- Friday, March 19, 2021

release-team@conference.openafs.org

Friday, March 19, 2021< ^ >

kaduk@jabber.openafs.org/barnowl has set the subject to: openafs release team

Room Configuration

Room Occupants

GMT+0
[12:20:48] meffie joins the room
[13:09:18] mbarbosa joins the room
[13:16:18] mbarbosa leaves the room
[13:16:51] mbarbosa joins the room
[13:57:08] Cheyenne joins the room
[15:48:13] yadayada joins the room
[15:48:22] mvita2 leaves the room
[15:58:24] <meffie> good day
[15:58:33] wiesand joins the room
[15:58:59] <wiesand> Hello
[15:59:08] <yadayada> hello all
[15:59:17] <Cheyenne> Hello
[15:59:42] <kaduk@jabber.openafs.org/barnowl> greetings
[15:59:56] <kaduk@jabber.openafs.org/barnowl> a bit distracted today by something else that is timely, unfortunately
[16:02:18] <wiesand> I pulled up the Linux-5.12 changes today. There was indeed one context skew. I haven't checked yet from which change that results and whether we''re sure we don't want it.
[16:03:10] <Cheyenne> If I remember correctly the context skew was fairly minor
[16:03:10] <wiesand> I also deployed a test build with everything pulled up to 1_8_x so far included.
[16:03:34] <wiesand> It was an additional function master has but 1_8_x doesn't.
[16:04:42] <wiesand> Next step is to merge what's in the test build - if review is sufficient and satisfactory. Then update NEWS.
[16:05:02] <wiesand> I'm wondering about Andrew's comment on 14520.
[16:06:36] <wiesand> I thought that by convention the symbols for EOL'ed older minor versions are not added in these changes?
[16:07:00] <kaduk@jabber.openafs.org/barnowl> My understanding (that I should put in gerrit) is that
[16:07:01] <Cheyenne> Ah.. yes -- it's the "inserting something at the end of the file"
[16:07:44] <kaduk@jabber.openafs.org/barnowl> when we make an X.0 file, we take all the (X-1).Y that are in the
latest version of the X-1 branch at that time, but if new minor
versions of (X-1) are added afterward, they are only on the (X-1)
branch and not retroactively added to X versions
[16:09:18] <wiesand> Thanks. I added that as a comment.
[16:10:07] <kaduk@jabber.openafs.org/barnowl> I added a comment too.  Hopefully we are consistent with each other :)
[16:10:52] <wiesand> Thanks :)
[16:11:20] <wiesand> That's about what I have for today.
[16:12:02] <wiesand> Anything else regarding 1.8.8?
[16:12:24] <Cheyenne> wiesand: The context conflict is 13721 .. which just moved a function from one place to another
[16:12:47] <Cheyenne> (or some conde into a static inline function actually)
[16:13:43] <Cheyenne> afs master still builds cleanly against linux 5.13-rc3+
[16:15:00] <wiesand> Thanks for the good Linux news. Re 13721, I think I'll submit a new Linux-5.12 stack with it included - thanks for checking.
[16:16:17] <wiesand> On to 1.9/master?
[16:16:30] <Cheyenne> 13721 seems like a easy/harmless commit to include in 1.8
[16:16:55] <kaduk@jabber.openafs.org/barnowl> Agreed, verifying 13721's (backport's) correctness should be
mechanical.
[16:17:07] <wiesand> Yes. Let's try to avoid code skew if it's that cheap.
[16:18:02] <kaduk@jabber.openafs.org/barnowl> master/1.9.  Um.  I pushed a 1.9.1 tag, with a NEWS update that I had
modified and nobody else reviewed...
[16:18:25] <wiesand> sorry…
[16:18:47] <kaduk@jabber.openafs.org/barnowl> but I got pretty tired after that (last night) and didn't trust myself
to review the bigsur changes.  It sounds like they should be good to
go, based on the previous comments and Andrew's latest notes, but I
didn't push the button on it.
[16:19:01] <kaduk@jabber.openafs.org/barnowl> I didn't give you any time to review it!  All blame is on me.
[16:19:25] <Cheyenne> 14559 was something I stumbled across.. silly mistake on a strcmp
[16:20:19] <kaduk@jabber.openafs.org/barnowl> *nods*
[16:20:30] <kaduk@jabber.openafs.org/barnowl> I hope that I can get to that quickly.
[16:21:07] <kaduk@jabber.openafs.org/barnowl> It looks like I had left a few questions on 14557 (fstrace) and I
wasn't sure if I should just merge it as-is or if we wanted to tweak
anything.  (That's Mark's change.)
[16:22:09] <wiesand> earmarked 14559 for pullup
[16:22:55] <kaduk@jabber.openafs.org/barnowl> I don't think there's much else for master that is blocking on people
other than me...
[16:23:55] <kaduk@jabber.openafs.org/barnowl> I have started looking at the infrastructure again, and making notes
to prepare for moving services from the EL6 machine to the EL7 one
(both at MIT).  Perhaps the highest priority task there is moving the
builbot coordinator, which IIRC Mike is mostly tending at this point.
[16:24:51] <kaduk@jabber.openafs.org/barnowl> Last I checked in Mike couldn't log in to the new machine, but maybe
could log in to the old one?  IIUC an ssh tunnel could use the old one
as a jump host in the interim while we sort out firewall issues.
[16:24:58] <meffie> yes, the one on EL7 was running one test builder last time i looked
[16:25:24] <meffie> oh, yes, i need to try to log in to the EL7 box.
[16:26:05] <kaduk@jabber.openafs.org/barnowl> I, of course, cannot really prioritize Mike's time, but that is
something I'd be interested in hearing about, if I can help get
firewall situations/etc. sorted out.
[16:26:31] <meffie> Last login: Thu Aug  6 11:32:29 2020 from www.meffie.org <http://www.meffie.org>
RHN kickstart on 2020-05-04
Managed by Puppet
[16:26:41] <meffie> i can ssh there now.
[16:26:46] <kaduk@jabber.openafs.org/barnowl> yay!
[16:27:11] <kaduk@jabber.openafs.org/barnowl> I guess I need to figure out the HTTPS story for it, still.
[16:28:01] <meffie> kaduk: sna has a new ipv4 address. could we ask mit to add it to the firewall rules?
[16:28:12] <kaduk@jabber.openafs.org/barnowl> definitely!  Send me mail, please.
[16:28:17] <meffie> will do.
[16:29:59] <kaduk@jabber.openafs.org/barnowl> I am sure there is a well-established practice for having one machine
that DNS points to serve as a frontend for another machine that hosts
the content, whether via straight-up netcat or an apache proxy
functionality.  I just have to find it.
[16:31:37] <meffie> i dont know, but i could ask around
[16:31:41] <Cheyenne> There was a ssh proxy that I discovered a while back that was very useful
[16:32:14] <meffie> this is for https traffic
[16:32:25] <Cheyenne> I think it could also handle https
[16:32:50] <kaduk@jabber.openafs.org/barnowl> Don't feel obligated to do anything, but I'm happy to hear anything
you do happen to find out.
[16:33:22] <Cheyenne> it might have been haproxy
[16:33:57] <kaduk@jabber.openafs.org/barnowl> And, with apologies for jumping around with topics, I did push the
1.9.1 tag but I didn't make tarballs or prepare a web change.  I could
probably do that over the weekend, but am happy to give people a
chance to volunteer if anyone is interested
[16:35:50] <meffie> i could make tarballs, that's just a make dist, no?
[16:36:55] <kaduk@jabber.openafs.org/barnowl> Something like that, yeah.
[16:37:13] <kaduk@jabber.openafs.org/barnowl> I think I call make-release directly when I do it
[16:37:18] <Cheyenne> kaduk_jabberopenafsorg: haproxy was what I was using in a past life.  It worked very well
[16:37:25] <meffie> ok
[16:37:33] <kaduk@jabber.openafs.org/barnowl> Thanks Cheyenne!
[16:38:36] <Cheyenne> ended up being very simple configuration as well.
[16:38:43] <kaduk@jabber.openafs.org/barnowl> Bonus :)
[16:39:02] <wiesand> I'm calling make-release to roll tarballs. I'm afraid it's even a local version with a few minor changes I should submit back to openafs-tools.git.
[16:40:41] <kaduk@jabber.openafs.org/barnowl> Well, I think you should submit it back to openafs.git (isn't it in
build-tools/ ?)
[16:40:57] <meffie> yes, it is in build-tools
[16:41:10] <meffie> that's the one i use when making tarballs and rpms
[16:41:25] <kaduk@jabber.openafs.org/barnowl> Anyway, I think I'm about wrapped up for today.  Any other topics?
[16:42:11] <meffie> got some results back from yadav on the ansible-openafs collection, some good results and some things still to look at.
[16:42:52] <meffie> i'll put an issues or fixes in github/openafs-contrib
[16:42:54] <yadayada> right Mike still testing it will try more next week mainly client role
[16:43:03] <meffie> thank you!
[16:43:44] <meffie> that all i have. thanks everyone.
[16:44:27] <Cheyenne> nothing here
[16:44:41] <yadayada> nothing from my side too
[16:45:04] <kaduk@jabber.openafs.org/barnowl> Let's adjourn, then.
Sorry again for the slow week on my side, and have a good weekend
[16:45:29] <yadayada> thanks all
[16:45:32] <Cheyenne> you too.  Everyone stay safe and have a wonderful week
[16:45:32] <wiesand> My local changes to make-release aren't substantial. I'll submit them anyway.
[16:46:00] <meffie> happy equinox
[16:46:31] <Cheyenne> Oh.. wedding anniversary this weekend :)  40 years
[16:47:18] <kaduk@jabber.openafs.org/barnowl> Oh, wow.  Congrats, Cheyenne!
[16:47:36] <Cheyenne> Also Bach's birthday
[16:47:44] <kaduk@jabber.openafs.org/barnowl> "You have been married longer than I have been alive."  :)
[16:48:41] <Cheyenne> well.. Jethro Tull's Aqualung is celebrating 50 years.  I remember when that came out ..
[16:49:02] <wiesand> so do I… [sobs]
[16:50:24] <wiesand> Anyway, All the Best!
[16:51:17] <wiesand> Thanks a lot everyone!
[16:58:53] yadayada joins the room
[17:06:39] yadayada leaves the room
[17:36:51] meffie leaves the room
[17:53:59] yadayada leaves the room
[18:02:23] yadayada joins the room
[20:26:38] yadayada leaves the room
[22:32:37] mbarbosa leaves the room
[22:58:06] wiesand leaves the room
[23:37:55] mbarbosa joins the room
[23:39:10] mbarbosa leaves the room