Hi Henrik, thx I get it. but... still have some doubts... On Wed, 2007-04-04 at 23:42 +0200, Henrik Nordstrom wrote: > ons 2007-04-04 klockan 16:37 +0100 skrev Joao Ferreira: > > 2 - How can I force them to terminate and restart, in order, for > > example, to reaload fresh information from disk ? > "squid -k rotate". ok, but... wouldn't this also force squid to rotate it's log files (assuming logfile_rotate != 0) ??? Is there a way to _only_ restart the helpers ?? > > Or add support in the helper for reloading it's information without > restart. For example a sighup handler like most software have, or > periodically monitoring the needed resources for change.. (done for > example by ncsa_auth). > This seems to be a good idea, but... how would my external application (the one that writes new data to the disk) know the PIDs of all the helpers in order to send them the sighups ? Would it work to send a signal to Squid and squid would propagate it to my helpers... this could have 'negative consequences' or is there some API in Linux that permits me to find exactlly the PID of my helpers (by their name maybe) ??? > Regards > Henrik Thank you joao jmf
