Availability report for clustered/redundant systems

Looking for ideas on how to create availability report for clustered/redundant systems.

Most of systems are redundant so if one is available and boss just want me to create a report which is tells her the email system is 99.5% available instead of saying:

Email ServerA is 100%

Email ServerB is 80%

Email ServerC is 99.5%

Email ServerD is 100%

My boss's arguement is that the users do not care whether Email ServerB is DOWN or UP as far as Email ServerA, C and D are UP and users can continue to send/receive emails the system is working.

However, she does not think using the average percentage of the four servers is a good approach.

I am keen to hear how other people have addressed this issue.

Thanks

Find more posts tagged with

clustered_systems_monitoring

availabiltiy_report

availability_report_redundant_systems

availability_report_clustered_systems

monitoring_redundant_systems

Accepted answers

All comments

jbiggley

I know that this a late reply, but I was searching for a similar report and thought I'd add this response. It sounds like your manager is confusing server availability with service availability. In order to accurately manage service availability you need Server and Application Manager so that you can view the critical services (like email) in the cluster instead of the status of the individual clusters.

Hope that helps (and that you already found your way to SAM on your own!)

bluefunelemental

Even later reply - I agree with jbiggley on SAM being good for monitoring services availability.

However for quick and dirty up/down time measurements I have been using group availability charting - add the clustered servers into a group and choose best status rollup option. Now when you use the group availability chart resource (which you can schedule as an emailed report PDF) it will show 100% as long as all servers are up - network takes them offline and you'll see 0%. Quick and dirty.