I've never seen condor_status so slow before for such a small Condor pool.
This is definitely unusual behavior.
Can you watch the CollectorLog while you do the condor_status query? If I
was on a Unix system, I would do:
* [On the central manager]
tail -f CollectorLog
* [On the client]
condor_status
I'm curious if there are any interesting messages in the CollectorLog that
indicate problems. If there aren't, you can increase the amount of
debugging information and try again.
Are you using any of Condor's security options, like GSI or Kerberos? What
OS are you running on?