I've been trussing the freshclam process (from clamav-20030317) for a few
weeks now and I think I know (at least one cause of) what's going wrong -
it's leaking file descriptors and eventually it gets an EMFILE error
whenever it tries to make update the database or talk to clamd. Is this
enough to
i ran it through gdb, and this is the backtrace:
#0 0x40025c77 in messageToBlob (m=0x86132f8) at message.c:545
#1 0x40024720 in insert (mainMessage=0x0, blobsIn=0x0, nBlobs=0,
textIn=0x0,
dir=0x860ed70 "/tmp/7e954d524eeddf1a", rfc821Table=0x860ed60,
subtypeTable=0x860ed20) at mbox.c:705
>it's me again!
>
>with the latest viruses.db all four virus emails are correctly detected,
and
>the segmentation violation disappeared
i'm getting the segfault again. I copied the virus database, just in case
regards,
René
-
Hi,
I've got the same problem...
the same message after hours of use..
I don't know where it comes from??
I try each new version til this disappears..
If there is a solution, i'm interested
Sendmail 8.12.9-3
Latest snapshot Clamav & Milter
Debian Linux
Le mer 02/07/2003 à 12:38, John Purcel
Hello all,
this may be partially off-topic, but looks like a problem most of us run into.
How does one prevent an exitus of his virus-scanner during runtime because of
archives unpacking to GBs of (empty) data flooding your hd? I mean somehow the
unpacking has to be solved by any user of clamav, o
I got Clam running over the weekend as a basic daemon. Freshclam seems to
be working. But I'm lost in trying to troubleshoot the problem below.
It appears that the problem is between Clamd and the milter. Clamdscan is
placed in a cron that seems to run fine (it finds the test virus files). In
