Re: Redundancy - why not do it on Unix?

You could protect from this by doing something like it on unix...

  * get two squid processes to run re-using the same TCP port but using
    seperate ICP ports (or the same, and both recieving the queries?).
    this shouldn't take much code modification. (like apache, two processes
    listening on one socket.)

  * setup a machine with an IP alias of your main cache machine somehow
    modified so it is slow at responding to ARP requests, or make a job
    insert the IP alias only when the other cache is detected as being down.

Not for mainstream squid, but maybe at least the first one would help maintain
an avaliable cache for a greater period of time.

David.
Received on Thu Nov 06 1997 - 17:57:28 MST