Server maintenance and usability

NathanA

Where I work we need to monitor server uptime, but we allow a maintenance window where we can reboot our servers without affecting the uptime that is reported on our reviews. Is there a way to have Orion ignore reboots/outages/etc. during a set maintenance window?

Also, is there any way to tell the difference between a server that is up and processing requests (i.e. an exchange server that users are able to connect to and send/receive/view mail) and a server that is up, attached to the network, but not usable from a user standpoint (exchange server is up, but no one can connect)?

Thanks for any help/advice.

edwardle

I'm sure that this won't be the answer you're really looking for, but we use third party tools to accomplish these tasks.

The event suppression and correlation in Orion couldn't do all that we wanted so we send specific Orion alerts into a third party syslogger/trap receiver tool (Kiwi Syslogger - it's really cheap!) as well as all syslogs/traps from our network kit and then suppress/correlate the events there before sending to our alerting software which telephones you. The syslogger tool allows us to create filter rules containing times of the day and we use this to supress and re-activate alerts before and after scheduled downtime.

To test if a server is actually usable, you'll need third party tools. Our server guys (I'm a networks guy) use NetIQ and Compaq Insight Manager/Remote Insight cards to monitor the services on the server and server health, but there are also many other tools out there that should do what you want. Many tools can do specific TCP/UDP port checks as well as the usual SMTP, HTTP checks for specific URLs and Windows service checks. The more expensive tools can test for almost anything and have specific tests for apps such as Exchange, Notes, SQL Server, SAP, Oracle etc...

Hope this helps,
Lee