hi,
as we discussed before, I can use acl maxconn for limiting the number of
parallel connections and if the maximum is exceeded, an 'access denied'
is returned.
Because of this I have to find a resonable value for the maxconn number.
I made some quick tests and these showed that it takes some time for the
connections to close fully, so it can happen that a request is denied
although there are 'inactive' connections that closed before. Is it
possible to exactly define this timeout interval? Does this depend on
squid or on the OS?
thanks,
Miki
Received on Fri Oct 26 2001 - 04:08:31 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:03:09 MST