- Friday, October 12, 2018

release-team@conference.openafs.org

Friday, October 12, 2018< ^ >

Room Configuration

Room Occupants

GMT+0
[01:24:54] meffie leaves the room
[12:51:50] mbarbosa joins the room
[13:00:06] wiesand joins the room
[13:00:15] meffie joins the room
[13:00:41] <kaduk@jabber.openafs.org/barnowl> greetings
[13:01:15] <wiesand> greetings back
[13:02:25] <wiesand> no news from me this week - on top of my usual excuses, I caught a flu :-((
[13:02:48] <kaduk@jabber.openafs.org/barnowl> Oh no -- get well soon!
[13:02:59] <meffie> oh no, get well soon!!
[13:03:31] <wiesand> thanks
[13:04:19] <wiesand> so, I noticed the mojave changes were merged on master - thanks Ben
[13:04:43] <wiesand> anything else remarkable that I missed pas week?
[13:04:45] <kaduk@jabber.openafs.org/barnowl> Yeah, that seemed like a good thing to avoid having it linger in limbo
forever
[13:05:02] <meffie> thank you ben! marcio has built dmg's and put them on our download site. the wiki should have a link to them.
[13:05:09] <kaduk@jabber.openafs.org/barnowl> user claims the getcwd bug is back
[13:05:19] <meffie> :(
[13:05:52] <meffie> maybe he should try andrew's native-mounts patchset?
[13:06:45] <wiesand> the case at DESY Martin brought up also smells a bit like getcwd to me - but it's hard to get useful information (like the actual error messages the user sees…)
[13:06:48] <kaduk@jabber.openafs.org/barnowl> The listed symptoms seemed more inline with the RHEL thing where you
just couldn't *list* a directory, but you could descend through it
okay
[13:07:40] mvita joins the room
[13:07:55] <mvita> sorry for my tardiness, the power was just restored here
[13:07:56] <wiesand> I have no reports of issues with our 1.6 clients on the kernels Neil mentioned
[13:08:46] <kaduk@jabber.openafs.org/barnowl> mvita: yikes, hope that you managed the outage and such okay
[13:09:02] <wiesand> I also tried my getcwd reproducer on Ubuntu 18.04 with the 1.8 ppa client - no problems found
[13:10:11] mvita leaves the room
[13:10:12] mvita joins the room
[13:11:23] <mvita> weisand:  I think their kernel level is the key
[13:11:26] <wiesand> Mike: I thought native-mounts is not complete yet?
[13:11:37] <wiesand> which one?
[13:11:44] <mvita> 3.10 mumble
[13:11:58] <wiesand> well it's the latest EL7 series
[13:12:09] <mvita> I have not had a chance to try it out myself
[13:13:11] <meffie> it is complete enough to test and get feedback
[13:13:42] <wiesand> EL7 will always have a 3.10.0-mumble kernel version number, til the end of time (and still no EL8 in sight - I doubt it will ever happen)
[13:13:47] <meffie> we are sending packages with it for people to give us feedback (when they are back)
[13:16:55] <wiesand> I should really give those a try…
[13:17:35] <meffie> excellent, when you are feeling better of course!
[13:20:01] <wiesand> of course… right now, boy am i feeling sick…
[13:20:16] <mvita> yuk, sorry
[13:20:57] <meffie> sorry.
[13:21:16] <wiesand> do you know this feeling when your hairs hurt?
[13:21:28] <mvita> yes
[13:21:52] <wiesand> now you're with me
[13:23:36] <wiesand> but back on topic, native-mounts to me really looks like the right solution to getcwd and a handful of other problems coming up every now and then
[13:23:57] <mvita> that's the hope
[13:24:39] <wiesand> is there a runtime switch? (meaning: any hope that we can get it into 1.8.x?)
[13:25:20] <mvita> idk
