check_icmp 'loses' perfdata labels on DOWN/CRITICAL

[log1-c]

When using check_icmp for checks on hosts/services the checks perfdata output changes based on the state, which is a problem for the old pnp4nagios graphing tool.

Check that reports ok/up:

Check that reports critical/down:

This results in the graph for packets lost containing rta ms values instead of the pl % ones:

pnp4nagios is not set up to have multiple files for the perfdata, as this is not suitable for the setup.

I would like to propose that check_icmp retains the other perfdata labels in the critical/down output and fills them with the warning/critical values (like check_ping does) or just zero..

system is CentOS 7, Plugin Version is check_icmp v2.2.1.git (nagios-plugins 2.2.1)

The check_icmp plugin (on Ubuntu) for the monitoring-plugins project does not have this problem

[sawolf]

Good catch - you're right that the performance data tags shouldn't change based on plugin status. I've added this for the next bugfix release

[sawolf]

The fix turned out to be relatively simple, so if you're willing to compile the plugin manually, check the branch linked in the pull request above.

[log1-c]

Tried it out, works.
Thanks for the quick fix!