Re: update of clamav fails - bad variable errors
On Tue, Apr 2, 2013 at 4:28 PM, francis picabia <fpicabia@gmail.com> wrote:
>
> I am going to leave the reinstall to another hour of the day when users
> won't be impacted much. In the meantime, I thought I'd test the
> freshclam run from the command line. First I need to stop
> the current daemon, and it triggers the same errors I'm seeing
> from attempts to update:
>
> # /etc/init.d/clamav-freshclam stop
> export: 266: DatabaseOwner
> UpdateLogFile
> LogVerbose
> LogSyslog
> LogFacility
> LogFileMaxSize
> LogTime
> Foreground
> Debug
> MaxAttempts
> DatabaseDirectory
> DNSDatabaseInfo
> AllowSupplementaryGroups
> PidFile
> ConnectTimeout
> ReceiveTimeout
> TestDatabases
> ScriptedUpdates
> CompressLocalDatabase
> Bytecode
> Checks
> DatabaseMirror
> DatabaseMirror: bad variable name
>
> At this time, some of the packages have been updated, but not others:
>
> # dpkg -l | grep clamav
> iU clamav 0.97.7+dfsg-1~squeeze1
> anti-virus utility for Unix - command-line interface
> iF clamav-base 0.97.7+dfsg-1~squeeze1
> anti-virus utility for Unix - base package
> iFR clamav-daemon 0.97.6+dfsg-1~squeeze1
> anti-virus utility for Unix - scanner daemon
> iFR clamav-freshclam 0.97.6+dfsg-1~squeeze1
> anti-virus utility for Unix - virus database update utility
> ii libclamav6 0.97.7+dfsg-1~squeeze1
> anti-virus utility for Unix - library
>
> It is rare for me to be the only person experiencing an error like this
> from a standard update procedure. The configuration file had not
> been changed for quite some time and should be OK, aside from any
> changes required for the new version - but it is complaining of too many
> of those for it to be real.
>
> I can go the kill route to resolve the init error easily. I'm looking
> for something
> more in depth of anyone else experiencing this update failure. Why
> does freshclam
> not stop?
>
> Same for asking for status:
>
> # /etc/init.d/clamav-freshclam status
> export: 266: DatabaseOwner
> UpdateLogFile
> LogVerbose
> LogSyslog
> LogFacility
> LogFileMaxSize
> LogTime
> Foreground
> Debug
> MaxAttempts
> DatabaseDirectory
> DNSDatabaseInfo
> AllowSupplementaryGroups
> PidFile
> ConnectTimeout
> ReceiveTimeout
> TestDatabases
> ScriptedUpdates
> CompressLocalDatabase
> Bytecode
> Checks
> DatabaseMirror
> DatabaseMirror: bad variable name
This is a strange problem. It is now 6 hours since I
saw the above errors. I put outbound smtp into
bypass (comment out content_filter in main.cf of postfix)
in preparation to shutdown amavis and clamav.
Then again I attempted to shutdown clamav-freshclam
with the init script.
This time it worked with no complaints.
I was able to run 'aptitude safe-upgrade' and
there were no issues at all. The full set of
clam packages are now at 0.97.7
Clamav packages were not purged and reinstalled to resolve this.
I hate problems which resolve like this because it is random and I
don't learn what was wrong.
All I can figure is freshclam did update the daily databases a few
minutes after I had the problem doing anything (stop or status) from
the init script with freshclam. Perhaps this somehow freed it up.
Reply to: