The first tests I did where with a routed topology through 2 StarVX routers
but these most recent tests is with all interfaces bridged and the StarVX
client as a wireless client bridge which I believe is based on WDS.

If your asymmetric performance with the RB532's is between 12 and 18mbps I
wouldn't say your seeing similar performance as I am seeing 40mbps and
27.7mbps when transmitting and receiving at the same time using random
(uncompressible) tcp based data.

The tests are based on default packet sizes however, we run M3P (ip packing)
between the Mikrotiks and when testing with small packets (50-100byte) I see
the same results. The only problem with M3P is that it adds 15ms of delay in
each direction under low loads. This means you see pings with a round trip
time of +30ms. This does however improve when more and more traffic is
passed over the link as M3P either waits for 1500bytes of traffic or 15ms to
expire (which ever comes first).

The difference in upload and download speed I have so far put down to one
system being a 2.4GHz CPU and the other being a 533MHz CPU. Hopefully I can
test with 2 P4's later today.



-----Original Message-----
Behalf Of David E. Smith
Sent: 20 October 2005 04:34
To: WISPA General List
Subject: RE: [WISPA] Atheros speed WRAP vs RB532

Paul Hendry wrote:
> Managed to scavenge a P4 system together for testing purposes.

Hey, as long as you're doubling as my personal IT lab... :)

My tests so far have just been traditional AP/client mode, and I'm seeing
the same sort of asymmetrical performance you are. With a pair of RB532s,
depending on the "random-data" and packet size settings, I'm floating
between 12 and 18Mbps. Thing is, I always get better performance with
traffic going client-to-AP than AP-to-client.

Have you, or will you, test with WDS mode? If you've already done so, are
you getting better or worse performance that way? (I may or may not be
able to try that particular setup before I leave for vacation on Friday.
Probably not, because I've not worked with RouterOS' wireless stuff before
and I may not be able to figure out how to set it up. :)

Also, are you using the default packet sizes or have you been playing with
those too? At least in wired network testing I've done before, the real
worst-case scenario involves using the smallest packets you can get away
with (because of the extra overhead in packet creation).

David Smith
WISPA Wireless List:



No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.12.4/142 - Release Date: 18/10/2005

No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.12.4/143 - Release Date: 19/10/2005

WISPA Wireless List:



Reply via email to