Using TimePeriod to avoid resource crunch... or not?

Hi… We have a series of databases which support multiple environments. We run application-oriented verification service checks on the databases to check various risky conditions. Icinga has been immensely useful keeping us ahead of environment error conditions which are indicated in the database.

The problem is we have hundreds of these checks and they all seem to run together. When we increase the check_interval they still run together - just at different intervals.

I would think this is not an uncommon problem… ? Can someone suggest the idiomatic solution? My thought was to create multiple timeperiods within any given hour and assign different service checks to those periods, e.g.,:

   object TimePeriod "ServiceSet1" {
     ranges = {
       "monday"    = "08:00-8:09", "09:00-09:09", "10:00-10:09", etc
       "tuesday"   = etc
     }
   }

   object TimePeriod "ServiceSet2" {
     ranges = {
       "monday"    = "08:10-8:20", "09:10-09:20", "10:10-10:20", etc
       "tuesday"   = etc
     }
   }

I’m not sure if ranges accepts an array of values like that… ?

Otherwise, any other thoughts or suggests? Thank you!

Andy

hi,

why not write a bash wrapper script? Start with the sleep command. The time is a random value. After run the correct check command like before

Hello @apj68!

You have to write it like in the second example here.

Also (if I remember correctly) you can specify one range: monday - sunday

Best,
AK