I have the same scenario: Servidor_A and Servidor_B. Each one have sitescope and manages 2500 monitors each.
1) Wich one of the following alternatives can we have for failover?
a) Can one act as the failover of the other? Servidor_A is failover of Servidor_B and Servidor_B is failover of Servidor_A. b) Can we have a third Servidor_C that acts as failover of the other two: Servidor_A and Servidor_B c) I need Servidor_C that acts as failover of A and Servido_D that as failover of B
You can configure one SiteScope server to be a failover for one or more other (primary) SiteScope server(s), but it must be a special failover install of SiteScope, not a regular SiteScope install. So if you already have 2 regular SiteScope servers (A and B) that monitor something, they cannot act as failover for each other.
If you do not use the High Availability feature and special install as listed above you could do this.
I am going to assume your two servers each have different monitors that ar enot on both servers.
Server A 2500 monitors.
Server B 2500 monitors.
Server A contains copy of server B monitors as well in disabled state
Server B contains copy of server A monitors as well in disabled state
Server A monitors SiteScope (Be it process or url monitor login whatever you choose) If server B is in error Run Alert to enable Disable Server B Monitors
Server B monitors SiteScope (Be it process or url monitor login whatever you choose) If server A is in error Run Alert to enable Disable Server A Monitors
Now you have fail over.
You could also have an alert that Server A logs into Server B (if server is up and running) runs start script. If fails to restart Run alert to enable Server B Disabled Monitors on server A.
Don't forget to add an alert that also will disable the fail over monitors if the SiteScope instance on the other server comes back up.
That's failover without special install or HA implimentation.
If you need servers C and D (because A or B can't handle 5000 monitors) well that's up to you just follow the same type of scenarios as above chanings A or B for C and D respectively.
You will also need a method or script that should post to the SiteScope URL of the downed instance to go to the other instance for monitoring until it's back up and running. Or use some hardware load balancer.
Disabled monitors will still consume license points. Also, you will have problems wrt reporting as monitor information stored in profile database for the original and copy will be in totally different places.
HA is the only reliable implementation wrt Sitescope failover.
If you don't have your metrics updating a database and use the standardstorage for metrics and reports on the SS server it is true you would have data in different locations. But it's still not a deal breaker. Nor without ways to manage it.
Yes it will utilize more points.
But HA is not the only reliable way.
It's the easiest.
Some would say the Best Practice.
But it's by far the only solution. }:>)
There's many ways and some are more cost effective than others.
SiteScope is one of THE BEST tools that prove there is more than one way to skin that cat!
Saying there's only one way is like saying SiteScope is not a swiss army knife of the monitoring universe. };>)