Checking NRPE certificate expiry without flooding logs

[xmasbofh]

For a while I've been using check_http to check the expiry of the TLS certificate and general health of my NRPE daemon with the command:

/usr/lib/nagios/plugins/check_http --ssl -H '$HOSTADDRESS$' -I '$HOSTADDRESS$' -p 5666 -C 30,7

This largely works fine, although has the mildly annoying side effect of generating the following in the log file:

Could not read request from client 192.168.0.116, bailing out...
INFO: SSL Socket Shutdown.

I guess this is because check_http doesn't speak NRPE and that upsets it. I Googled the protocol and it looks like it's not terribly easy to configure an existing tool to play nicely. That said, I thought I'd ask the question: Anyone know how to check the cert expiry without causing the log file noise?

It occures to me that I could create two checks and use check_nrpe to check the health of NRPE and check_http every 12 hours or so to check the cert. Alternatively I could just learn to live with the log noise. Does anyone have any other suggestions, please?

[lgute]

Hi @xmasbofh, thanks for reaching out.

Which log file has these messages?

[xmasbofh]

Sorry for the delay, I neglected to enable notifications.

The logs are written to syslog, which I have filtered off into /var/log/nrpe/nrpe.log