Dear Clive,
Clive Messer wrote:
On Friday 11 Feb 2005 14:53, George Chelidze wrote:
Hi George,
I don't use FC3, neither rpms. I compile it from source. Both sources (mentioned above) seems to contain the same clamav-milter.c code, so why the last should work and the first not. #define SESSION /*bla bla bla*/ is commented in both of them so I think it's the same as #undef SESSION, isn't it?
I don't know. What I do know is that I had the issue when I installed clamav-0.82 from the RPM which originally had 'SESSION' defined eg.
LibClamAV Warning: Session 0 has gone down LibClamAV Warning: Session 0 restarted OK LibClamAV Warning: Session 1 has gone down LibClamAV Warning: Session 1 restarted OK LibClamAV Warning: Session 2 has gone down LibClamAV Warning: Session 2 restarted OK LibClamAV Warning: Session 3 has gone down LibClamAV Warning: Session 3 restarted OK LibClamAV Warning: Session 4 has gone down LibClamAV Warning: Session 4 restarted OK LibClamAV Warning: Session 5 has gone down LibClamAV Warning: Session 5 restarted OK LibClamAV Warning: Session 6 has gone down LibClamAV Warning: Session 6 restarted OK LibClamAV Warning: Session 7 has gone down LibClamAV Warning: Session 7 restarted OK LibClamAV Warning: Session 8 has gone down LibClamAV Warning: Session 8 restarted OK LibClamAV Warning: Session 9 has gone down connect: Network is unreachable connect: Connection refused connect: Connection timed out Wed Feb 9 13:29:23 2005 -> ERROR: ScanStream: accept timeout. Wed Feb 9 13:29:23 2005 -> ERROR: ScanStream: accept timeout.
Since rebuilding clamav-0.82 with '#undef SESSION' I no longer have a clamd log full of errors and a maillog full of ... eg. 'sb75g2 sendmail[14264]: j18J8ro6014264: Milter: data, reject=451
4.3.2 Please try again later'.
Milter and clamd were restarted 27 hours ago after rebuild with '#undef SESSION'. No errors since. If '#undef SESSION' does not explain that then I really don't know. YMMV.
I understand your point. I have just replied to Nigel who adviced me to use devel-20050210 and stated that I think it wouldn't help as it's the same code of clamav-miter.c. Anyway, it's up and running and waiting for a new db update to hang it or pass through, who knows:)
Regards
Clive
Best Regards, -- George Chelidze _______________________________________________ http://lists.clamav.net/cgi-bin/mailman/listinfo/clamav-users
