RESOLVED: Active Checks Not Working w/ Service Dependencies
Posted: Fri Mar 17, 2017 10:58 am
Hello,
We are running Nagios XI 5.4.3 on RHEL 6.8 64-bit.
I've noticed an issue today, but am not sure if it's related to the upgrade to 5.4.3. Many of our active checks are not updating. By this, I mean the last checked time is old and the next check time simply keeps moving ahead in time. Issuing a forced immediate check updates the last checked time.
For example, our SNMP uptime check is working as expected using a plugin for that. On the same server, though, an SNMP check of interface traffic using a different plugin, check_traffic, isn't updating.
Another example are MySQL checks using a plugin, check_mysql_health. Using the command to check connection time is working, but not for uptime.
Active checks show enabled for the host and the services. How can I troubleshoot this?
Tom
EDIT: I think I see a link in the behavior, as I did add service dependencies yesterday. In this case, I placed all SNMP services dependent on SNMP uptime. Additionally, all MySQL checks are dependent on MySQL connection time.
If this causes this, how can I configure Nagios to only send me one alert when MySQL goes down (i.e. connection time failed) and not for every service (log waits, uptime, etc.)?
I guess I'm looking for a parent/child relationship, like you can do with hosts.
EDIT 2: Indeed, service dependencies cause the issue. So my question above is relevant.
RESOLUTION: I had the service dependency execution and failure options configured with Up, in addition to Warning and Critical. By removing this, the checks are performing as expected.
Side Note: Why is it that I can troubleshoot for a few hours and find nothing, then when I post an entry here, it all starts coming together?
We are running Nagios XI 5.4.3 on RHEL 6.8 64-bit.
I've noticed an issue today, but am not sure if it's related to the upgrade to 5.4.3. Many of our active checks are not updating. By this, I mean the last checked time is old and the next check time simply keeps moving ahead in time. Issuing a forced immediate check updates the last checked time.
For example, our SNMP uptime check is working as expected using a plugin for that. On the same server, though, an SNMP check of interface traffic using a different plugin, check_traffic, isn't updating.
Another example are MySQL checks using a plugin, check_mysql_health. Using the command to check connection time is working, but not for uptime.
Active checks show enabled for the host and the services. How can I troubleshoot this?
Tom
EDIT: I think I see a link in the behavior, as I did add service dependencies yesterday. In this case, I placed all SNMP services dependent on SNMP uptime. Additionally, all MySQL checks are dependent on MySQL connection time.
If this causes this, how can I configure Nagios to only send me one alert when MySQL goes down (i.e. connection time failed) and not for every service (log waits, uptime, etc.)?
I guess I'm looking for a parent/child relationship, like you can do with hosts.
EDIT 2: Indeed, service dependencies cause the issue. So my question above is relevant.
RESOLUTION: I had the service dependency execution and failure options configured with Up, in addition to Warning and Critical. By removing this, the checks are performing as expected.
Side Note: Why is it that I can troubleshoot for a few hours and find nothing, then when I post an entry here, it all starts coming together?