[ewg] EWG/OFED meeting minutes for Apr 6, 09

[Tziporet Koren]

> This is OFED meeting agenda for today (April 6):
Meeting Summary:
==============
a. OFED 1.4.1 release status: RC3 done, decided on RC4 date and bugs
that should be fixed for the release.
b. Reviewed the changes in OFED build process that will take place in
OFED 1.5, following the BOF we had in Sonoma 
c. Reminder: OFED roadmap is available on the web:
http://www.openfabrics.org/txt/woody/roadmap.txt
d. All participants are requested to review bugzilla bugs and update
status (if they are still reproduced)

Details:
======
> a. OFED 1.4.1 release status:
> Reminder for OFED 1.4.1 schedule:
* RC1 & RC2 - done 
* RC3 - Apr 6
* RC4 - Apr 20
* GA - Apr 23
> Critical bugs :
We reviewed the high priority bugs and decreased the list to 11 bugs:
bug_id	bug_severity	op_sys		assigned_to
short_short_desc
1538	blocker		Other		swise at opengridcomputing.com
how to deal with /usr/src/ofa_kernel/include/linux/autoconf.h?
1578	blocker		All		swise at opengridcomputing.com
ofa_kernel/include/linux/autoconf.h undefines macros
1589	critical		RHEL 5
jon at opengridcomputing.com	FRMR registration errors logged by cxgb3
during NFSRDMA iozone runs
1571	critical		RHEL 5		vu at mellanox.com
nfsrdma server crash @test5 connectathon basic test,
1528	major		RHEL 5		jackm at mellanox.co.il
IPoIB get stack when running Hadoop application.
1529	major		RHEL 5		jackm at mellanox.co.il
Opensm cannot be stopped following openib failure.
1545	major		Other		jackm at mellanox.co.il
Performance degradation in ofed 1.4.1 in TCP BW for some packets size
1570	major		SLES 10	Jeffrey.C.Becker at nasa.gov
rnfs-utils fails to compile on sles10sp2
1579	major		RHEL 5		jsquyres at cisco.com
OpenMPI-1.3.1-1: segfault during close
1581	major		Other		vlad at mellanox.co.il
Unable to uninstall OFED1.4 due to dependencies on tgt and
scsi-target-utils
1587	major		Other		yosefe at voltaire.com
Adding and removing pkey from partition.conf cause to machine to hung

> b. OFED 1.5 release process:
> Make sure all participants are inline with decisions we took in Sonoma
> in the OFED/distro discussion:
> 1) Identify all user components which are not packaged as tarballs and
> notify them that it will be a condition of inclusion in OFED 1.5 that
> they are packaged as tarballs.
> 2) Ensure that each of the tarballs for OFED 1.5 has a unique version
> number, and that version numbers are updated appropriately as the
> tarball contents change.  
> 3) Ensure that each use level component that depends on specific
> kernel feature actually checks for the existence of that kernel
> feature, and politely declines to compile/install if that component is
> not available.
> 4)  Investigate making the OFED install work as non-root - Doug has
> offered tutorials using BUILD_ROOT and chroot mechanisms to do
> non-root builds.
5) bugzilla status: we have too many old bugs in bugzilla that no one
know what is their status.
    Thus we request that any one that opened a bug will see if it still
reproduced.
    If not mark as fixed and if yes change version number to 1.4.1

c. Bill raised the question how are we going to address MPI requests
that were raised in Sonoma.
Tziporet updated that API for Memory registration notification is being
derived by Jeff S. in MPI forum.
Maybe in next meeting Jeff will be on the call and he will be able to
update us
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/ewg/attachments/20090407/ed7361e3/attachment.html>