[OpenAFS-devel] release-team meeting minutes 2014-04-02

[Stephan Wiesand]

Log: =
http://conference.openafs.org/release-team@conference.openafs.org/2014-02-=
04.txt

Participants:

* Andrew Deason
* Ben Kaduk
* Daria Brashear
* Marc Dionne
* Mike Meffie
* Stephan Wiesand

=3D=3D Linux news =3D=3D

Marc tested the current 1.6 head against the final Linux 3.14, and we're =
still good. The official merge window for the next Linux release is 2 =
weeks, thus it's likely that we'll be able to address any Linux 3.15 =
issues in 1.6.8 if needed.

=3D=3D Problem reports =3D=3D

Linux stack overflows (RT #131831):

Not uncommon. Probably not a regression and rather due to recent ext3 =
using more stack space. Saving some stack space on our side is likely to =
be the solution. Gerrit 10964 by Chas [meanwhile merged on master] is at =
least a start. Not a blocker for 1.6.8, but it would be good to address =
if we feel confident about the change(s).

=3D=3D 1.6.7 security release =3D=3D

After his whining at EAKC, Stephan will probably be more involved in the =
1.6.x security releases. A CVE has been obtained, thus the clock is =
ticking. 1.6.7 will have its own branch, which will have to be merged =
into 1.6.x behind gerrit's back. Stephan asks whether such branches =
could be handled like all the other's, with just restricted read access =
before going public. Alas, this seems to be impossible because =
git.openafs.org is actually the same tree.

=3D=3D 1.6.8 release =3D=3D

Andrew and Ben volunteer for providing FBSD/Solaris binaries for pre1. =
[done, uploaded and announced, together with SuSE/RedHat binaries from =
Christof and Stephen].

=3D=3D Testing =3D=3D

The lack of testing, and the missing guarantees for what is tested =
before a release, were brought up at EAKC. Several ideas were brought up =
for improving this:

* Mike's Robo Test, which is a smoke test at this point (see =
https://github.com/openafs-contrib/openafs-robotest)
* Regression tests for known bugs, simulating all known operations etc.
* Reviving the -testers list, because potential testers may be shy of =
posting on -info.=20
* A group of volunteers we rely on to test prereleases.

Daria reports that a while ago there was a similar idea of having known =
testers endorsing releases for "their" platform.

Andrew to approach potential volunteers. Once those are found, the =
discussion should be continued, including them.

=3D=3D Next stable branch(es) =3D=3D

Mike & Stephan would prefer the successor of 1.6 to be 1.8, if there's =
no general change to handling/naming branches.

Daria proposes keeping the Windows/Others branches unlinked, allowing =
different release schedules for them, while all changes should still go =
through the master branch. Naming should then change to something like =
"Portable OpenAFS" vs. "Windows OpenAFS", rather than "OpenAFS 1.8" vs. =
"OpenAFS 1.9".

Stephan would be willing to care for 1.6 and the next stable series if =
asked to.

=3D=3D Meeting time slot =3D=3D

The regular participants are generally ok with the current slot (4pm in =
central Europe, 10am at the US east coast), but having the meeting one =
hour later would be an option if it would allow others to participate =
too. Unfortunately, both known candidates didn't show up although this =
discussion was held during the potential new slot and the agenda was set =
up to make that likely. Stephan to bring this up on the list.

--=20
Stephan Wiesand
DESY -DV-
Platanenenallee 6
15738 Zeuthen, Germany