Re: Prepare to be assimilated^Wanalysed; resistance is futile

[Ken McDonell <kenj@xxxxxxxxxxxxxxxx>]

On 18/07/13 09:47, Nathan Scott wrote:

> [unless I hear otherwise, at this stage I'm planning to set an increased
> timeout using Jeffs suggestion, an extra -T 24h10m and the -r option...]
-r also reports what groups are logged, so it is easier to find "actual 
config file does not match assumption" issues ... and it is a once trip 
per group of metrics in the control file, so very low overhead.
> The compressing/culling of their old /var/log/pcp/pmlogger/*/* files?
> IIRC, that is done using find and not the contents of the control files,
> so I *think* that one's a non-issue - files from rogue loggers should be
> dealt with there (if not, its at least fixable by using find).
It _is_ control file driven at the moment.  We do this

        find . -type f -mtime +$CULLAFTER

and this

        find . -type f -mtime +$COMPRESSAFTER

but both within the context of a ...

cat $CONTROL \
| sed -e "s/LOCALHOSTNAME/$LOCALHOST/g" \
      -e "s;PCP_LOG_DIR;$PCP_LOG_DIR;g" \
| while read host primary socks dir args
do
        ...
        cd $dir
        ...
done

But we could add a catch all cleanup in another loop at the end, the 
same as is done for the "trace" files ... see pmlogger_daily, around 
line 835.