Hi !
Thank you for your suggestions.
I'm afraid that none of these suggestions are appropriate to the current
situation.
This is no cluster installation, it is a single instance of a VM running
on the same physical machine as SOGo.
There are no other services running on this machine.
The machine is idling, there is no overload on the machine nor the network.
The disconnects/failures seem to happen overnight when there is little
or no activity.
There are no error messages in other logs indicating any errors.
The mail server (which is on a different physical machine on the same
network) uses the same Windows VM for authentication via LDAP as SOGo,
and the mail server gives no errors and works fine.
I can ping the Windows VM (LDAP server) from the shell where the SOGo
service is running, so the network connection is fine.
The mail server uses the same account for binding as SOGo, so it's not
an account problem either.
Since a simple restart of only the SOGo service fixes the problem, I
think the problem lies with SOGo.
Is there perhaps a permanent LDAP connection that times out ?
Regards,
Sigurd Holter
Den 27.10.2014 10:58, skrev Christian Mack:
Hello Sigurd Holter
Am 2014-10-26 um 09:02 schrieb [email protected]:
The past few days we are seeing this in our logs :
NAME:LDAPException REASON:operation bind failed: Can't contact LDAP server
(0xFFFFFFFF)
Our users cannot log on to the web interface.
Restarting the SOGo daemon makes everything work again.
The LDAP server in this case is a Windows 2012 R2 server.
Since our mail servers also authenticate against this server - and mail flow
is normal - I doubt there is a problem with the Windows server.
The same setup has been working fine up until 4-5 days ago.
SOGo and WIndows 2012 R2 server are on the same machine - the WIndows server
is virtualized.
This is SOGo 2.2.9a
Any tips ?
Can't contact LDAP server (0xFFFFFFFF)
Means literally, that your sogo server can not reach your LDAP server.
This can happen eg. while one or both of these servers are moving
between cluster nodes in your VM environment.
This can happen, when your Network is overloaded.
This can happen, when you run out of port numbers.
This can happen, when your VM has not enough compute power or RAM.
So check your environment and log your performance.
Kind regards,
Christian Mack
--
MVH
Sigurd Holter
Bookingleder/prosjektleder
West Audio A/S
Consul Sigval Bergesens vei 41
4016 Stavanger
Telefon 51537030
--
[email protected]
https://inverse.ca/sogo/lists
