[389-users] repl5_inc_waitfor_async_results

[shardulsk]

Hi,

Running 389-ds 1.1.2 on Centos 5.

We have suddenly seen repl5_inc_waitfor_async_results  errors crop up in
our error log during peak traffic hours.

The Master loses sight of Hubs and replication stalls. Most of the times it
comes right back up after a couple of mins. However we have been restarting
Hubs to make the process quicker.

tail -f errors
[08/Jan/2015:02:42:08 -0800] NSMMReplicationPlugin - agmt="cn=add -> hub1"
(hub1:2390): Simple bind resumed
[08/Jan/2015:09:04:38 -0800] - repl5_inc_waitfor_async_results timed out
waiting for responses: 0 34222
[08/Jan/2015:09:05:18 -0800] - repl5_inc_waitfor_async_results timed out
waiting for responses: 0 33499
[08/Jan/2015:09:05:37 -0800] NSMMReplicationPlugin - agmt="cn=add -> hub2"
(hub1:2390): Warning: unable to receive endReplication extended operation
response (Can't contact LDAP server)
[08/Jan/2015:09:05:37 -0800] NSMMReplicationPlugin - agmt="cn=add -> hub2"
(hub1:2390): Simple bind failed, LDAP sdk error 91 (Can't connect to the
LDAP server), Netscape Portable Runtime error -5961 (TCP connection reset
by peer.)
[08/Jan/2015:09:05:59 -0800] NSMMReplicationPlugin - agmt="cn=add -> hub2"
(hub1:2390): Simple bind resumed
[08/Jan/2015:09:07:43 -0800] NSMMReplicationPlugin - agmt="cn=add -> hub1"
(hub1:2390): Warning: unable to receive endReplication extended operation
response (Can't contact LDAP server)
[08/Jan/2015:09:07:43 -0800] NSMMReplicationPlugin - agmt="cn=add -> hub1"
(hub1:2390): Simple bind failed, LDAP sdk error 91 (Can't connect to the
LDAP server), Netscape Portable Runtime error -5961 (TCP connection reset
by peer.)
[08/Jan/2015:09:08:05 -0800] NSMMReplicationPlugin - agmt="cn=add -> hub1"
(hub1:2390): Simple bind resumed


Any idea what is causing this? I checked ADD/DEL operations during the
outage and none of them stand out. No MOD errors and all MODs completed
within a second.


~Shardul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/389-users/attachments/20150108/fb242463/attachment.html>