[OpenAFS-devel] release-team meeting minutes 2013-12-18

[Stephan Wiesand]

Log: =
http://conference.openafs.org/release-team@conference.openafs.org/2013-12-=
18.txt

Participants:

* Andrew Deason
* Derrick Brashear
* Marc Dionne
* Mike Meffie
* Stephan Wiesand

=3D=3D Linux 3.13 =3D=3D

Marc tested the head of 1.6.x on 3.13-rc4+ this morning, and we're still =
good.

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

The report in =
https://lists.openafs.org/pipermail/openafs-info/2013-December/040327.html=
 lacks the details required to understand what's going on.

Andrew and Mike encountered a problem where SELinux causes delays when =
accessing uncached AFS data. A proposed fix was merged on master and =
pulled up to gerrit 10598 during the meeting. The commit message also =
explains well what's going on. Stephan believes this problem has existed =
at least since the early EL5 days. Marc points out that it uses a =
different code path within the kernel. Decision: this comes a bit late =
for pre2.

Marc reported (already last week) about a problem where McAfee antivirus =
software prevents the client from accessing his disk cache. This seems =
to be new with the RHEL 6.5 kernel. The workaround is to exclude the =
disk cache from virus scanning. A note should be added to the release =
notes and the announcement. Mike proposes posting this on openafs-info. =
It's not clear whether this can be fixed or worked around in the openafs =
code.

=3D=3D 1.6.6pre2 =3D=3D

The EL 6.5 fix (10578) and the pending changes for Mavericks support =
(10541/2/3/8/9), as well as the "make pre2" and NEWS update changes, =
received sufficient review and were merged during the meeting. Tarballs =
were uploaded to =
/afs/.grand.central.org/software/openafs/candidate/1.6.6pre2 . We'll =
wait for some positive smoke test results before tagging, but builds not =
depending on a tag could start now. [update: Derrick pushed the tag]

There was some discussion on 10548. Replacing a macro with a static =
inline function this way may not generally work, and a cleaner solution =
should be found. Andrew volunteered to look into it. But at least for =
1.6.6pre2, this darwin-only change should do.

=3D=3D 1.6.5.2 =3D=3D

Stephan to work on it right after the 1.6.6pre2 smoke tests. [update: =
the list of changes foreseen to go in can be found under =
http://gerrit.openafs.org/#q,status:open+project:openafs+branch:openafs-st=
able-1_6_5_x,n,z ]

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