efa-project.org

It appears I had a problem with my internal mailserver (OS crash) overnight which has occured before but I seem to also have had a problem in EFA.

When i've seen an internal mail server problem i just see messages sitting in the outbound mail queue... fix the problem and alls good.

However this time it seems that there was a problem on the inbound side, with messages being bounced back to the sender and then constantly redelivered. Consequently it looks like 350K messages processed by EFA !!
There's now a lot of data in the mySQL database and everything is a bit sluggish ....
Any idea if this is a EFA bug that needs fixing and is there any easy way to clean the DB up .....

I'm new to EFA and loving it so far .... if there's any diagnostics i can provide you with please let me know.

Andy

Likely clamd has crashed on your system. Reboot to stabilize your system temporarily.

How much memory do you have?

How many mailscanner threads are you running?

Hi,

i rebooted the system when i first noticed the issue. The VM was set up to use 2Gig mem and 4 scanner threads, i've now changed that to 4 Gig and 2 threads.

if there has been a clamd problem, is there any logs i can provide that would be useful ?

Andy

Known issue, but you would see the problem in the following log:

/var/log/maillog

You would see MailScanner attempting to submit to clamd for scanning, failing, and requeing the same batch over and over.

We found that 2GB is not sufficient, so all new builds start with 4GB of memory. I encourage everyone to bump up their memory on older installations.

well upgrading from 2GB to 4GB is not a huge problem, I am wondering why clamd is not monitored at all ?
it seems to be a vital part of the system, and when it dies for whatever reason it will not be fired up again ?

it could be me, but I think vital parts should be monitored and restarted if necessary, eventually notify the administrator by email if it's an very unusual process that had died