[mrtg] Re: Whither the 95th percentile parser?

Henry Steinhauer H1STEINH at hewitt.com
Fri Dec 24 00:32:56 MET 1999


Pete (and others)

I built on your idea and added a few others if anyone is interested, I
could pass it on to them.

Just e-mail at h1steinh at hewitt.com and I'll pass it on after the holidays.

It::
1 - records the higer of the two (in/out) for each observation and then
reports the 95th based on that.
2 - Gives the number of observations that the sample is based on,
3 - Gives the 95th for the In, Out and Overall in the output.

I also plan to add weekly information so it might be possible to spot
trends but I have to think through a good way to archive the information.

Also  a way to strip out the "extra" information that gets into the daily
archive process since we are really keeping about two days worth each day.
I have more than 1,000 log entries now and this begins to add up the disk
space.

But -  Thanks much to Pete, I at least have some Perl Code to modify and
hack.

I have played with also creating Mean, 50th, 75th, 85th, 90th percentiles
since they are going to a file and space is not a real concern.  Looking at
the spread helps tell you information about the spikeness of the data.

Thanks again Pete -  I greatly appreciate it.

Henry Steinhauer -  h1steinh at hewitt.com


--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:mrtg-request at list.ee.ethz.ch?subject=help
Archive     http://www.ee.ethz.ch/~slist/mrtg



More information about the mrtg mailing list