All, I have a customer reporting significant difference between the wireless performance on new MC9190s that they are testing in an environment that currently has an MC9090 deployment. Looking at some of the info they have sent I'm not convinced the wireless install is great, but it would seem that the MC9090s are coping with it significantly better than the new MC9190s. I had a similar issue with MC3190s, is it just that the JEDI radio is poor compared to the Photon radio? I will get them upgraded to the latest Fusion build, but would appreciate any feedback from your experience. Thanks IJ
MC9190 v MC9090 wireless performance// Expert user has replied. |
4 Replies
Hi Ian Assuming your Fusion build is .19R or .08R. I do know from experience that with respect to excessive roams, reassoc's, missed Beacons, etc... and of course this would also impact Tx Retries... multiple bugs were addressed between .019R and .025R. Good luck. Very odd remark about ICMP responses being set to "very low response priority in WLAN response algorithm". Those darned pings anyway. :) Jud
I just went through this at my customer. My customer was seeing Avg RTT during ping testing10 times greater for the MC9190 over the MC9090. After the Fusion upgrade to 25r, the MC9190 is comparable and in some cases better.
Hi IJ, Not sure what you mean by WLAN performance issue when comparing MC9190 to MC9090 ... RF coverage or throughput (with Robs reply specifically related to ping response). In going along the assumption that Rob followed up with ... performance relating to ping responses ... not sure which OS you have but even the latest BSP releases for the MC9190 on support centeral is VERY OLD. As you and Rob mentioned, try the newest fusion version first. I ran also into the slow ping RTT with our very first sale of the MC9190 a year or so ago ... then when set Fusion to CAM mode and re-run ping test the ping response was very fast. The helpdesk answer to this back at that time was the MC9190 will have slower ping response times as they are trying to conserve battery and not respond as quickly to ping responses (set to very low response priority in WLAN response algorithm ... hmmm).They stated the correct test to prove true transmit/receive speeds would be to compare throughput (not ping response) between the two MC9090/MC9190 with NetIQ (aka Chariot) network test software, or even FTP of a 1 MB file with a software that can report data rates and transfer times. If after trying this and the MC9190 with the most recent Fusion version and transmit/response file transfer speed is still slower ... then next step would be for you to open a case with the helpdesk, have them duplicate your issue, and result in opening an SPR You can also try the reg file below to disable WMM if you do not need it, as others have reported seeing improvements with this. [HKEY_LOCAL_MACHINE\Comm\JEDI10_1\Parms] "APSDConfiguration"=dword:0 "CCXParams"=dword:00000005 Paul
Thanks Gents, I'm getting them to upgrade to -025R now which hopefully will give us the same improvement as Rob has seen. As to what we are seeing as poor performance ... 10x as many roams/reassoc's even when standing still, 10-20% more TxRetries, 20% more missed Beacons and signal strength typically 1 "bar" lower (i.e. good instead of very good, poor instead of fair etc.) even when standing in the same place. Fingers crossed -025R sorts it. You'd think we'd have this stuff worked out by now!! IJ