AP's Functioning very well now with two provisio

I have finally got ALL my APs functioning well after globally making two very thorough changes - 1) All - and I mean ALL - CPE MUST have RTS/CTS off - it only takes one on an busy AP to take it down - and it (CPE) must have the newest firmware available loaded!
2) Not ONE of my APs will work with the NV2 wireless package loaded - which blows out all combo boards in the network * all edge and secondary towers - and only in v4.16 do I have a higher performance then v3.30. But it is definitely higher performance - technically speaking it kicks ass! AP works better and I truly believe that WDS Backhauls are faster - even using 5mhz bands. And - I’m going out on a limb here - ethernet ports seem to handle packets better.

Having said all this and - over the last several months - totally pissing off my customer base trying to make NV2 and AP work togather. NV2 by itself on a board only doing Backhauls is pretty damn awesome - both in v4.16 and RC7 (haven’t tried RC8 - still have PTSD from the last action). Not so sensitive to noise and very quick! I even have several boards doing both NV2 and WDS(AP BH) and they work great!
Good luck All!
Rod

what exactly happening on the APs when the clients uses RTS/CTS? And AP is running regular wireless package or Nv2 wireless package?

Hey Uldis,

With the NV2 the AP’s will not function at all almost immediately if there is a CPE with RTS/CTS active.
With v4.16 it will show up 15 - 60 minutes later. I’m assuming it has to do with the processing power of the individual board that determines how quickly it looses control. Timing? Latency goes to hell - 3000+ms and then complete time outs.

Once all the units have that protocol deactivated then the whole works really good!
RC7 does not work well at all for 802.11 with a NV2 connection functioning! Have not tried RC8 yet and probably won’t until full release.
RC7 NV2 operating as a BH bridge with multiple P-to-MP connections seems to work really good.
Thank you,

Rod

RTS/CTS as designed should only be used when EVERYTHING has it configured. By definition if only one client has it set up, it’s going to wreck the sector. This is to be expected.