[SCore-users] sc_watch problem
Atsushi HORI
hori at swimmy-soft.com
Thu Dec 4 19:03:53 JST 2003
Hi,
>Under heavy (IO ??) load from cluster members the sc_watch seems to
>get some kind
>of a timeout and it decides to rerun itself. This usualy happens
>once or twice
>per night, when we are running many jobs.
The sc_watch periodically communicate with SCore-D, and when a
timeout happens (there is no answer with in a certain time duration),
it assumes something worng.
Under a heady load situation, a timeout can happen because compute
hosts are very busy and response to sc_watch is late.
To avoid this, you may set the timeout duration by "-t" option
followed by a number of time duration in minute. The default timeout
is set to ten minutes.
----
Atsushi HORI
Swimmy Software, Inc.
More information about the SCore-users
mailing list