Oleg Kobchenko wrote:
> My point is the statistic like you are giving is not
> the true availability of Y!, but the artifact of
> the retry logic of particular MTA. Y! is always available
> within seconds after a sufficient number of retries.

I don't think that changing retry timing is a good solution.

In general, retries are increasingly delayed because
the retry cases include the case where the server
is overloaded.

-- 
Raul

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to