NRPE check for failures of persistent volume mounting

On a customer cloud during OpenStack upgrades, we encountered https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1847361 which created problems with persistent volume mounting for K8s on OpenStack:

Failed to open module: /usr/lib/x86_64-linux-gnu/qemu/block-rbd.so: undefined symbol: RbdAuthMode_lookup
Failed to open module: /var/run/qemu/_Debian_1_2.11+dfsg-1ubuntu7.31_/block-rbd.so: failed to map segment from shared object

This was resolved by restarting the K8s worker instances after we root caused and identified the issue.

We may want to look into whether we can detect failures in persistent volume mounting in some way, so as to better detect a case such as the above, assuming such a check makes sense overall.