I/O Wait more than 20%

[avandemore]

For best Nagios performance it should be a single purpose system. That means you should remove tomcat from this system along with any other applications not for the purpose of monitoring.

The system could also benefit from more ram but that may not be required if you follow the first suggestion.

[tmcdonald]

Just checking in since we have not heard from you in a while. Did @avandemore's post clear things up or has the issue otherwise been resolved?

[ericssonvietnam]

Just checking in since we have not heard from you in a while. Did @avandemore's post clear things up or has the issue otherwise been resolved?

I have observed in past and found that as soon as we restart the nagios service below is the trend in message queue and as we start the monitoring engine manually again the message queue starts decreasing but we have more than one message queue.
Can you tell me the reason for the behavior of the application need Root cause for the issue.

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0xa9050002 4947968 nagios 600 0 0

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0xca050002 4980736 nagios 600 7818240 7635

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0xca050002 4980736 nagios 600 4893696 4779

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0x01050002 5046272 nagios 600 6282240 6135
0x8e050002 5013505 nagios 600 0 0

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0x01050002 5046272 nagios 600 4819968 4707
0x8e050002 5013505 nagios 600 0 0

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0x01050002 5046272 nagios 600 3789824 3701
0x8e050002 5013505 nagios 600 0 0

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0x01050002 5046272 nagios 600 1311744 1281
0x8e050002 5013505 nagios 600 0 0

[root@nagiosxi ~]# ipcs -q

------ Message Queues --------
key msqid owner perms used-bytes messages
0x01050002 5046272 nagios 600 59392 58
0x8e050002 5013505 nagios 600 0 0

[tgriep]

When the Nagios process writes data to the MYSQL database it uses the ndo2bd daemon and that creates a Queue.
Where I think the extra Message Queue is coming from is that when you restarted the nagios process, it could not connect to the ndo2db daemon as it was busy processing other messages so it creates a new queue and uses that one.
Or if the ndo2db process was down when the nagios process was restarted, that could cause the extra queues as well.