[Prev][Next][Index][Thread]

Re: good practice for clients

"N.G.Smith" writes:

| For those who don't follow uk newsgroups, this comes after some heated
| discussion, where I was trying to encourage users of caches to accept
| the fact that they were at least partly to blame in the case where
| their WWW access was interrupted by their browser's inability to
| degrade gracefully in the light of a proxy failure.

Well, you did kind of imply that everyone ought to be running Netscape >
= v2 ! <ducks>

| >  [ + detect and act appropriately when servers have multiple IP
| >      addresses ?  Problems with DNS resolver code here ? ]
| 
| I don't know why this is in brackets. What's stopping resilient clients
| being written now (other than nscd of course), what problems do you
| anticipate in the DNS resolver code?

Yep - cheers, Sun!

One particular problem I've seen here and there is with resolvers only 
returning a single IP address even though there are several associated 
with the DNS entry.  Caching DNS lookups for the duration of the 
session (e.g. while DOS TSR loaded), even if the address(es) stop 
working, can also be a bit of a bugger.

I may be being pointing the finger at the wrong culprit here, though - 
perhaps it's the browser or the TCP/IP implementation which isn't doing 
anything with those multiple IP addresses, and they are all being 
returned by the DNS lookup ?

There isn't the time or the inclination for me to go around checking 
the innards of all these wretched Mac/M$ setups (unless someone is 
offering to make it worth my while :-), but a little light testing 
suggests that we at least have a problem in this department.  I suspect 
others do too.  Any corroborating evidence ?

| >Does the above sound reasonable as a baseline ?  I realise that this
| >stuff isn't universally supported, but that's another story!
| 
| I find it unbelievable that it isn't universally supported.

Yep - in-lined infomercials must be a higher priority!

From my experiences hacking at the X Mosaic proxy support (feel sorry 
for the NCSA guys, I'm a pretty lame programmer) it's certainly not a 
big deal.  But...  lots of people are working with code which is only 
expecting a single proxy, or if you're lucky maybe one proxy per 
protocol and no fallbacks ?

PS I think it would be quite useful if the CERN server were capable of 
surviving the outer proxy's death automatically, e.g. by using other 
proxies and/or direct connections.  Before I get out my lame 
programmer's hacking kit, does anyone have patches to make it do this ? 
 Haven't come across any...
Follow-Ups: