Adding a server instructions needed

[rex007can]

Done.

But now, what does this mean from the logs. Syncing hasn't gone through...
And I get the same entry in the logs of both servers, with the other server's name.


ERROR Directory Sync(Directory Sync) Tue Aug 8 11:15:00 2006
[OM.DS 1506] (DS IMPORT)
A REPLY Updates has been timed out and the
configured transaction retry count has been exceeded.

Pid of logging process: 1817
Current errno value: 42


ERROR Directory Sync(Directory Sync) Tue Aug 8 11:15:00 2006
[OM.DS 2600] Client details for fault (1506)

Import Directory : **
Export Directory : **
Export OR Address: +DIRSYNC////NODE-A/DOMAIN//////////
Thread Number : 0
Propagated By Id : 4864
Timestamp : 000000000001
Request Id : 0011549808820018
Local Path : loc1
Remote Path : rem2
Pid of logging process: 1817
Current errno value: 42

[AussieOwner]

Please execute the following commands on both machines and post the output.. Thx

Code: Select all

omshowrt -q all

omlistds -e
omshowds -e <number from line above>

omlistds -i
omshowds -i <number from line above>

[jim mullady]

Try these steps and see how it works

Scalix Route Configuration

When using more than one scalix server in an environment we need to be able to add routes for directory synchronization and mail routing.

In this example we will use 3 servers each with one mail node.

Scalix1 will have a mail node of mt1
Scalix2 will have a mail node of mt2
Scalix3 will have a mail node of mt3

So with those configurations we setup our scalix servers as follows;

Scalix1 Server
Step 1. run omresetmn
Step 2. Run command. omaddrt -m mt2 -q SMINTFC -i scalix@scalix2.domain.com
Step 3. Run command omaddrt -m mt3 -q SMINTFC -i scalix@scalix3.domain.com
Step 4. Restart the service router. omoff -d0 "Service Router"; omon "Service Router"
Step 5. omaddds -i -m +DIRSYNC/mt2 -t "000101 00:00"
Step 6. omaddds -i -m +DIRSYNC/mt3 -t "000101 00:00"
Step 7. omaddds -e -m +DIRSYNC/mt2
Step 8. omaddds -e -m +DIRSYNC/mt3
Step 9. In /var/optscalix/sys/general.cfg enter in
DS_CUST_SEND_REQ_NOW=TRUE
DS_CUST_MSGQ_TIMEOUT=2
Step 10. omoff -d0 dirsync ; omon dirsync

Scalix2 Server
Step 1. run omresetmn
Step 2. Run command. omaddrt -m mt1 -q SMINTFC -i scalix@scalix1.domain.com
Step 3. Run command omaddrt -m mt3 -q SMINTFC -i scalix@scalix3.domain.com
Step 4. Restart the service router. omoff -d0 "Service Router"; omon "Service Router"
Step 5. omaddds -i -m +DIRSYNC/mt1 -t "000101 00:00"
Step 6. omaddds -i -m +DIRSYNC/mt3 -t "000101 00:00"
Step 7. omaddds -e -m +DIRSYNC/mt1
Step 8. omaddds -e -m +DIRSYNC/mt3
Step 9. In /var/optscalix/sys/general.cfg enter in
DS_CUST_SEND_REQ_NOW=TRUE
DS_CUST_MSGQ_TIMEOUT=2
Step 10. omoff -d0 dirsync ; omon dirsync

Scalix3 Server
Step 1. run omresetmn
Step 2. Run command. omaddrt -m mt2 -q SMINTFC -i scalix@scalix2.domain.com
Step 3. Run command omaddrt -m mt1 -q SMINTFC -i scalix@scalix1.domain.com
Step 4. Restart the service router. omoff -d0 "Service Router"; omon "Service Router"
Step 5. omaddds -i -m +DIRSYNC/mt2 -t "000101 00:00"
Step 6. omaddds -i -m +DIRSYNC/mt1 -t "000101 00:00"
Step 7. omaddds -e -m +DIRSYNC/mt2
Step 8. omaddds -e -m +DIRSYNC/mt1
Step 9. In /var/optscalix/sys/general.cfg enter in
DS_CUST_SEND_REQ_NOW=TRUE
DS_CUST_MSGQ_TIMEOUT=2
Step 10. omoff -d0 dirsync ; omon dirsync

