Re: Cobalt

From: Stefan Rompf <>
Date: Sat, 03 Oct 1998 12:18:27 +0100

At 21:16 03.10.98 -0500, Brian Feeny wrote:

>async I/O as in an async filesystem? Is that a reality on linux yet?
>Ultimitley, the plan is to grow a "farm" of Cobalts. I like the idea of
>multiple processors/machines for speed and redundancy. Could you comment
>for example, would 2 CacheRaQ's of the above config do better than my
>squid box? I know my RAID0 Ultra/Wide disk array is superior to an
>UltraATA single drive, but I was thinking some of whats lost in the cobalt
>could be made up with the 64bit processor, dma io, and possibly a farm of
>cobalts (a la alteon).

Using multiple caches for load splitting is a VERY bad idea because you
will not only share the load, but also the hit ratio. There is no chance
setting up a sibling relationship between all boxes as this will give more
load to every box than a single system has.

If you need redundancy, use a second cache that can overtake first's IP
address or do some magic with the browser's proxy auto configuration (this
won't work for MSIE). For caching, a "real" computer which allows you to
add another SCSI drive easily seems more usable than a Cobalt box. PCI SCSI
and ethernet controllers use DMA since decades ;-)

cu.. Stefan

 | Customer: I'm using Windows '95. Hotline: Ok, got that one. |
 | Customer: It's not working. Hotline: You already said that. |
Received on Mon Oct 05 1998 - 01:34:24 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:42:20 MST