Earlier today, Jimmy Stewpot wrote:
> Is there any way with squid that you can select where it fetches the data
> from if different from the default gateway on the system? Without using
> Routing protocols of some sort.
Squid knows it has to fetch data from some target IP address, chosen from its
configuration of peers, destination systems, parents and whatnot. *How* the
request passes between Squid and the target IP address is none of Squid's
concern.
What you want to do is *exactly* the reason why routing protocols exist - for
path selection given multiple paths. If you don't have multiple paths, why
bother using routing protocols at all? Just hardcode the default route.
> Basically we are after some method that if one link fails it would
> automatically start to fetch the information from a different link.
Configure your routing accordingly and the traffic will automatically flow
from one link to the next in the event of a failure. If you only want
Squid-related traffic to automatically failover, then that will depend on the
functions/features of your routing software (beyond just kernel routing
tables).
Cheers..
dave
Received on Mon Apr 10 2000 - 22:13:02 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:52:53 MST