Hi there,

I'm spending a lot of time trawling through logs at the moment. Our 
primary mail server is experiencing high load yet CPU usage is less than 
50%, memory less than 75% and disk usage minimal.

I've turned on debugging and set absolutely everything to maximum 
logging and have noticed some oddities that I could do with some help 
understanding. They could be normal, or could mean something.

After turning session logging to max I get loads of these, I'd say 
anywhere up to 50 each message:

2014-11-28 11:41:40 m1-xxxxx-xxxxx [Worker_7] 1.1.1.1 
<[email protected]> to: [email protected] info: Maillog - no log - 
log-condition is zero

As far as I can tell, it logs one of these for each line of data that is 
processed as follows:

 >2014-11-28 11:47:07 [Worker_1] <doing line ###DATA#
 >2014-11-28 11:47:07 [Worker_1] <Maillog
 >2014-11-28 11:47:07 [Worker_1] <matchSL - [email protected] - noCollecting
2014-11-28 11:47:07 m1-xxxxx-xxxxx [Worker_7] 1.1.1.1 
<[email protected]> to: [email protected] info: Maillog - no log - 
log-condition is zero

So this begs two questions, firstly when there is nothing to be logged, 
does it need to write a line saying so?

Secondly, is there a reason for matchSL running for every single line? 
My instinct is that the decision on whether to log or not based on 
sender address should be made once when the header is received and 
repeating it many times is wasting resources.

I also get a constant stream of these:

2014-11-28 11:40:54 [Worker_1] <error
 >2014-11-28 11:40:54 [Worker_1] <error
 >2014-11-28 11:40:54 [Worker_1] <error
 >2014-11-28 11:40:54 [Worker_1] <error
 >2014-11-28 11:40:54 [Worker_1] <error
2014-11-28 11:40:54 [Worker_1] <error

All workers except worker 1 log them once, immediately after a seterror 
and it seems to follow a greylisting rejection.

Worker 1 also logs occassional seterrors that match up with greylisting 
rejections. The oddity is that it also logs a continual stream of errors 
that don't seem to match up with anything.

I've tried to get some more information out of strace, but it leaves me 
equally puzzled.

I see a lot of EAGAIN (Resource temporarily unavailable), but no 
information about which resource is unavailable. I also understand that 
these are logged as normal behaviour when a process listens for 
connections on a TCP port and there is nothing there.

Monitoring file and network activity shows it all pretty low with every 
call being returned quickly.

Still, I see ASSP not responding to connections on the admin interface 
for minutes at a time. When I turned on debug mode, I turned it on at 
11:20:15.  It wasn't until 11:27:14 that ASSP wrote the following to the 
logs:

[Main_Thread] Info: starting partial debug mode to file 
/usr/local/assp/debug/1417174034.dbg

Any thoughts or suggestions on this one?

Thanks,
Colin

------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
_______________________________________________
Assp-test mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-test

Reply via email to