Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
Thanks, Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration
Quanah Gibson-Mount wrote:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
All tests passed, ubuntu 8.10alpha x86_64 dualcore
Thanks, Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc
Zimbra :: the leader in open source messaging and collaboration
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Sep 30, 2008, at 22:59 , Quanah Gibson-Mount wrote:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
All tests except for test050 with HDB pass. test050 with BDB works, interestingly. This is on OS X 10.5.5/Intel on a Core 2 Duo.
jens
Quanah Gibson-Mount quanah@zimbra.com writes:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
test001 still hangs with db4.7 on openSuSE-11.0 x86_64 single core. the last line of strace:
:~> strace -ff -p 7671 Process 7671 attached - interrupt to quit futex(0x2ae2021ba924, FUTEX_WAIT, 1, NULL) = ? ERESTARTSYS (To be restarted)
-Dieter
--On October 1, 2008 2:47:42 PM +0200 Dieter Kluenter dieter@dkluenter.de wrote:
Quanah Gibson-Mount quanah@zimbra.com writes:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
test001 still hangs with db4.7 on openSuSE-11.0 x86_64 single core. the last line of strace:
Did you apply the patch from Oracle that Howard forwarded to the ITS?
--Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration
Quanah Gibson-Mount quanah@zimbra.com writes:
--On October 1, 2008 2:47:42 PM +0200 Dieter Kluenter dieter@dkluenter.de wrote:
Quanah Gibson-Mount quanah@zimbra.com writes:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
test001 still hangs with db4.7 on openSuSE-11.0 x86_64 single core. the last line of strace:
Did you apply the patch from Oracle that Howard forwarded to the ITS?
No, there is only 1 patch available at oracle's website and as far as I can see ITS #5707 doesn't provide any patch. Could you provide a URL to the patch to apply?
-Dieter
"Dieter Kluenter" dieter@dkluenter.de writes:
Quanah Gibson-Mount quanah@zimbra.com writes:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
test001 still hangs with db4.7 on openSuSE-11.0 x86_64 single core. the last line of strace:
:~> strace -ff -p 7671 Process 7671 attached - interrupt to quit futex(0x2ae2021ba924, FUTEX_WAIT, 1, NULL) = ? ERESTARTSYS (To be restarted)
I don't know wether the kernel has any influence on this behaviour, but if it does, the OpenSUSE default kernel is a smp version, even for single core units.
-Dieter
Dieter Kluenter wrote:
Quanah Gibson-Mount quanah@zimbra.com writes:
--On October 1, 2008 2:47:42 PM +0200 Dieter Kluenter dieter@dkluenter.de wrote:
Quanah Gibson-Mount quanah@zimbra.com writes:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
test001 still hangs with db4.7 on openSuSE-11.0 x86_64 single core. the last line of strace:
Did you apply the patch from Oracle that Howard forwarded to the ITS?
No, there is only 1 patch available at oracle's website and as far as I can see ITS #5707 doesn't provide any patch.
It's the base-64 encoded attachment in Followup 3.
Could you provide a URL to the patch to apply?
Unfortunately Oracle does not provide this patch on their web site.
Ciao, Michael.
Quanah Gibson-Mount quanah@zimbra.com writes:
--On October 1, 2008 2:47:42 PM +0200 Dieter Kluenter dieter@dkluenter.de wrote:
Quanah Gibson-Mount quanah@zimbra.com writes:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
test001 still hangs with db4.7 on openSuSE-11.0 x86_64 single core. the last line of strace:
Did you apply the patch from Oracle that Howard forwarded to the ITS?
Success! After applying the patch from http://www.openldap.org/its/index.cgi/Incoming?id=5707;page=5 followup 3
-Dieter
test008 missed a couple entries:
+dn: cn=Manager,dc=example,dc=com +dn: cn=All Staff,ou=Groups,dc=example,dc=com
presumably due to:
bdb(dc=example,dc=com): previous transaction deadlock return not resolved
testrun directory: https://www.nbcs.rutgers.edu/~richton/test008fail-20081001.tar.gz
On Tue, 30 Sep 2008, Quanah Gibson-Mount wrote:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
Thanks, Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc
Zimbra :: the leader in open source messaging and collaboration
Aaron Richton wrote:
test008 missed a couple entries:
+dn: cn=Manager,dc=example,dc=com +dn: cn=All Staff,ou=Groups,dc=example,dc=com
presumably due to:
bdb(dc=example,dc=com): previous transaction deadlock return not resolved
Thanks for the report. Looking into it now.
testrun directory: https://www.nbcs.rutgers.edu/~richton/test008fail-20081001.tar.gz
On Tue, 30 Sep 2008, Quanah Gibson-Mount wrote:
Please test RE24. Issues remain with test050, but are not considered release blocking as they are not a regression.
