I have to report Network Availability Metrics on a monthly basis. My current reports do not consider any load balancing in the calculations. I would like to be able to do that, but don't really know where to start. For example, let's say that I have Server 1 and Server 2 that are load balanced for Service 1. Now if Server 1 becomes unavailable (reboot, crash, or etc...) which is unplanned, my Service 1 is still available and there is no interruption of service. I want to be able to capture that in my metrics. Currently, I gather all of my availability data as a percentage and I sum the percentages and divide by the total number of nodes monitored. So in this case if Server 1 becomes unavailable, Service 1 is unavailable for that node, but that is not a true reflection of the availability of Service 1. My calculation would be
Service 1 Server 1 95%
Service 1 Server 2 100%
Availability = 97.5%
when in actuality my Availability for Service 1 is 100%. How do I account for load balancing, like in the example, in Report Writer? Can it even be done? How does everyone else handle this type of situation?
Message was edited by: Lamar Ross
What about monitoring the VIP address as it's own ICMP-only node? You can still get the detailed stats you want on the servers, but ICMP will give you availability and response time for the service as a whole.
SolarWinds solutions are rooted in our deep connection to our user base in the THWACK® online community. More than 150,000 members are here to solve problems, share technology and best practices, and directly contribute to our product development process. Learn more today by joining now.