check_nrpe using NSClient++ gives random socket timeout

[futureweb]

Hey there,
already roamed through dozens of postings here - but unfortunately none of them helped me solving this problem. Maybe one of you got another idea which helps me?! :-)

I'm using check_nrpe to NSClient++ for Monitoring some Windows Servers.
"Normal" Checks (Disk/CPU/Memory/...) all working fine and reliable, but the "Check HyperV Replication Status" Check is flapping massively. About every other check it gets "Socket Timeout" ... Already tried different Versions of check_nrpe - example:

#  /dl/nrpe-4.1.0/src/check_nrpe  -H server.dom.tld -t 60 -u -p 5668 -c CheckHyperVReplica  -2 -4
Normal
#  /dl/nrpe-4.1.0/src/check_nrpe  -H server.dom.tld-t 60 -u -p 5668 -c CheckHyperVReplica  -2 -4
Normal
#  /dl/nrpe-4.1.0/src/check_nrpe  -H server.dom.tld -t 60 -u -p 5668 -c CheckHyperVReplica  -2 -4
CHECK_NRPE STATE UNKNOWN: Socket timeout after 60 seconds.
#  /dl/nrpe-4.1.0/src/check_nrpe  -H server.dom.tld -t 60 -u -p 5668 -c CheckHyperVReplica  -2 -4
Normal
#  /dl/nrpe-4.1.0/src/check_nrpe  -H server.dom.tld -t 60 -u -p 5668 -c CheckHyperVReplica  -2 -4
CHECK_NRPE STATE UNKNOWN: Socket timeout after 60 seconds.

So it's completely random when it fails and when it succeeds. When executing the Check Hyper-V Replica PS Script on Server it always responds in time (5-15 seconds) and never failed / timeout.

Anyone any idea on this? :-)

Thank you
Andy