[PROXYPER] Is 30x a lemon?

Petr Novotny Petr.Novotny at antek.cz
Wed Feb 17 10:02:36 EST 1999

> The use of memory buffers to cache blocks from buffers is a
> performance enhancement designed to reduce the latency involved in
> serving clients and also to reduce disk accesses by combining access
> of the buffers into single operations.

And how does keeping block in memory forever come into the picture? I 
can understand that you have performance gains not writing it 
immediatelly - but not writing it at all, I can see much less of a 
gain. Plus, you're not going to avoid wasting work - in a crash the 
blocks simply disappear. Why would you not then create "checkpoint 
files" for the proxies as much as you do for the clients? I can see 
only a little rationale...

> As I've brought up a number of times before, although it is possible
> to run a proxy on a low-end user machine, it should really be
> treated with the respect that you would offer a dedicated server
> machine.  One would certainly take efforts to avoid powering down
> their multiuser enterprise Solaris or NetWare machine without
> properly shutting down and allowing disks to be synced and umounted.

Our Internet proxy is a seriously underpowered Linux computer on a 
486/33 with hard discs that should be already in a museum; due to irq 
timeouts, it crashes once or twice a week. squid has no problems with 
crashes. qmail has no problems with crashes. No mail has ever been 
lost due to the crash. A few hundred blocks have already been lost, 
and more certainly will. There's something fishy with your 
assumptions I would think.
Petr Novotny, ANTEK CS
Petr.Novotny at antek.cz
-- Don't you know there ain't no devil there's just God when he's drunk.
                                                             [Tom Waits]
To unsubscribe, send 'unsubscribe proxyper' to majordomo at lists.distributed.net

More information about the proxyper mailing list