We had this issue on a unix server, where the HPOV agent process is running fine, but the agent status complains with the below error marked in red.

The agent status command also takes a long time to finish responding.

# opcagt -status
scopeux Perf Agent data collector (26896) Running
midaemon Measurement Interface daemon (26727) Running
ttd ARM registration daemon (1144) Running
perfalarm Alarm generator (26915) Running
perfd real time server (26715) Running
coda OV Performance Core COREXT (26765) Running
opcacta OVO Action Agent AGENT,EA (26786) Running
opcle OVO Logfile Encapsulator AGENT,EA (26790) Running
opcmona OVO Monitor Agent AGENT,EA (26792) Running
opcmsga OVO Message Agent AGENT,EA (26774) Running
opcmsgi OVO Message Interceptor AGENT,EA (26788) Running
rtmd HP Real Time Measurement AGENT (26762) Running
Could not contact Message Agent to query buffering state.
Setting SocketPoll to TRUE and clean-starting agent fixed the issue in a few cases

ovconfchg -ns xpl.net -set SocketPoll TRUE

opcagt -kill
ps -ef |grep opc (use kill -9 if any process still exists)
rm /var/opt/OV/tmp/OpC/* (These are your agent Queue files)

opcagt -cleanstart
opcagt -status