Trigger question...

[joshuamcdo]

Since correcting a major issue with the zabbix installation that I have been stuck with. Some new problems have arisen.
I am pasting a trigger that is supposed to trigger in the event that connectivity fails from the front end to the 4 mongodb servers.

<triggers>
{front-end-severname:net.tcp.port[mongo1dbhostname.com,27017].last(3)}=0 | {front-end-severname:net.tcp.port[mongo2dbhostname.com,27017].last(3)}=0 | {front-end-severname:net.tcp.port[mongo3dbhostname.com,27017].last(3)}=0 | {front-end-severname:net.tcp.port[mongo4dbhostname.com,27017].last(3)}=0
</triggers>

If I understand this correctly. I use IF very prominently.
{front-end-severname:net.tcp.port[mongo1dbhostname.com,27017].last(3)}=0

The above says "If connectivity to from front-end-servername to mongodbhostname.com on port 27017 fails for more than 3 seconds. Then trigger.. Am I understanding that correctly?

The problem is that we keep getting alerts that this connectivity has beeb interrupted but then immediately after we get a recovery email.
Any help is appreciated.


Thanks,
J

[steveboyson]

It means that
if
the frontend server failes ONE of the connections in the last 3 seconds
then
trigger fires.

I assume that would require that you run these tests quite frequently and very likely it will have a connection loss on 4 backend servers within 3 seconds.

[joshuamcdo]

This doesn't sound like to the best way to test for these...
I need to find a better way, something more like .. If these things fail to connect, and do so for more than 20 seconds, then trigger.. More reading.. YaY!

J

It means that
if
the frontend server failes ONE of the connections in the last 3 seconds
then
trigger fires.

I assume that would require that you run these tests quite frequently and very likely it will have a connection loss on 4 backend servers within 3 seconds.