- Wednesday, June 28, 2017

release-team@conference.openafs.org

Wednesday, June 28, 2017< ^ >

Room Configuration

Room Occupants

GMT+0
[00:51:01] kadukoafs@gmail.com/barnowl0F10C5CB leaves the room
[00:51:12] kadukoafs@gmail.com/barnowl0F10C5CB joins the room
[12:06:53] meffie joins the room
[14:59:08] wiesand joins the room
[15:00:12] <wiesand> Hello
[15:00:51] <meffie> good evening.
[15:02:17] <wiesand> It was a quite week...
[15:02:37] <wiesand> We're still good with Linux-next, right?
[15:03:15] <meffie> i believe so
[15:04:15] <wiesand> rc7 was released, likely the final one
[15:04:30] <wiesand> any test results for 1.6.21pre1
[15:04:31] <wiesand> ?
[15:05:18] <meffie> i've not gotten any test reports, seems everyone is pretty busy. (i've been running pre1 without problems)
[15:06:04] <wiesand> So the plan is still to release 1.6.21 next week, w/o further changes.
[15:06:16] <meffie> ok, sounds good.
[15:06:57] <wiesand> It would be nice to have hte ubik fies ready then for an early 1.6.22pre1...
[15:07:29] <meffie> agreed.
[15:08:03] <wiesand> Any other major topics for .22 ?
[15:08:05] <meffie> marcios patch from yesterday is just a sanity check, not a bug fix.
[15:08:14] <wiesand> (besides arm ;-)
[15:08:18] <meffie> arm?
[15:08:57] <wiesand> arm.
[15:09:02] <meffie> oh, do you mean the arm64 (aka ararch64) patches?
[15:09:09] <wiesand> you put arm64_linux26 on the wish list a while ago
[15:09:10] <jhg> greetings! I am around for about 10 more minutes if there is any Linux stuff to discuss or actions for my plate
[15:10:01] <meffie> ah, yes, arm64_linux26 is still on the wish list. we've been building and running with those patches for a while now.
[15:10:27] <meffie> it would be nice to make them available to a wider audience at some point.
[15:10:58] <meffie> good morning jhg.
[15:11:08] <wiesand> Hello Joe. We hope Linux is currently boring after your changes for 4.12?
[15:11:16] <jhg> good morning, meffie_!
[15:11:25] <jhg> wiesand: it has been... =)
[15:11:32] <meffie> good news
[15:12:36] <meffie> linux-daily-x86_64-builder has been happy for a long time now.
[15:13:45] <wiesand> Things have improved a lot at the linux front in the last year. Tahnks.
[15:14:23] <meffie> thanks wiesand!
[15:14:26] <wiesand> re arm, some will say that should rather be left for 1.8
[15:15:07] <wiesand> I'd tend to agree, unless the changes are quite simple and low risk...
[15:15:20] <wiesand> I haven't looked at them yet
[15:15:43] <meffie> the changes for arm are really rather minor. mostly build system stuff.
[15:16:32] <meffie> arm64 that is.
[15:16:47] <wiesand> In that case, 1.6.22 is probably the window to pull them up
[15:16:50] <meffie> but, yes, they need to be merged on master first.
[15:17:25] <wiesand> Ah, I wasn't aware of that little complication...
[15:17:40] <meffie> ben might want to wait until 1.8.0 is out, and i could agree with that.
[15:18:06] <meffie> it's the old carrot and the stick approach.
[15:18:29] stephan.wiesand joins the room
[15:18:40] stephan.wiesand leaves the room
[15:18:58] <wiesand> alter ego… clicked the wrong icon...
[15:19:04] <jhg> gtg. ttyl!
[15:19:22] <wiesand> bye joe
[15:19:40] <wiesand> yuck. we have a cooling problem.
[15:19:43] <meffie> since the arm64 patches were pushed to gerrit, it's much easier to build them these days. i can be done on a raspberry pi 3 since suse released a 64 bit linux for the pi
[15:20:10] <meffie> oh, no. sorry to hear about your cooling problem.
[15:21:58] <meffie> btw, made some progress on Gaja's bug report. hopefully we will have a root cause and a fix soon.
[15:22:12] <meffie> (for 1.8.0)
[15:23:00] <wiesand> great news
[15:23:44] <wiesand> `is there more to discuss today?
[15:23:48] <mvita> I will miss the next 2 meetings.
[15:23:53] <meffie> with that, motion to adjourn?
[15:24:04] <meffie> with that == great news
[15:24:17] <wiesand> mark: enjoy your vacation I hope?
[15:24:25] <mvita> I hope!
[15:24:31] <meffie> yes, enjoy!
[15:24:47] <mvita> thank you very much
[15:24:55] <wiesand> Let's adjourn then. Thanks a lot everyone.
[15:25:12] <meffie> thanks. may your pager be silent.
[15:25:30] <wiesand> I don't carry a pager :)
[15:25:52] <wiesand> Nor a company phone :) :)
[15:25:58] <meffie> lol. then may your inbox be empty :)
[15:26:26] <wiesand> Ouch, that hits the sore spot….
[15:26:41] <wiesand> Bybye now.
[15:26:44] wiesand leaves the room
[15:26:44] <mvita> bye
[17:11:21] <kadukoafs@gmail.com/barnowl0F10C5CB> Re Gaja's thing, I got some "interesting" behavior in my local test
setup but I am not convinced that it is the same thing that she sees.
Namely, I can sometimes log in and not have a homedir, but I also have
neither tickets nor tokens, and a "lost contact with vlserver" message
is logged at the same time.
I still need to check my hypothesis that this is just caused by trying
to log in "too soon" and networking or something is not fully up yet.
[17:31:26] <kadukoafs@gmail.com/barnowl0F10C5CB> meffie: that seems pretty different from what Gaja sees, right?
[17:34:55] <meffie> i saw, i had a token, but always got ENOENT for one of the components of the path to the home directory.
no amount of fs flush*/fs check* helped. then cleared the linux direntries and that cleared the ENOENT errors, even after logging out and back in.
[17:36:44] <meffie> so, i assume it's related to the negative dentry cache.
[17:36:58] <kadukoafs@gmail.com/barnowl0F10C5CB> Right.
[17:39:27] <meffie> e.g. gerrit topic topic:linux-negative-dcache
[17:41:57] <kadukoafs@gmail.com/barnowl0F10C5CB> But you've only seen it on Gaja's machine, not in your local lab?
(That is, we can't just add debugging diagnostics in the changed
codepaths from that topic?)
[17:51:45] <meffie> i've only seen it on gaja's machine. not sure why i can't see it on my machines. gaja granted me sudo on her test machine, so i think you are right, i'll need to add some more fstracing/syslogging in that error of code and reload the kernel module.
[23:24:53] meffie leaves the room