ITS#2862 slurpd needs a configurable timeout

[hyc@symas.com]

Just noting this so that it doesn't get lost - another slurpd error 
that's come to light recently - it is possible for slurpd to 
successfully send an update to a slave that is too busy to reply. If 
slurpd's default timeout expires before the reply arrives, the 
connection will be aborted and slurpd will record that the update needs 
to be retried. When the retry occurs, it will fail (because the update 
already succeeded the first time) and generate a (spurious) reject log 
record.

The ability to specify a longer timeout can reduce the occurrence of the 
problem, but I'm not sure what else can be done to eliminate it completely.

-- 
  -- Howard Chu
  Chief Architect, Symas Corp.       Director, Highland Sun
  http://www.symas.com               http://highlandsun.com/hyc
  Symas: Premier OpenSource Development and Support