[9fans] NOW it's like lightning ... almost

[9fans] NOW it's like lightning ... almost

[Matt Pidd-Cheshire]

> I just did a quick timing test and it would appear that the slowness
> is a local problem, not a generic Plan 9 Ethernet problem.  Reading
> ~4MB (/386/lib/*) from a Ken fs with client-side caching disabled took
> 17.89 seconds the first time (Ken fs doesn't use DMA for IDE ever [for
> historical reasons], though I might fix that), which is still roughly
> 260KB/s, an order of magnitude faster than the ~250Kb/sec cited.

Sorry ISC; KB/s is what I /should/ have written. Some measuring will 
tell me if the ATA-5 disc is being DMA'd. WRT to your previous
post, logical that 9 will enable what works and leave us to take it on 
from there where hardware permits.

The ether interface is an Intel PRO/100 using IP4.

> Reading it a second time, with the data in RAM cache on the file
> server, took 1.07 seconds, or 4.2MB/s, far faster than the 2.5Mb/sec
> cited between Windows and BSD.  And that's over slow 100Mb/s Ethernet
> because I haven't got around to upgrading the file server to Intel's
> gigabit card yet.

Mmmm, gigabit. Nope, I just don't have the bandwidth need for that or 
the necessary router - meaning I get clonked on the head with the 
skillet if I make eyes at either.

Again, s/Mb/MB/ and 2.5 is a rough figure depending on what else may 
pass through my WAP/router at the time. It does indeed peak at just 
above 4MB/s and the problem vis a vis 9's /net/ether0 is likely local.
I'll have to set up my combined cpu-auth-fs machine and do some jiggery 
pokery with it.


BTW: is this a moderated group? seems to be a persistent sprinkling of 
snake-oil weight-loss postings ...