[rex007can]

Done.
but:
On server1-->
[root@server1]# omlistds -i
1 ** ** +DIRSYNC /node2
[root@server1]# omlistds -e
1 ** ** +DIRSYNC /node2
[root@server1]# omshowrt -q all
UNIX internet MIME
UNIX internet,tnef TNEF
SMINTFC node2 scalix@server2.domain.com
LOCAL node1,domain
[root@server1]# omshowds -i 1
SYNCHRONIZING DIRECTORY : **
EXPORT OR ADDRESS : +DIRSYNC /node2
DS-STRT-AT : 000101 00:00
DS-PERIOD : 15
DS-IR-FLG : N
[root@server1]# omshowds -e 1
SYNCHRONIZING DIRECTORY : **
IMPORT OR ADDRESS : +DIRSYNC /node2
DS-EL-FLG : N
<--

And on server2-->
[root@server2 ]# omshowrt -q all
UNIX internet MIME
UNIX internet,tnef TNEF
LOCAL server2,domain
SMINTFC node1 scalix@server1.domain.com
[root@server2 ]# omlistds -i
1 ** ** +DIRSYNC /node1
[root@server2 ]# omlistds -e
1 ** ** +DIRSYNC /node1
[root@server2 ]# omshowds -i 1
SYNCHRONIZING DIRECTORY : **
EXPORT OR ADDRESS : +DIRSYNC /node1
DS-STRT-AT : 000101 00:00
DS-PERIOD : 15
DS-IR-FLG : N
[root@server2 ]# omshowds -e 1
SYNCHRONIZING DIRECTORY : **
IMPORT OR ADDRESS : +DIRSYNC /node1
DS-EL-FLG : N
[root@server2 ]#

But this doesn't work.
I'm still getting the errors in /logs/fatal on both servers (Tis is from server2)

ERROR Directory Sync(Directory Sync) Wed Aug 9 13:45:00 2006
[OM.DS 1506] (DS IMPORT)
A REPLY Updates has been timed out and the
configured transaction retry count has been exceeded.


Pid of logging process: 5051
Current errno value: 42


ERROR Directory Sync(Directory Sync) Wed Aug 9 13:45:00 2006
[OM.DS 2600] Client details for fault (1506)

Import Directory : **
Export Directory : **
Export OR Address: +DIRSYNC////NODE1///////////
Thread Number : 0
Propagated By Id : 480
Timestamp : 000000000001
Request Id : 0011551373890001
Local Path : loc1
Remote Path : rem5
Pid of logging process: 5051
Current errno value: 42

[rex007can]

ahhh...nothing?
Anyone?
Anyone?
Bueler?

[AussieOwner]

Ok.. lets check something simple.

from server1 can you telnet <server2> 25 ??

try the same on server2

telnet <server1> 25 ??

[dkelly]

I've added a write-up on the wiki to give some troubleshooting tips for dirsync

http://www.scalix.com/wiki/index.php?title=HowTos/DiagnosingDirsync

Cheers

Dave

[rex007can]

Ok.. lets check something simple.

from server1 can you telnet <server2> 25 ??

try the same on server2

telnet <server1> 25 ??

Yes.

[rex007can]

Found the problem, now how do I fix it?

The problem is that on both test machines, the messages get delivered to the sendmail local inbox instead of the Scalix inbox. I don't know how that happened. One machine is a test machine updated from 9.4 to 10.0.1 and it didn't display that behaviour when it was still and 9.2.4.

the other machine is an install from scratch.
Both are running on RHEL.
Why did Sendmail stop handing the mail over to Scalix?

thanks.

[dkelly]

Does this mean that you have sendmail listening on an address other than localhost ?

Cheers

Dave

[rex007can]

I don't know. I didn't change it so it doing whatever the default behaviour is.

But if I telnet to either servers on port 25, then manually send an email to sxadmin, or an existing scalix AND Unix user, the message gets delivered to /var/mail/username.

And if I do this but send a message to a Scalix only user, the message is blackholed.

Which would explain why none of the sync messages are going through.

[rex007can]

FOUND it!

It was the AV configuration that was faulty and cause emails to get blackholed on one of the servers...
that was long...