Benchmark results

[David Collier-Brown]

Odd:  http://www.zdnet.com/sr/stories/issue/0,4537,387506,00.html
say the server has a singe 4GB ide drive. Not a terribly fast device...

  However, the chart at 
http://www.zdnet.com/sr/stories/issue/0,4537,2196106,00.html
says we got about 32 "mbits per second".

  32 mbits/s == 4 mbytes/s, assuming 8-bit bytes.
  Purely for argument, assume no overhead for SMB and TCP/IP headers.

  A single 7200 rpm scsi drive will start about 70 i/o operations a
second,
and if we assume that 100% of them are for data, and the buffer size is
8 KB, we'll get 560 KB/Sec of throughput from one drive.
(In fact, less than 100% are for data).

  So we're getting 4 mbytes/s on the wire, while reading .560 mbyes/sec
off the disk.

  Methinks this benchmark allows lots of server-side caching (;-))
Anyone care to comment on the method(ology)?

--dave
--
David Collier-Brown,  | Always do right. This will gratify some people
185 Ellerslie Ave.,   | and astonish the rest.        -- Mark Twain
Willowdale, Ontario   | http://java.science.yorku.ca/~davecb
Work: (905) 477-0437 Home: (416) 223-8968 Email: davecb at canada.sun.com