How prevent flapping triggers?

[Tux]

Hi,

how can i change my trigger to get only problem state, if problem still exist after X mins/checks?

example Trigger:

{Template DB Server:vfs.fs.size[{#FSNAME},pfree].last(0)}<5

Im not really sure, how to use the "sum#" correctly:

{Template DB Server:vfs.fs.size[{#FSNAME},pfree].sum(#3)}=0

i want: if free space on #FSNAME ist under 5% about 3 checks, then alert this problem (want prevent flapping and a lot of OK,PROBLEM mails...). Or ist last(3)<5 better for this?

can anyone help?

thx

PS: Ups, maybe wrong section. would it be better in "Zabbix Troubleshooting and Problems" ?

[cferraro]

Tux,

You may want to use the .count trigger function as described in

https://www.zabbix.com/documentation/doku.php?id=1.8/manual/config/triggers

Something like this should work for you:

{Template DB Server:vfs.fs.size[{#FSNAME},pfree].count(#3,5,"lt")}>2

This function returns the number of values which are less than 5 over the last 3 checks. If that number is greater than 2 it'll trigger.

Basically, this would trigger if your last 3 checks were less than 5%.

[Tux]

Thx m8, i will test it soon and report it here.

[Michael0]

I had the same problem for our server, I solved it this way:

For example with CPU utilization:

{servername:system.cpu.util[,].avg(600)}>80

If the average CPU utilization is higher then 80% for the last 10mins, then the trigger will alert me.

Also you could use someting like this:

{servername:system.cpu.util[,].last(0)}>80 &
{servername:system.cpu.util[,].prev(0)}>80 &

If the CPU utilization from the last and the previous check is higher than 80%

Hope this helps