Clamav Ping Timeout During Update

Rick Cooper rcooper at dwford.com
Tue Mar 4 21:13:19 GMT 2008


 

 > -----Original Message-----
 > From: mailscanner-bounces at lists.mailscanner.info 
 > [mailto:mailscanner-bounces at lists.mailscanner.info] On 
 > Behalf Of Neal Morgan
 > Sent: Tuesday, March 04, 2008 2:17 PM
 > To: MailScanner discussion
 > Subject: RE: Clamav Ping Timeout During Update
 > 
[...]
 > 
 > Thanks Julian and Rick for your responses.  I am running 
 > clamd, and when
 > I run the lint test MailScanner correctly finds it.  I am not
 > referencing clamavmodule.  (Lint confirms this as well).
 > 
 > I did have the "Monitors for ClamAV Updates" configured in
 > MailScanner.conf (upgrade_MailScanner_conf complained about 
 > them on my
 > last upgrade, so I updated the line to point to the proper 
 > locations).
 > Rick, per your suggestion, I have taken this setting out.
 > 
 > I am using Unix sockets.  To be sure MailScanner isn't 
 > trying TCP/IP, I
 > took the port number out leaving only the path to the 
 > socket.  One thing
 > that might be of interest, my clamd does create a socket and 
 > pid file,
 > but doesn't create a lock file (that I can find).  Could this have an
 > impact?
 > 
 > As far as other messages in the clamav.log file, the only 
 > other ones I
 > find are 1) self check, and 2) the occasional "file not 
 > found."  Those
 > seem to correspond with MailScanner reporting "not spam (too large)".
 > 
 > Anyway, I am still left with the behavior that I get the ping timeout
 > messages within 1 to 3 seconds of the time clamav.log 
 > reports "Database
 > correctly reloaded".  Weird.
 > 
 > I agree that 90 seconds is more than enough.  Changing this has no
 > effect on the problem.
 > 
 > If others running clamd aren't getting these
 > 	"MailScanner[4899]: ClamD Timed Out During PING Check!" 
 > 	and " MailScanner[3841]: Clamd::ERROR:: CLAM PING TIMED OUT! ::
 > ." messages, then it must be something I have 
 > mis-configured.  ...I just
 > have no idea what that is.  
 > 

Ok, I assume the lock file setting is blank (I think that is default) since
you would get an error stating clamd doesn't appear to be running instead of
the time out. The lock file isn't used in all installations.

You have never actually stated what version of clamav you are running. A few
versions back there was a real issues with the time the databases were
taking to load. What version are you running currently?

I have attached a script to do some testing with. The perl modules required
are, I believe, nothing that isn't required to run MS. Run PingClamd.pl
--help to get the options list. If you are running Unix Sockets you will
have to set -s /path/to/socket  and -d. The default timeout is 20 seconds.
If you can open two terminal windows restart freshclam in one window and run
PingClamd.pl in the other and see if you can cause the timeout there. If not
then we probably have to look for something in MS, otherwise something in
the OS. You can increase the timeout via the -t switch.

Rick

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: PingClamd.pl
Type: application/octet-stream
Size: 5674 bytes
Desc: not available
Url : http://lists.mailscanner.info/pipermail/mailscanner/attachments/20080304/2c3570a1/PingClamd.obj


More information about the MailScanner mailing list