I'll second that by adding, not only have consistent testing tools for testing each product, but also taking the time to test the testing tools as well. It can make a BIG difference which PC a testing tool is running from. Taking for granted that one P4 router will perform the same as another P4 router is not always the case. recent tests we showed, The chip set or NIC card in the testing machine could make as much as a 50% difference. On the bench I always test betwee nthe two testing machine via patch cable, before injecting the radio we are testing. Its also importnat to correctly set the buffer size of the test tool, to guarantee results. A Radio may pass faster capacity than it can generate itself for a test.
In our tests Mikrotik had always lagged behind, but it was a hardware limit not a software limit. It will be really interesting to see how MT software performs on faster hardware, like their newer processor type models. We had also found that the Ligowave 533 gear performed excellent, in both our lab and field trials, as far as actual throughput, apples to apples comparisons. Tom DeReggi RapidDSL & Wireless, Inc IntAirNet- Fixed Wireless Broadband ----- Original Message ----- From: "Harold Bledsoe" <[EMAIL PROTECTED]> To: "WISPA General List" <[email protected]> Sent: Friday, February 08, 2008 11:43 AM Subject: Re: [WISPA] MT vs StarOS > I'm not trying to make things more complicated, but the tool used to > test is also very important. For example, if you use a BW test tool > that pumps compressible data into a wireless link that has compression > turned on, then you might get misleading results. > > IMO, if you are showing your best results, then full packets are OK but > the packet fill should be random or at least consistent across the test > platforms. You also, need to know what the results are showing you from > the test. An example is Atlas's built-in test tool that reports in the > manual that: > > "Throughput numbers represent the absolute maximum bi-directional > capacity including the management header (which is not available for > user payload) so actual user throughput measurements will be slightly > less." > > I am not saying that this is good or bad, just that it should be a > factor in comparisons (as is the case with any benchmark testing). > > -Hal > > -----Original Message----- > From: Matt Larsen - Lists <[EMAIL PROTECTED]> > Reply-To: WISPA General List <[email protected]> > To: WISPA General List <[email protected]> > Subject: Re: [WISPA] MT vs StarOS > Date: Fri, 08 Feb 2008 00:31:42 -0700 > > OK, here are some real world examples: > > 5.3ghz 40mhz channel, 8.5 miles. WAR4/CM9 attached to 29db PacWireless > Antennas. 50meg one way. > 5.7ghz 20mhz channel, 10 miles. WAR2/CM9 attached to 29db PacWireless > Antennas. 45meg one way (CPU maxes out on WAR2s) > 5.7ghz 10mhz channel, 42 miles WAR4/SR5 attached to 34db Radiowaves > Antennas. 18meg one way. > FDD 5.2ghz 20mhz channel, 10 miles X4000/R52H attached to 29db > Pacwireless/21db MTI panel. > 5.8ghz 20mhz channel, 10 miles X4000/R52H attached to 21db > MTI Panel/29db Pacwireless 30 down, 30 up > > for comparison: Motorola Gemini Lite (30meg bh) on the same link - 3meg > down, 3 up > > We have several RB532 backhauls in the air, and even with all the > optimization we can do to them, we never get better than 20 meg in one > direction. Replacing the RB532s with WAR2s (266mhz CPU, comparable to > an RB532) nearly doubles the speed. > > On comparable hardware, I have not come across ANY Mikrotik system that > will keep up with StarOS when it comes to dedicated backhaul. > > Matt Larsen > Inventive Media > > > Mark Nash wrote: >> 5.3GHz 40MHz channel, 1.5 miles. WAR4-METRO/SR5 inside 24dbi RooTennas. >> Testing beyond each ethernet port. 2 weeks non-stop stress-testing after >> we >> installed and before we deployed gave us 70 megs one way. The MRTG graph >> looked like a solid block without any deviation. >> >> Mark Nash >> UnwiredWest >> 78 Centennial Loop >> Suite E >> Eugene, OR 97401 >> 541-998-5555 >> 541-998-5599 fax >> http://www.unwiredwest.com >> ----- Original Message ----- >> From: "Mac Dearman" <[EMAIL PROTECTED]> >> To: "'WISPA General List'" <[email protected]> >> Sent: Thursday, February 07, 2008 3:08 PM >> Subject: Re: [WISPA] MT vs StarOS >> >> >> >>> I haven't seen those results, but I have seen 12MbpsFDX with a -63 on >>> both >>> sides running Nstream2. If we are going to talk REAL THROUGHPUT - - lets >>> get >>> real and everyone use real figures. I ain't talking bench test and >>> "maybe" >>> if I hold one hand in the air, twist my lips standing on one leg. >>> >>> I mean real world - whatcha getting?? Whatcha see and is it a bench test >>> or >>> are you in the real WISP world? :) >>> >>> >>> Mac >>> >>> >>> >>> >>> >>> >>> >>> >>>> -----Original Message----- >>>> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On >>>> Behalf Of Dennis Burgess - Link Techs Inc >>>> Sent: Thursday, February 07, 2008 2:51 PM >>>> To: 'WISPA General List' >>>> Subject: Re: [WISPA] MT vs StarOS >>>> >>>> 50 both ways with N-Stream dual and Turbo mode... >>>> >>>> Dennis M. Burgess >>>> Mikrotik Certified Consultant >>>> Link Technologies, Inc., St. Louis, Missouri >>>> --WISP/Network Support Services-- >>>> +1 314-686-1302 >>>> >>>> >>>> -----Original Message----- >>>> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On >>>> Behalf Of Mike Hammett >>>> Sent: Thursday, February 07, 2008 2:46 PM >>>> To: WISPA General List >>>> Subject: Re: [WISPA] MT vs StarOS >>>> >>>> Mikrotik can do 70 megs or more over 40 MHz as well. >>>> >>>> >>>> ---------- >>>> Mike Hammett >>>> Intelligent Computing Solutions >>>> http://www.ics-il.com >>>> >>>> > > > > -------------------------------------------------------------------------------- > WISPA Wants You! Join today! > http://signup.wispa.org/ > -------------------------------------------------------------------------------- > > WISPA Wireless List: [email protected] > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ > > > -------------------------------------------------------------------------------- > WISPA Wants You! Join today! > http://signup.wispa.org/ > -------------------------------------------------------------------------------- > > WISPA Wireless List: [email protected] > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ -------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: [email protected] Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
