PreFork — specify whether Interchange server should pre-fork processes that wait for client connections
For each new client connection, Interchange must spawn a new process that will
handle communication with the client.
By using PreFork, it is possible to pre-fork
(or in other words, "start ahead of time") some number of Interchange servers
that will be ready to serve client requests immediately as they come in.
If you're running PreFork, you should always try PreForkSingleFork too
to start. Only turn off PreForkSingleFork if something pathological
happens (the most likely outcome being a slew of zombies piling up that
are never reaped). Using it successfully creates the absolute minimal
number of forks necessary.
This method reduces system overhead due to forking and is the fastest and best way to run a busy Interchange server.
Regardless of the PreFork setting, each spawned server will serve
MaxRequestsPerChild requests, before being shut down and respawned.
(This technique is used to prevent any memory leaks).
For an introduction to Interchange run modes and practical implications, please see the ic run mode glossary entry.
Interchange 5.7.0:
Source: lib/Vend/Config.pm
Line 5260 (context shows lines 5260-5272)
sub parse_yesno {
my($var, $value) = @_;
$_ = $value;
if (m/^y/i || m/^t/i || m/^1/ || m/^on/i) {
return 1;
}
elsif (m/^n/i || m/^f/i || m/^0/ || m/^of/i) {
return 0;
}
else {
config_error("Use 'yes' or 'no' for the $var directive\n");
}
}