__/ [ Jim Richardson ] on Saturday 03 June 2006 22:20 \__
> < snip />
>
>>> logresolve in the apache toolchain, in debian, it ships in
>>> apache2-utils. You can do it at the time you log the hits, via apache,
>>> but most busy sites shut that off, as it slows down the logging. man
>>> logresolve for details, but this example should help
>>>
>>> logresolve -c resultsfilename < logfile > newlogfilewithrevDNSentries
>>
>> Ta muchly, Jim. Much appreciated! Found it, got it, compiled it and
>> running it at the moment. I wonder if I ought to just convert all my
>> monthly logs in this way. This could come handy when grepping, but I
>> rarely ever do this...
>>
>><snip />
>
> Better yet, feed the logs to something like webalizer or awstats, google
> for apache log parsers, you have a multitude of options there.
Well, here's what happened:
I woke up this morning and Firefox at home was not responding. Fine. I shut
it down. When arriving at the office I discover that the drive has 0 bytes
available and the swap partition is filled to the rim. Sadly, while
processing and replicating a 250 MB log file, I forget to make some vacant
space. No big deal. Linux didn't even cough, let alone froze.
Anyway, see my reply to Borek. I grepped half of the output generated
overnight (space drained at about 2AM, killing some computer vision
experiments along the way). I found several more Microsoft employees who
visited the site.
As for the suggestion made above, awstats.pl and Webalizer (maybe Analyze
also) make use of DNS resolvation whenever they are run (nightly cron job).
However, their input/output is not 'greppable'.
Many thanks,
Roy
--
Roy S. Schestowitz
http://Schestowitz.com | Free as in Free Beer ¦ PGP-Key: 0x74572E8E
9:15am up 37 days 14:48, 12 users, load average: 0.50, 1.21, 1.30
http://iuron.com - semantic engine to gather information
|
|
