ClamAV constantly failing

[Jason Voorhees]

Hello guys:

I'm running CentOS 7 x86_64 with MailScanner 4.85.2-3 and clamav
0.99.1 from EPEL. These are my clamav packages installed:

clamav-data-0.99.1-1.el7.noarch
clamav-server-systemd-0.99.1-1.el7.noarch
clamav-unofficial-sigs-3.7.2-1.el7.noarch
clamav-server-0.99.1-1.el7.x86_64
clamav-0.99.1-1.el7.x86_64
clamav-lib-0.99.1-1.el7.x86_64
clamav-scanner-0.99.1-1.el7.noarch
clamav-scanner-systemd-0.99.1-1.el7.noarch
clamav-filesystem-0.99.1-1.el7.noarch
clamav-update-0.99.1-1.el7.x86_64

What I'm experiencing is that several times a day clamd is failing.
There are several errors like:

1. MailScanner cannot find clamd's socket even if it exists in
/var/run/clamd.scan/clamd.sock -> I have to stop clamd, remove the
socket file and start clamd again.

2. MailScanner says clamd isn't running like this " Clamd::ERROR::
COULD NOT CONNECT TO CLAMD, RECOMMEND RESTARTING DAEMON" -> I have to
restart clamd or do some other troubleshooting.


3. Clamd detects a failed or corrupted database file (i.e. main.cvd)
-> I have remove such db file, restart clamd and/or run freshclam.

I think there might be a bug in ClamAV because the same thing occurs
seldom in other similar installation from different customers.

Ok, I know this is not a ClamAV mailing list, but I wonder if there's
a way to tell MailScanner to not depend on ClamD if this isn't
running. Because, what's currently happening is that when one of these
clamd errors occur, MailScanner can't process messages correctly and
my mail queue starts to grow until I manually solve this.


I wonder if it's possible to tell MailScanner that must skip Antivirus
checking if AV isn't running but allows messages to be delivered
anyway instead of be holded in the queue.


Do you get my point? I hope someone can help me with some ideas.


Thanks in advance.