Page 1 of 1

performance logging - how to make cyclic

Posted: Tue Jul 07, 2020 1:36 pm
by gmills
Hello,

I'm trying to find out if it is possible to have the performance logging be cyclic when it reaches a certain size ?

so that if the service-perfdata reaches say 1Gb, then rollover to write over the existing file? Not sure exactly how this logging mechanism works.

this is for
host-perfdata
host-perfdata.out
service-perfdata
service-perfdata.out

thank you

Re: performance logging - how to make cyclic

Posted: Wed Jul 22, 2020 1:19 pm
by gmills
am I on the right site for this question ?

Re: performance logging - how to make cyclic

Posted: Fri Jul 24, 2020 5:13 pm
by benjaminsmith
Hi @gmills

The performance graphs are store time-series RRD files so they always stay the same size. This keeps the disks space requirements for XI in check.

If you have a bunch of older hosts and services that you no longer use using up disk space, you can set up the Deadpool processor to remove those along with the perf data if desired.

How To Use Deadpool In Nagios XI

Let me know if that answers your question or not.