Re: Redundancy

  Yes , I sometimes have the same
  problem. I solution would be to
  add an use the ip belonging to
  the domain (cache.whatever) on the
  two servers as an alias , and to
  remove it off the one server , and
  to set up on the second server
  /sbin/ifconfig eth0:0 ipaddress.
  
  There are things like clashes , and getting
  the ciscos to refresh and so forth ,
  but it might be a solution when redundancy
  is an issue. I find the problem is when clients
  run there own cache only dns servers , and having
  two top level domains with an six hour (default I think)
  refresh rate , might be an idea to set it down to an hour
  seeing they are on there the same network.

  Also , I saw an psoting earlier about someone who
  wrote the code to simply listen on the port , see if the
  cache is up , otherwise do direct query , I'm very interested ,
  otherwise , do you ppl find automatic proxie does not give problems ?

  Hope this adds to the thread , love to write more
  but time is a factor :)

--Daniel Schroder
  Networld
  http://livewire.new.co.za (F) 419 3212
  mailto:daniel@new.co.za (T) 419 4430
  #linux (IRC) see www.zanet.org.za
  sms +2782 950 7747

On Fri, 7 Nov 1997, Andrew Cormack wrote:

> On Fri, 7 Nov 1997, Daniel Schroder wrote:
>
> > I reboot every friday. Prior to that and any reconfigurations
> > we change the dns to point to the machine were squid was initially
> > developed , wait until all connections are on the other machine
> > and have full access to the main proxie
>
> Daniel,
> I'd be interested to hear how long you have to wait! I followed the same
> procedure last time I had to do a software upgrade on our cache, and an
> hour after changing over the DNS pointer, requests were still coming in to
> the "old" cache. It looks as if browsers take one look at the DNS when
> they start, and never (or at best seldom) go back to it. So as long as
> someone has still got a browser running, connections to the "old" cache
> will keep on coming.
>
> I'm not sure what effect shutting down the cache would have on such a
> browser. I'd hope that the browser would check the DNS again before giving
> up in disgust, but probably I'm being too optimistic!
>
> Andrew
>
> --
> +-----------------------------------------------------------------+
> | Andrew Cormack | email: cormack@cardiff.ac.uk |
> | Information Systems Support, | snail: 40/41 Park Place, Cardiff |
> | postmaster and webmaster | CF1 3BB |
> | University of Wales, Cardiff | |
> | Web page and PGP key: http://www.cf.ac.uk/People/cormack.html |
> +-----------------------------------------------------------------+
>
Received on Fri Nov 07 1997 - 02:40:29 MST