Hello,
Does anybody know if can I change the default values of "Assume Initial States" and "First Assumed Host State" in Nagios XI reporting, from yes to NO and from Host UP to unspecified? I don't see an option in the nagios.cfg file. Has anybody done that?
Kind Regards,
George
Assume Initial States - First Assumed Host State
Re: Assume Initial States - First Assumed Host State
Actually I found a bigger issue, where if the state change does not exist in the log (I suppose) file, the device seems to be DOWN for the whole duration of the reporting period. In the database, I keep the log entries for 30 days, but it doesn't seem to affect it. I guess it's the logs written on Nagios XI server itself?
Below the example I have:
I have a host that was down since the last 6 days. If I create a report with Assume Initial States - First Assumed Host State as yes and HOST UP, for the last 24 hours I get 57% UP and 43% DOWN, which is obviously wrong. If I use No and Unspecified, I get 100% unavailability for the last 24 hours (which is correct), but I also get 100% DOWN for the whole month, which is again wrong.
As I said, I suspect it's the log files on Nagios XI server that do not have the state change of this specific host, hence Nagios cannot understand when it went down. Can someone help me to understand which log files are these (to check if I have a retention on the server itself) or maybe I have not understood the root cause correctly?
Kind Regards
Below the example I have:
I have a host that was down since the last 6 days. If I create a report with Assume Initial States - First Assumed Host State as yes and HOST UP, for the last 24 hours I get 57% UP and 43% DOWN, which is obviously wrong. If I use No and Unspecified, I get 100% unavailability for the last 24 hours (which is correct), but I also get 100% DOWN for the whole month, which is again wrong.
As I said, I suspect it's the log files on Nagios XI server that do not have the state change of this specific host, hence Nagios cannot understand when it went down. Can someone help me to understand which log files are these (to check if I have a retention on the server itself) or maybe I have not understood the root cause correctly?
Kind Regards
Re: Assume Initial States - First Assumed Host State
Hi @coitop,
The reporting options you asked about initially are form options for the reports and don't have default values that you can edit, unfortunately.
For the second issue: Reporting only looks at the period of data specified in the report settings -- it will not check the log history to find the initial state, which is the purpose of the Assume Initial States and First Assumed Host State values. If the service was down during the reporting period and you start with initial values assuming it was up, the data would be wrong, that makes sense.
To get accurate data from a report, you will need to know the state at the time the report stars and set the "Assume Initial States" value to yes, and "First Assumed Host State" to the actual host state at the time. This should produce an accurate result. I don't believe retention is a factor here because as long as there is log data covered by the reporting period, the initial state comes from the report settings, not the log data. Omitting state changes in the log data would lead to incorrect results because the report will have no way of knowing that the assumed initial state has changed. Setting the starting values to be accurate should produce results as long as the log data has measurable values, which it sounds like in this case it may not.
Let me know if you have any other questions about this!
The reporting options you asked about initially are form options for the reports and don't have default values that you can edit, unfortunately.
For the second issue: Reporting only looks at the period of data specified in the report settings -- it will not check the log history to find the initial state, which is the purpose of the Assume Initial States and First Assumed Host State values. If the service was down during the reporting period and you start with initial values assuming it was up, the data would be wrong, that makes sense.
To get accurate data from a report, you will need to know the state at the time the report stars and set the "Assume Initial States" value to yes, and "First Assumed Host State" to the actual host state at the time. This should produce an accurate result. I don't believe retention is a factor here because as long as there is log data covered by the reporting period, the initial state comes from the report settings, not the log data. Omitting state changes in the log data would lead to incorrect results because the report will have no way of knowing that the assumed initial state has changed. Setting the starting values to be accurate should produce results as long as the log data has measurable values, which it sounds like in this case it may not.
Let me know if you have any other questions about this!