[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [Condor-users] negotiating with schedds when a client has FW



If I can try to summarize only the 'practical' outcomes of this thread:

To overcome the problem of the negotiate cycle halting once the Negotiator
encounters a machine that apparently fails to respond, the best course of
action is to increase

NEGOTIATOR_TIMEOUT

on the submitting machines. Suggested values might be 45-60 seconds. This is
to be preferred to setting 

NEGOTIATE_ALL_JOBS_IN_CLUSTER = True

which appears to work well in practice but fails to address the true reason
for the 'fallover'.

Would the above appear to be reasonable ?