>>>>> "MY" == Michael Yount <csf@moscow.com> writes:
MY> The latest changes seem to be working well, but the new RCPT TO code
MY> probably hasn't been exercised yet.
I've tried it and it seems OK, but it's hard to test this kind of thing.
There is something I did in TLB which might be useful. I had a debug mode
that timed out 50% of all SMTP transactions just to make sure the host
failover worked. It might work to just replace
sub RCPT { shift->transact("RCPT TO: <".shift().">",5)}
in SMTP.pm with
sub RCPT { return 0 if (.50 > rand()); shift->transact("RCPT TO: <".shift().">",5)}
or change the 5 to something small like .1 so you see more timeouts.
- J<
References:
|
|