[13:26:16] <wiesand> "LINUX: Introduce 'afsd -automount'" inspires some hope there is…
[13:27:02] <kaduk@jabber.openafs.org/barnowl> meffie: just an early heads-up, the stalled-fileservers thing needs to
sanity-check the value as it's passed into the kernel, in particular
to not be negative.  I'll hopefully be commenting shortly...
[13:27:54] <meffie> ok, thanks, marcio mentioned i should check for negative values, i should have listened to him.
[13:28:11] <meffie> also, i think we wanted it off by default, if i recall?
[13:28:23] <mvita> naturally
[13:28:52] <kaduk@jabber.openafs.org/barnowl> Marcio does seem to be getting the hang of things, yeah -- he's sent
some good stuff by recently
[13:29:04] <mvita> oh, sorry, I thought we were still talkng about about the native mounts
[13:29:42] <kaduk@jabber.openafs.org/barnowl> I'm the one who switched topics randomly
[13:30:57] <meffie> i have a topic switch as well, something i'm not sure is a good idea, but my be helpful for wiesand and me
[13:31:18] <kaduk@jabber.openafs.org/barnowl> I think we are still nominally in Stephan's part of the meeting
[13:31:42] <meffie> i yield to stephan then.
[13:32:19] <meffie> or did you mean, i should ask now?
[13:33:15] <kaduk@jabber.openafs.org/barnowl> Stephan should say whether you should bring it up now
[13:33:51] <wiesand> sorry, was distracted by fetching more hankies… please go ahead
[13:35:53] <meffie> ok, last week we were talking about how gerrit always sorted by last gerrit updated, which makes it harder to track new gerrits. it seems this is a problem with new versions of gerrit, from what i've seen in the gerrit issue tracking. so as a work around i wrote a script yesterday that posts the open gerrits (per branch) in descending order on wiki.openafs.org
[13:36:02] <meffie> https://wiki.openafs.org/devel/GerritsForMaster/
[13:36:18] <kaduk@jabber.openafs.org/barnowl> Hmm, that seems pretty nifty
[13:36:18] <meffie> https://wiki.openafs.org/devel/GerritsForStable/
[13:36:54] <meffie> i was going to run in in a cronjob daily, and i would only update if there  was a change.
[13:37:19] <meffie> you can check "history" to see the diff for changes.
[13:37:32] <wiesand> neat!
[13:37:41] <meffie> it's a hack, but it might be helpful?
[13:37:55] <kaduk@jabber.openafs.org/barnowl> One commit per day to the wiki should be within the bounds of
acceptable resource usage for a script
[13:38:21] <meffie> yes, i think so.
[13:39:13] <meffie> ok, lets try it for a while and see if it is helpful.
[13:39:48] <wiesand> Chances are it is. I'll use it and provide feedback on how it works out.
[13:39:58] <meffie> thank you
[13:40:14] <wiesand> Thank *you*!
[13:40:48] <wiesand> On to "Ben's part of the meeting"?
[13:41:28] <kaduk@jabber.openafs.org/barnowl> Okay.
Not a whole lot of news, though, sadly.
[13:42:39] <kaduk@jabber.openafs.org/barnowl> I merged a few things last night/this morning, and I see we've had
some updates on (e.g.) ipv6-prep and pthreaded-xstat that I haven't
looked at yet, so thank you to everyone for working on that
[13:43:27] <kaduk@jabber.openafs.org/barnowl> Mark, did you get a chance to look at the apparmor thing?
[13:43:49] <meffie> thank you. also, i think there are more reviews for rxgk-phase1 to be done soon.
[13:44:04] <mvita> I barely got started then was pulled off to several tough issues
[13:44:23] <kaduk@jabber.openafs.org/barnowl> I know how it goes :)
[13:44:26] <mvita> the last few weeks have been unusually grueling
