Nagios Support Forum

before version 5.6 performance graphs had gaps in them where the check was not returning graph data

But since the upgrade I have noticed that what would previously have been a gap is now a constant line of the last value recorded until the check responds again

has this actually changed or am I just going mad?

The drawing of the Straight line is how the graphing system works in the newer version of the application that draws the graph.
If the check stops returning performance data, as the check (Host or Service) is down, when it comes back, the application back fills the data to the last time the check ran.
This is especially true for passive checks. If the check was not received, there is not any Performance data, so it is filled in the next time a check is received.

Ah ok, so, like if there is disconnect between the Nagios server and the host it stores the data locally on the host and then back fills?

If the host check spools up the data and when it can reconnect, it should back fill in the data.
What I feel is happening it that there is not any stored data on the remote host so the graphing function draws a line from the last data point to the current data point.
That is where the straight line comes in.

any documentation on how we spool and store the information?

What agent are you using for Passive checks?

Its not a passive check - its active, and runs directly from the nagios box to a switch via SNMP

OK, the data for active checks are not spooled anywhere.
As long as the nagios process it running, it will run the checks.
If the host that you are checking was down and the plugin did not return performance data like SNMP checks do when they cannot connect to the remote host, that would cause the issue you are seeing.

Is it possible to deactivate this feature? Its causing havoc with our bandwidth monitoring

Is the check in question a bandwidth check using the check_rrdtraf plugin?