Maciej Kedzierski wrote:
> I have another problem (probably continuation my earlier problem) > > This morning clamav-milter has died completely and I didn't start it > again. > # /etc/init.d/clamav-milter start (restart) didn't work until I restarted > clamav. > After this clamav-milter began work. > > In maillog I found this, and much more: > Oct 22 06:13:31 melmak clamav-milter[23181]: clamfi_connect: connection from > [69.6.27.97] [69.6.27.97] > Oct 22 06:13:32 melmak clamav-milter[23181]: hit max-children limit (504 >= 40): > waiting for some to exit > Oct 22 06:13:47 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > Oct 22 06:13:51 melmak clamav-milter[23106]: pthread_cond_timedwait: (rc = 110) U^P^H > Oct 22 06:13:55 melmak clamav-milter[23198]: clamfi_connect: connection from > mta04ps.bigpond.com [144.135.25.158] > Oct 22 06:13:55 melmak clamav-milter[23198]: hit max-children limit (505 >= 40): > waiting for some to exit > Oct 22 06:14:07 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > Oct 22 06:14:08 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > Oct 22 06:14:10 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > Oct 22 06:14:18 melmak clamav-milter[23176]: pthread_cond_timedwait: (rc = 110) > \230N^P^H > Oct 22 06:14:19 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > Oct 22 06:14:25 melmak clamav-milter[23161]: pthread_cond_timedwait: (rc = 110) > PT^P^H > Oct 22 06:14:32 melmak clamav-milter[23181]: pthread_cond_timedwait: (rc = 110) > X[^P^H > Oct 22 06:14:55 melmak clamav-milter[23198]: pthread_cond_timedwait: (rc = 110) > hS^P^H > Oct 22 06:15:01 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > Oct 22 06:15:07 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > Oct 22 06:15:15 melmak clamav-milter[30065]: ClamAv: accept() returned invalid > socket (Too many open files), try again > > There was above 500!!!! clamav-milter processes (they ate 1GB my server's > memory !!!!!) That's intersting, as that's exactly the same problem that I see on my server. As I mentioned in my earlier mail, I'm still running the CVS snapshot from 20030926, so it's possible that some of this has been fixed. It always seems to happen to me after clamd runs its database selfcheck, and I believe that clamd is the actual root cause, and the proliferating milter processes are just the end result. One thing that occurs to me - the server I'm running it on is a dual CPU machine. Could that create more opportunity for a race condition to generate the problem? Also possible worth mentioning - generally the symtom I see first is slow SMTP sending, as the milter times out, but occasionally, a clamd process has run away with the CPU, chewing 99% in top. On a dual CPU machine this is not nearly as bad as on a single CPU system, and I just kill the offending process, but maybe it gives a clue as to whats happening. I'm happy to help with any debugging I can reasonably manage. Mike. Mike. ------------------------------------------------------- This SF.net email is sponsored by OSDN developer relations Here's your chance to show off your extensive product knowledge We want to know what you know. Tell us and you have a chance to win $100 http://www.zoomerang.com/survey.zgi?HRPT1X3RYQNC5V4MLNSV3E54 _______________________________________________ Clamav-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/clamav-users
