Local Delivery stopped working

[Doc3rror]

Hello,

we have since this morning a strange problem.

We are running SBS with 50 users, all runs fine over 1 year with version 10.

Now one week before the planned upgrade of the server to version 11 we got following error.

The Local Delivery everytime stops directly after the start (1 second later). The server received all the mails and placed it in the quoue. The external mail quoue is working, but the delivery to the users is not working. The dellivery to public folders is still working.

Here are some entry's from some log files.

[root@mail1 ~]# /etc/init.d/scalix status
*** mail1
PC Monitor Started NON-STOP 0
Directory Relay Server Started 19:25:45
Notification Server Started 19:25:45 0
Shared memory daemon Started NON-STOP
Notification Monitor Started NON-STOP
Session Monitor Started NON-STOP
Container Access Monitor Started NON-STOP
Item Structure Server Stopped
Database Monitor Started 19:25:45
Licence Monitor Daemon Started NON-STOP
LDAP Daemon Started 19:25:45
Queue Manager Started NON-STOP
Item Delete Daemon Started NON-STOP
IMAP Server Daemon Started 19:25:45
SMTP Relay Started 19:25:45
Mime Browser Controller Started 19:25:45
Service Router Started 19:25:45 0
Local Delivery Partially Abor 20:03:01 630
Internet Mail Gateway Started 19:25:45 0
Local Client Interface Enabled 19:25:45 0
Remote Client Interface Enabled 19:25:45 7
Test Server Started 19:25:45 0
Request Server Started 19:25:45 0
Print Server Started 19:25:45 0
Bulletin Board Server Started 19:25:45 0
Background Search Service Started 19:25:45 0
CDA Server Started 19:25:46 0
POP3 interface Started 19:25:46 0
Omscan Server Started 19:25:46 0
Archiver Started 19:25:46 0


The output in /var/opt/scalix/logs/fatal

SERIOUS ERROR Local Delivery(Local Delivery) Wed Sep 26 20:03:02 2007
[OM 10270] Process about to terminate due to error.
Signal (Segmentation Violation) trapped by process 12375
Procedure trace follows:
-> nm_AppendFieldMem
-> nm_AddSeparators
<- nm_AddSeparators
-> nm_ParseORN
<- nm_ParseORN
<- nm_AppendFieldMem
<- nm_PutFieldMem
-> nm_PutFieldMem
-> nm_AppendFieldMem
<- nm_AppendFieldMem
<- nm_PutFieldMem
-> nm_ParseORN
<- nm_ParseORN
<- ul_utUnpackUserEnt
<- ul_FindPrimeUser
<- ul_IsEnu
Pid of logging process: 12375


SERIOUS ERROR Local Delivery(Local Delivery) Wed Sep 26 20:03:02 2007
[OM 10272] BACKTRACE:
/opt/scalix/lib/libom_er.so(er_add_backtrace+0xb6)[0xf7f3be76]
/opt/scalix/lib/libom_er.so[0xf7f3c162]
/opt/scalix/lib/libom_er.so(er_DumpProcAndExit+0x1f)[0xf7f3c31f]
[0xffffe500]
[0x614d223d]
ÿ
Pid of logging process: 12375


omstat -s
Service Router Started 19:25:45 0
Local Delivery Partially Abor 20:03:01 630
Internet Mail Gateway Started 19:25:45 0
Local Client Interface Enabled 19:25:45 0
Remote Client Interface Enabled 19:25:45 7
Test Server Started 19:25:45 0
Request Server Started 19:25:45 0
Print Server Started 19:25:45 0
Bulletin Board Server Started 19:25:45 0
Background Search Service Started 19:25:45 0
CDA Server Started 19:25:46 0
POP3 interface Started 19:25:46 0
Omscan Server Started 19:25:46 0
Archiver Started 19:25:46 0

I have restarted the server and all the services more than 20 times over the day.

The mail storage is running on a LVM RAID 10 array with Scalix version 10.0.1.3.

I have started Omcheck in script mode and fixed all error's with no succsess.

If you need more information please feel free to request them.

Best regards

Carsten

[gren]

Hi,

Is the first message shown in the output of

Code: Select all

omstat -q LOCAL

always the same message?
If so, it may be a message that causes serious problems for Local Delivery. In that case, a copy of the message for diagnostics would be useful. To create a copy,
run "omqdump". When it says
"If you do not wish to continue, hit return now."
enter the password which is 'A##E' where ## is the current day of the month + 10.
Use the "g" command to GET the message from the queue (LOCAL in this case)
Use the "o" command to output the open message to files.

If you tar up and gzip the files and send that to me, that would be great.

My email is gren.elliot and the domain is guessable

If it isn't the same message every time, the problem may be different. It might be instructive to run the "strace" utility against the local.delivery process.

Probably easiest to do this running local.delivery interactively as follows :

Code: Select all

omoff -d0 -w ld
omreset -o on ld
strace local.delivery -i

When you have finished gathering information (and presumably local.delivery has aborted) reset things to normal by repeating the omoff command and then use omon to switch ld on again.

Regards,
Gren.