Hi Kees,
Is changelogmaxage=-1 after the upgrade ?
would you send a full pstack when it hangs ? If pthread_rwlock_wrlock is
trim_changelog then you may hit another flavor of [1] (without known
reason).
regards
thierry
On 5/12/21 2:40 PM, Kees Bakker wrote:
Sorry to revive an old thread. I'm getting deadlocks again. See
below
On 20-04-2020 15:16, thierry bordaz wrote:
> [...]This is a known bug [1].
> With the same bug there are two deadlock scenario but only one is
> fixed (for example in slapi-nis-0.56.4-1 [2]).
> A fix for the second one is under tests.
>
> At the moment I would recommend the workaround [3]. The drawback is a
> growth of the retroCL database but unless you have a very high rate
> of update it should not be a concern.
>
> [1]
https://bugzilla.redhat.com/show_bug.cgi?id=1751295
> [2]
https://koji.fedoraproject.org/koji/buildinfo?buildID=1457771
> [3]
https://bugzilla.redhat.com/show_bug.cgi?id=1751295#c5
>
> best regards
> thierry
>
I followed the recommendation to set nsslapd-changelogmaxage to -1.
The system
has been running successfully for a year.
Recently I upgraded all packages in this CentOS7 system. Ever since
that moment
the server is quite unusable.
[root@linge ~]# gdb -ex 'set confirm off' -ex 'set pagination off' -ex
'thread apply all bt full' -ex 'quit' /usr/sbin/ns-slapd `pidof
ns-slapd` |& grep '^#0.*lock'
#0 0x00007f96eaf3939e in pthread_rwlock_wrlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf39184 in pthread_rwlock_rdlock () at
/lib64/libpthread.so.0
#0 0x00007f96eaf3939e in pthread_rwlock_wrlock () at
/lib64/libpthread.so.0
[root@linge ~]# rpm -qa 389\*
389-ds-base-libs-1.3.10.2-10.el7_9.x86_64
389-ds-base-1.3.10.2-10.el7_9.x86_64
389-ds-base-debuginfo-1.3.10.2-10.el7_9.x86_64
[root@linge ~]# rpm -qa slapi\*
slapi-nis-0.56.5-3.el7_9.x86_64
[root@linge ~]# rpm -qa centos-release
centos-release-7-9.2009.1.el7.centos.x86_64
Are there any new hints to avoid the deadlock?