SUMMARY: monitoring tool

From: Rene Degen (degen@srzts306.alcatel.ch)
Date: Mon Jun 14 1993 - 07:41:35 CDT


A couple of weeks I posted:
>
> I am looking for a program (set of programs) which
> is able to check the proper working of some workstations.
> I want check them for being up, not running out of disk space,
> existing of some special processes (e.g. named)
> and maybe some filtering of syslog.
>
> Detecting a problem should trigger an configurable reaction,
> like sending a mail message, forking off some process and so on.
>
> We have Suns and DECstations here, clients should be able
> to run on both, server doesn't matter.
>

Many thanks for your replies to:

- alan@nabeth.cxo.dec.com (Alan Rollow - Alan's Home for Wayward Tumbleweeds.)
- tim@andy.hssc.scarolina.edu (Tim White)
- marym@hol.Eng.Sun.COM (Mary Morris)
- navarro@azur.enet.dec.com (EIC Valbonne 10-May-1993 1405)
- meekj@pt.Cyanamid.COM (Jon Meek)

Short summaries:

- Tim White recommended "Watcher". It is very flexible and
  allows to react on any command output you want it to.
  I looked at it and was a little surprised that it has not
  been updated for a few years now.

- Jon Meek as well was concerned about "Watcher" not being
  updated since 1987. He started to do these things in perl
  as he used it for other jobs already.

- Mary Morris recommended SunNet Manager. Using SNMP
  you can check everything a (SNMP) client is available.

- Alan Rollow gave me the hint for a package in the
  DEC ASSET library running. It runs on Ultrix only,
  but he considered to convince my local DEC office for
  a port. (Read on Alan: you may be surprised!)

  He did not mention the name, but I think it is called
  DCMX, an Ultrix port of the VMS DCM utility we already
  use here.

- Jean Navarro told me that he works on an replacement for DCMX
  called PSWX (Polycenter System watchdog for UNIX).
  Short description from him:

> Client-server application composed of two part a consolidator installed on
> a central node which polls remote Agents installed on remote node to monitor.
 
> The consolidator automatically polls remote node through DECnet or TCP/IP
> and triggers actions routines which can be a
 
> - Character Cell Display
> - Graphical Display
> - DECmcc
> - Action routines like mail,beeper, logfile ...everything you
> want to be interfaced with
>
> The agents on remote node can detect the following events :
>
> - Hardware error (Cpu/disk/mem) by
> - Disk near full (threshold attached to a disk)
> - File size limit to monitor file size activity
> - process missing
> - process looping
> - printer stalled
> - Ethernet errors and collision
> - Interactive login limit
> - Network reachability pbs
> - mailq /swap/
> - customizable events using shell script to add your own events detectors

> The consolidator runs on ULTRIX and VMS , Agents runs on VMS, ULTRIX, SunOS,
> HP-UX and will run very soon on IBM.

  My sales man denied it to be available right now but thought
  it should be available mid-year. He had no idea about pricing
  at that moment.

So well:

Right now I will continue to use "DCM" for checking our VMS stuff
(which I did not mention in my request). For the UNIX stuff and
our communication equipment I will use SunNet Manager of an
adjacent departement.

As available I will further investigate the DEC stuff. That is
because I do not want to go through the process of deciding
what to check from scratch, as with "Watcher" or some own
perl scripts.

                                        -- red

-- 
 R. Degen, Alcatel STR AG, AST 1.691, Friesenbergstr. 75, CH-8055 Zuerich
  RFC822:  rene.degen@alcatel.ch		  Phone:  +41 1 465-3269
  X.400:   c=CH a=arCom p=Alcatel s=Degen g=Rene  FAX:	  +41 1 465-2411
  DECnet:  PSI%[0228]479512391::TRZCL1::R_DEGEN



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:55 CDT