[ Return To FAQ Index | Search The FAQs ]
|Title:||Troubleshooting undetermined times in trends and availability CGIs|
|Submitted By:||Ethan Galstad |
When user generates host/service reports using the trends and availability CGIs, they get a percent of time reported as being 'Undetermined'. Sometimes the report will show '100% Undetermined' time. What does this mean?
The trends and availability CGIs were designed to not lie to you (or your manager). They make the best attempt to give an honest report of the state of a host or service based on alert entries found in the log files. If there are chunks of time in the reporting period that the CGIs cannot determine the state of a host/service for, those chunks get counted as undetermined.
If you get 100% of the period being reported as undetermined, the problem lies with the fact that the CGIs could not determine the initial/start state of the host/service for that period. You can increase the chances of the CGIs finding an initial state for the host/service by increasing the value of the Backtracked Archives option. If this does not help and you know what the state of the host/service was at the start of the report period, you can tell the CGIs what that was by selecting it from the First assumed state dropdown box.
If only a portion of the reporting period is being shown as undetermined, it may be due to two things:
- Nagios was not running during the undetermined times. If this is the case, you're out of luck. The CGIs cannot be forced to report the state of a host/service when Nagios was not running.
- The Nagios process was restarted one or more times during the reporting period. If you have state retention enabled, set the Assume state retention option to yes. If you do not have state retention enabled, you might try setting the Assume initial states option to yes.
|F0135||What does the backtracked archives option in the trends and availability CGIs do/mean?|
|Keywords:||undetermined time trends availability cgis |