|
> -----Original Message-----
> From: Benjamin Lee [mailto:benjaminlee@xxxxxxxxxxxxxx]
> Sent: Wednesday, November 01, 2000 10:59 AM
> To: Trevor Marshall
> Cc: Linux Virtual Server Mail List
> Subject: Re: HTTPS and cookies
>
>
> Trevor, see inserted comments below...
>
> On Wed, Nov 01, 2000 at 06:44:26PM +0000, Trevor Marshall wrote:
> > Client1 connects from AOL and is routed to RIP1
> > Client2 connects from AOL before Client1's session times out,
> so Client2 is
> > routed to RIP1.
>
> Yes. However the distinction must be made that a 'session' in LVS
> connection affinity is NOT related to 'sessions' as created by
> HTTP cookies.
This got me wondering... Why does persitance matter to cookies? If you're
keeping sessions, wouldn't you just store the info in a shared database and
be done with it? The same shared database that your dynamic content comes
from, could be used to store the session information.
Lets not get into the HA aspects, but just load balancing. If one could turn
off persistance, one would not have the problem of the thousands of AOL(or
large corporate organization) users sharing a proxy hammering one server. I
know Slashdot uses 3 Debian web servers, pointing at one RedHat MySQL
server(strangely, they don't use LVS or BigIP, but someone else's hardware
solution). Doesn't this sound like the best idea?
>
> > The cookies can then be sent to ensure the two users' sessions aren't
> > crossed?
>
> See above.
>
> > I'm sorry Joseph's 1999 lvs-howto gave me the impression that ppc used
> > source port in routing decisions, or is the current version of ipvs not
> > using ppc.
>
<snip>
>
|