[13:44:37] <kaduk@jabber.openafs.org/barnowl> Oh, cool, good news about rxgk-phase1
[13:45:29] <kaduk@jabber.openafs.org/barnowl> And yeah, the past few weeks have been pretty busy here, too.
[13:45:44] <kaduk@jabber.openafs.org/barnowl> Though I suppose you probably don't have a lot of openssl 1.1.1
fallout to deal with yet
[13:46:05] <mvita> no, I'm not familiar with that one
[13:46:54] <kaduk@jabber.openafs.org/barnowl> It tries to be both ABI/API-compatible with openssl 1.1.0 and enable
TLS 1.3 by default, which is kind of like trying to square the circle.
So there's a bunch of weird subtle stuff that broke, or at least had
behavior change.
[13:47:30] <meffie> yuck.
[13:47:37] <kaduk@jabber.openafs.org/barnowl> Indeed.
[13:47:55] <mvita> Is that a reference implementation for TLS 1.3?
[13:48:00] <meffie> we've been beset by a string of bad luck, like power outages, and other issues.
[13:48:11] <meffie> dealing with fallout for those
[13:48:22] <kaduk@jabber.openafs.org/barnowl> There's no single reference implementation, but we did have like six
implementations doing interop before the final RFC was published
[13:48:30] <mvita> wow
[13:49:08] <kaduk@jabber.openafs.org/barnowl> It should be pretty solid as a standalone TLS 1.3 thing, yeah.
[13:49:16] <kaduk@jabber.openafs.org/barnowl> back to openafs, I guess I should ask my usual question: anything
still sitting in gerrit for master that I should prioritzie looking
at?
[13:49:48] meffie thinking
[13:50:55] <meffie> you already got the new macos changes, i dont see anything else time critical.
[13:51:27] <kaduk@jabber.openafs.org/barnowl> Okay.  I'm on holiday this weekend, but will try to be less of a
slacker than last week.
[13:52:04] <meffie> oh, have a good holiday. i'm on my way to the Ohio Linux Fest
[13:52:31] <meffie> no openafs talks this year. maybe i should start a BOF :)
[13:53:02] <mvita> and then next week, company mtgs in Winchester VA - there may be an OpenAFS talk there ;-)
[13:53:41] <kaduk@jabber.openafs.org/barnowl> :)
[13:54:04] <kaduk@jabber.openafs.org/barnowl> Any other topics?
[13:54:16] <meffie> hopefully the openafs-contrib/openafs-robotest suite will be converted to ansible soon :)
[13:54:17] <wiesand> openstack summit here in a few weeks - did you know there's an openstack.org afs cell?
[13:54:28] <mvita> si
[13:54:34] <meffie> (for cell setup and teardown)
[13:54:54] <wiesand> ah
[13:55:20] <meffie> i didnt know openstack.org had a cell, that's great!
[13:55:31] <wiesand> I have no further topics today (shame on me again).
[13:55:52] <kaduk@jabber.openafs.org/barnowl> your main task this week is to get healthy; don't worry about the
other stuff
[13:56:02] <meffie> +1
[13:56:07] <mvita> meffie:  it's listed in the avdb
[13:56:15] <wiesand> Sounds like really good advice, thanks.
[13:57:13] <wiesand> Let's adjourn then. Thanks a lot everyone!
[13:57:15] <meffie> get well soon wiesand. have a good weekend kaduk@jabber.openafs.org
[13:57:36] <mvita> get well soon Stephan
[13:57:43] <kaduk@jabber.openafs.org/barnowl> thanks everyone, and get well soon!
[13:57:51] wiesand leaves the room
[14:28:11] mvita leaves the room
[14:40:32] mvita joins the room
[14:46:12] meffie leaves the room
[15:16:28] mbarbosa leaves the room
[16:05:35] mbarbosa joins the room
[16:48:06] mbarbosa leaves the room
[17:18:34] mbarbosa joins the room
[18:13:48] mbarbosa leaves the room
[19:38:55] mvita leaves the room
[20:14:43] mvita joins the room