Messaging Gateway

Hi, we have a SmsSmtp 5.01 patched to 189, on a W2K3 server - currently serving some 70 local domains. Suddenly when we add another local domain that domain isn't serviced. Nothing appears in the logs, and the sender only gets a "553 We do not Relay".

I've checked the MySQL database, the local domain exists there in table settings_local_domains. The Local domain also exists in a file called locals.txt.

So something must be broken, but I haven't got a clue where.

Anyone have a tip?

Cheers / Peter

No, neither does restarting the server.

We had this problem in May, and then I solved it by deleting first one domain, that didn't fix it, then I deleted another domain - and re-applied the domain in question and voila it worked.

Today, I haven't got any domains to delete - but I can try to delete the domain in question and re-apply it again. However the domain is mentioned within mysql table settings_local_domains, and in the following two files:

C:\Program\Symantec\SMSSMTP\mta\delivery\etc\routing.txt

C:\Program\Symantec\SMSSMTP\mta\shared\locals.txt

So I suspect that either the MySql relationships are broken, or that another config file / queue pair doesn't exist - or some hash coded list perhaps - and that the cause could be some timeout occuring when we enter the domain through the control panel. It is though very troublesome for us. The server is very busy, often working at a 100% - conduit.exe and filter-hub.exe often fight for power.

/Peter

Hello Peter,

I would probably say if you are consistently running at 100% on the server you may want to look into more hardware. This could be part of what caused the issue in the first place.

The only other thing I could think of as a quick test would be to go into the user interface for SMS, go to the Settings tab, select Hosts on the left side, edit your host, go to the SMTP tab, and toggle the scanner role. If you change it to another role, save it, change it back and save again. This will force a rebuild of the metabase and can sometimes resolve these type of issues.

Other than that I would suggest calling into support.

I hope this helps!

Tom

Tom, I tried your suggestion and it didn't work. However, I did fiddle about quite long and found this document: http://service1.symantec.com/support/ent-gate.nsf/854fa02b4f5013678825731a007d06af/d769106d59463e768825733f006cafd4?OpenDocument

I did then go to our inside server and removed the Windows Firewall, then removed the domain, and added it again - and now relaying worked. I added another domain and relaying worked for that one too. What I'd like to have done is wireshark the actual traffic going on at the time of adding a domain to verify if there are any tests going on at the time of adding a local domain with host forward. It's not the way I would have designed a system.

Regarding the 100%, it appears that whenever there are messages flowing the processes grabs all they can to do their work. Smssmtp is very resource demanding - all the queue directories, memory consumtion etc.

The overview logs showed a lot of firewall errors and reading up on your support docs they state to increase memory - so I did up the memory. We run on a virtual server, so adding memory is quite easy.

I dropped the badmail folders content manually, as well as tidying the queue directories - they have a few xml files that can't be parsed for some reason.

So I can just hope the problem doesn't come back again.