🚀 go-pugleaf

From article <9104181711.AA11246@hpnmd.hp.com>, by darrylo@HPNMXX.SR.HP.COM:
>> Well, the system management here wasn't comfortable making emacs sgid
>> system so it could read /dev/kmem, so I decided to sit down and write a
>> short elisp routine to parse uptime(1).  I don't have too much experience
>> programming in elisp, so if you see a way to speed it up, feel free to tell
>> me.
>
I've also written emacs lisp code to parse uptime, then I wrote it
in C, then I realized just how useless the uptime statistics are.

To the best of my knowledge, the uptime value = processes in run queue
 plus processes swapped out.  A very general overview.

I found that if you can gather paging stats, disk io, network traffic, 
etc., that you get much more meaningful graphs and you have data that
you can interpret to mean something.  More detailed stats like this
are usually available using routines provided by your OS vendor.

Of course, you can still parse them in the language of your choice.

Steve Anastasi
3M