audit-enhancement stuff.
[17:45:51] <kaduk@jabber.openafs.org/barnowl> You're welcome, and thank you for working on it; it's a good thing to
have
[17:45:56] <meffie> i have nothing more.
[17:46:27] <wiesand> No further topics here. And I have to do some shopping - I'm literally out of food.
[17:46:27] <yadayada> I am testing bigsur changes. Intial tests looked fine, we are running our stress test on it, saw some issue with @sys looking into it
[17:46:51] <wiesand> Ah, Big Sur, right.
[17:47:04] <meffie> thanks yadav
[17:47:48] <kaduk@jabber.openafs.org/barnowl> SNA folks do you know if Mark is "actually here"?  If not I can just
push the 13801 update that I made locally before realizing it was the
one I had just told him about...
[17:48:22] <meffie> i was just talking to him about 1380x before this meeting.
[17:48:45] <kaduk@jabber.openafs.org/barnowl> Ah, thanks.  I will wait!
[17:49:01] <meffie> looks like "mvita" is just a stale connection?
[17:49:30] <meffie> but hes it is updating that stack.
[17:49:40] <meffie> er, he is
[17:51:26] <meffie> motion to adjourn so wiesand can go shopping?
[17:51:36] <kaduk@jabber.openafs.org/barnowl> seconded
[17:51:44] <meffie> have a good weekend all!
[17:51:47] <cwills> 3rd
[17:52:02] <kaduk@jabber.openafs.org/barnowl> Thanks everyone!
[17:52:12] <kaduk@jabber.openafs.org/barnowl> Happy shopping Stephan!
[17:52:48] <wiesand> NB:
During upload by pushing to refs/for/* or refs/heads/*, Gerrit will try to find an existing review the uploaded commit relates to. For an existing review to match, the following properties have to match:
    •    Change-Id
    •    Repository name
    •    Branch name
The following applies in the different scenarios:
    •    Create a new change
If no matching review is found, Gerrit will create a new change for review.
[17:53:07] <wiesand> (from https://gerrit-review.googlesource.com/Documentation/user-changeid.html)
[17:53:38] <wiesand> I guess it is safe to keep change-ids over pullups.
[17:54:01] <kaduk@jabber.openafs.org/barnowl> That's the docs for the latest version of gerrit; it's probably safer
to look at "our" docs
[17:54:29] <kaduk@jabber.openafs.org/barnowl> https://gerrit.openafs.org/Documentation/user-changeid.html
[17:54:54] <kaduk@jabber.openafs.org/barnowl> Which seems to say the same thing, so yes, looks safe
[17:55:28] <wiesand> Looks identical at first glance.
[17:55:48] <kaduk@jabber.openafs.org/barnowl> Thank you for checking!
[17:55:55] <kaduk@jabber.openafs.org/barnowl> (Go shopping already!)
[17:56:32] <wiesand> Thanks a lot everyone! Off now to go for some fresh mushrooms and other vegetables, cream, … :)
[17:58:56] <yadayada> Thanks all
[18:02:57] yadayada leaves the room
[18:08:41] meffie leaves the room
[18:42:46] mbarbosa leaves the room
[18:43:35] mbarbosa joins the room
[21:44:29] mbarbosa leaves the room