Thanks, Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc
Zimbra :: the leader in open source messaging and collaboration
Aaron Richton wrote:
test008 missed a couple entries:
+dn: cn=Manager,dc=example,dc=com +dn: cn=All Staff,ou=Groups,dc=example,dc=com
presumably due to:
bdb(dc=example,dc=com): previous transaction deadlock return not resolved
Aaron, can you pick up new back-bdb/filterindex.c and search.c from CVS HEAD. You seem to be the only one who can reliably reproduce these issues so far.
On Wed, 1 Oct 2008, Howard Chu wrote:
Aaron, can you pick up new back-bdb/filterindex.c and search.c from CVS HEAD. You seem to be the only one who can reliably reproduce these issues so far.
These worked OK for about 14 hours. Eventually, the machine started running out of memory, but I'm pretty sure that's a kernel bug. I patched, rebooted and will keep test008 on a loop today.
--On Wednesday, October 01, 2008 7:15 PM +0200 Dieter Kluenter dieter@dkluenter.de wrote:
Did you apply the patch from Oracle that Howard forwarded to the ITS?
Success! After applying the patch from http://www.openldap.org/its/index.cgi/Incoming?id=5707;page=5 followup 3
I've added the patch to the build/ directory in HEAD and 2.4, along the lines of what we've done for these types of patches in the past.
--Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration
--On Thursday, October 02, 2008 10:13 AM -0400 Aaron Richton richton@nbcs.rutgers.edu wrote:
On Wed, 1 Oct 2008, Howard Chu wrote:
Aaron, can you pick up new back-bdb/filterindex.c and search.c from CVS HEAD. You seem to be the only one who can reliably reproduce these issues so far.
These worked OK for about 14 hours. Eventually, the machine started running out of memory, but I'm pretty sure that's a kernel bug. I patched, rebooted and will keep test008 on a loop today.
Multiple runs passed for me with these patches (linux 64-bit)
--Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration
On Thu, 2 Oct 2008, Quanah Gibson-Mount wrote:
Multiple runs passed for me with these patches (linux 64-bit)
At this point, I was able to:
(a) replicate the issue on Fedora 9 x86_64
and
(b) have not yet run into the issue, once the patch was applied, on Solaris 9 sparc nor Fedora 9 x86_64.
I'll let things keep going, but I think it's OK at this point.
Quanah Gibson-Mount wrote:
--On Wednesday, October 01, 2008 7:15 PM +0200 Dieter Kluenter dieter@dkluenter.de wrote:
Did you apply the patch from Oracle that Howard forwarded to the ITS?
Success! After applying the patch from http://www.openldap.org/its/index.cgi/Incoming?id=5707;page=5 followup 3
I've added the patch to the build/ directory in HEAD and 2.4, along the lines of what we've done for these types of patches in the past.
There are two identical files now there. I guess file name db.4.7.52.patch is a typo.
Ciao, Michael.
--On Friday, October 03, 2008 12:27 AM +0200 Michael Ströder michael@stroeder.com wrote:
Quanah Gibson-Mount wrote:
--On Wednesday, October 01, 2008 7:15 PM +0200 Dieter Kluenter dieter@dkluenter.de wrote:
Did you apply the patch from Oracle that Howard forwarded to the ITS?
Success! After applying the patch from http://www.openldap.org/its/index.cgi/Incoming?id=5707;page=5 followup 3
I've added the patch to the build/ directory in HEAD and 2.4, along the lines of what we've done for these types of patches in the past.
There are two identical files now there. I guess file name db.4.7.52.patch is a typo.
No, the 4.7.52 ones have been deleted from CVS itself. The web repository doesn't seem to have caught up to that fact yet.
http://www.openldap.org/devel/cvsweb.cgi/build/Attic/db.4.7.52.patch?hideattic=1&sortbydate=0
--Quanah
--
Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration
With the patch from Oracle and the one in CVS for 4.7.25, all tests pass for me on a single core (Centrino) 32-bit in RE24, when they didn't before the latest fixes.
Thanks.