Proactive monitoring of servers

[PVR]

I am planning proactive monitoring of our HP servers.

Can you give your ideas to take it forward?

Also pl let me know what all are the free tools available for monitoring.

Many thanks in advance !!

[Con O'Kelly]

Hi

Depending on the number of servers you could write some shell scripts to monitor your systems. Main things to monitor are filesystem space, key processes, performance, system log files and security (eg failed logins, su's etc).

I have implemented this in the past where all detected events for all servers are written to a central event log file. This logfile is displayed on a webpage. In this way you can easily monitor all your systems.

Cheers
Con

[Michael Tully]

As suggested by Con, you could make use of scripts to do the job for you. As a result of scripts running, outputs could be into log files, emails or smnp traps.

You could go a lot further and purchase actual software that does this for you, with the likes of openview/network node manager and other products like patrol tivoli and CA unicentre. All of these have snmp traps that can change colour on a screen for operators, create email warnings and to some extent could be used to send SMS.

[Jeroen Peereboom]

L.S.

Some free tools:
OpenNMS (Network Management System)
Big Brother (www.bb4.org).

Of course you can also have a look at the forum for OpenView Operations, also known as OVO or ITO.

JP.

[Steven E. Protter]

We do almost all of it with scripts and cron.

If there is a problem, sendmail is used to notify appropriate staff.

In addition to that there is event monitoring service. Thats built into sam under resource management. You can set all kinds of triggers for certain events to notify staff. You can go beyond email as well.

Sam also has peformance monitors. Good place for beginners to start.

SEP

[PVR]

SEP

Can you help me with such system monitoring scripts

Thanks !

[MarkSyder]

Don't forget to check files in /var/adm which can be known to take up large amounts of space.

cd /var/adm
ll btmp
ll sulog
ll wtmp
ll cron (checking for OLDlog and log)
ll lp
ll syslog (checking for mail.log, OLDsyslog.log and syslog.log).

DO NOT delete any of these files: if they are large, null them: e.g.

cd /var/adm/syslog
> syslog.log

Do a daily check for core files.

Mark Syder (like the drink but spelt different)

[Steve Kruger]

I would recommend ISEE (Instant Support Enterprise Edition).

URL: http://www.hp.com/hps/hardware/hw_standard.html

The HP Instant Support Enterprise Edition (ISEE) Standard Configuration transmits incidents and scheduled system information collections from monitored ISEE clients in your datacenter. This data is automatically sent to your regional HP Support Center for reactive support in response to incidents and for proactive system monitoring.

[Ravi_8]

Hi,

scripts as told by con, and schedule using cron.
But i have Tivoli working in my environment which send alerts in case of malfunctioning.

Even HP's openview does the same.