Appreciate your feedback with this request. Ive inherited a Kaseya environment and the original guy (now left) has enabled every alert / monitor set available.
I would like to ask for your knowledge and experience to provide examples of the core services which should be monitored and alerted. This could be the obvious perfrmon counters eg CPU, Memory, Disk Space etc through to core Windows Services, Exchange services, Citrix services, SQL services, Hardware etc.
My goal is to start small then increase / fine tune these alerts over time. The issue is, right now, I have so much noise from all these alerts, I dont know whats worth investigating and what should be dismissed.
Just looking for a mandatory baseline to start with and then grow this into a more detailed baseline. Appreciate your feedback.
Sadly, that's a common problem among VSA users... We spent 6 months developing our monitor sets based on the prior year's monitoring data and processed that against the help-desk response to each. Figured out what to monitor, what to ignore. It's not a simple task, but my first suggestion is to ditch every performance alert you have. Of all our clients, only 1 has allowed us to do performance optimization that actually warrants monitoring their server performance. We use KNM for this, and our automation discards any alert from performance outside of business hours or anytime for customers that haven't been optimized. Once that crap is removed, you can focus on the most important monitors, always adding back once the basics are under control. The presentation that I did at the Automation Exchange kick-off has some good info about monitors -
I reviewed an MSP's VSA 2 weeks ago and found monitors with over 70 PAGES of alert objects defined - that's almost 1500 objects monitored in just one alert! We have about 90 monitor sets now and layer them - one group for SQL, another for AD, etc. I wrote an audit tool that controls what monitor sets get applied based on what's discovered - keeps admin overhead to a minimum.
It's worth the effort.. we moved from a "monitor everything/automate nothing" environment that I inherited to a fresh VSA, optimized monitors, tons of automation, and we even developed "Smart Monitors" that self-adjust thresholds, suppress transient events, and self-remediate before generating an alert. After the first year, we found that we had an average of 62% fewer alerts, and the engineers spent almost 50% less time on alert tickets (because the basics were completed by automation). This allowed us to more than double our managed agents without increasing help-desk staff - we typically operate at 1 engineer per 1000 endpoints.
Our automation/monitoring package is available in Automation Exchange if you want a fresh start. :)
Thanks for your response. If you could advise the name of your package, would be appreciated. Thanks.
Aaron -Its the "MSP Builder - RMM Suite for Kaseya VSA", on the Featured Product section. or more info at www.mspbuilder.com.