Kolab 389 directory server not responding

[Ramiro Aukam]

Hello
Our 389 directory server is freezing after some running time (time
varies , can be multiple days or only few minutes) , even restarting the
computer doesn't work but if i run/service ***@kolab restart/ it
takes ~ 1 min to stop but then works for some time again , also if i
wait for ~ 30 min or longer it also starts working again.

I have no idea where to search, i searched in all Log files there is no
error from the directory server.

Is there a documentation about the 389 directory server in kolab because
it seems to me that it varies from the standart 389 directory server i
couldn't even find a way to turn debug mode on ?

I am running Kolab 16 in Ubunut 16.04 , installed with this guide
https://docs.kolab.org/installation-guide/ubuntu-16.04.html

Thanks
Ramiro

[Jan Kowalsky]

Hi Ramiro,

Post by Ramiro Aukam
Hello
Our 389 directory server is freezing after some running time (time
varies , can be multiple days or only few minutes) , even restarting the
takes ~ 1 min to stop but then works for some time again , also if i
wait for ~ 30 min or longer it also starts working again.

we struggle a lot with 389-ds. While a single server runs fine we get
performance problems spontaneously hangs and crashes in a multimaster
replication environment.

Post by Ramiro Aukam
I have no idea where to search, i searched in all Log files there is no
error from the directory server.

the error log goes to /var/lib/dirsrv/YOURINSTANCE/error.log

Post by Ramiro Aukam
Is there a documentation about the 389 directory server in kolab because
it seems to me that it varies from the standart 389 directory server i

maybe depends from your distro. on debian since jessie it's the
unchanged 389-ds from repository.

Post by Ramiro Aukam
couldn't even find a way to turn debug mode on ?

Most of configuraion is done with ldap attributes - some of them are
even configurable during runtime via ldapmodify.

https://access.redhat.com/documentation/en-us/red_hat_directory_server/10/html/configuration_command_and_file_reference/core_server_configuration_reference#cnconfig-nsslapd_errorlog_level_Error_Log_Level


It's all very well documented in either

https://access.redhat.com/documentation/en-us/red_hat_directory_server/10/

or

https://directory.fedoraproject.org/

Search the 389-users mailing list
https://lists.fedoraproject.org/archives/list/389-***@lists.fedoraproject.org/message/ZNZ2ADZKCXFHQKNUTHJ7IC57EGGLCFLJ/
for help.

Regards
Jan

[Timotheus Pokorra]

Hello Ramiro,

Post by Ramiro Aukam
Hello
Our 389 directory server is freezing after some running time (time
varies , can be multiple days or only few minutes) , even restarting
it takes ~ 1 min to stop but then works for some time again , also if
i wait for ~ 30 min or longer it also starts working again.

I have posted some months ago a description how we were able to
stabilize dirsrv for TBits.net:

https://kolab.org/hub/topic/321/directory-server-389-stuck/2

setting LimitNOFILE and sslapd-idletimeout and nsslapd-maxdescriptors as
desribed on that linked post.

I hope this helps,
Timotheus

----------------------------------------------------------------
Diese Nachricht wurde versandt mit Webmail von www.tbits.net.
This message was sent using webmail of www.tbits.net.

[Ramiro Aukam]

Hello

Thanks for the answers ,  the thread  whith your description i already
found after i wrote to the Mailing list.

I tried it directly , the LDAP server runs much better after that , but
we still have problems with connectivity , i found out that Guam is
crashing so i disabled it , i don't think that we need it because we
only have 15 mailboxes for now.

I disabled guam only today , and the problems mostly happens at morning
so i will have to monitor it further if it runs stable.

Ramiro

Post by Timotheus Pokorra
Hello Ramiro,

Post by Ramiro Aukam
Hello
Our 389 directory server is freezing after some running time (time
varies , can be multiple days or only few minutes) , even restarting
it takes ~ 1 min to stop but then works for some time again , also if
i wait for ~ 30 min or longer it also starts working again.

I have posted some months ago a description how we were able to
https://kolab.org/hub/topic/321/directory-server-389-stuck/2
setting LimitNOFILE and sslapd-idletimeout and nsslapd-maxdescriptors
as desribed on that linked post.
I hope this helps,
  Timotheus
----------------------------------------------------------------
Diese Nachricht wurde versandt mit Webmail von www.tbits.net.
This message was sent using webmail of www.tbits.net.

[Evilham]

Post by Ramiro Aukam
Hello
Thanks for the answers ,  the thread  whith your description i already
found after i wrote to the Mailing list.
I tried it directly , the LDAP server runs much better after that , but
we still have problems with connectivity , i found out that Guam is
crashing so i disabled it , i don't think that we need it because we
only have 15 mailboxes for now.
I disabled guam only today , and the problems mostly happens at morning
so i will have to monitor it further if it runs stable.
Ramiro

For what it's worth, my issues stopped the moment I disabled Guam.

You may want to run `kolab remove-user-subscription` for all your users
on the IMAP folders that Guam hides by default.

Some clients don't respect that, but roundcube will and IIRC thunderbird
will too.