I’ve got a (client/AP) link about 10km away, running on RB133C with a compex CX-WLM54AG card. AP has RB333 with R52H. On first install I used 2.9.46 and things were ok, then I thought to upgrade client to 3.0rc10, after which the link was VERY bad, about nly every 5th ping would get a response in the range of 100-400ms. Then I downgraded to 2.9.49 and pings were back solid between 2 and 20ms.
The client card was running at full power all the time (20dbm) so in my opinion there is something not 100% in the 3.0rc10 software.
Anyone had similar experience or know of an issue? Would the same issue affect my APs that is running on RB333 with 3.0rc10 software, giving not best performance? I’d just like to know when would be good to upgrade all my radios to 3.0…
I have tried 3.0 on many 133c boards and have switched back to 2.9 on every single on of them. It just doesn’t work well on them. CPU usage is always high and performance is aweful. I won’t be even trying it again until 3.0 is finished and the reports of it being so unstable are gone.
Check CPU load. It seems to spike when using winbox. web administration doesn’t hit as hard, and cpu usage doesn’t appear to be as bad when passing traffic as long as you are not actively monitoring the system. I’ve made a backup of the config, reset the rb, then restored the config and it worked much better after that. Dunno what the difference is.
this may be a dumb question, but have you disabled any unnecessary packages running on the board? i’ve setup dozens of rc10’s on the 133c with no problems, AFTER disabling everything but DHCP, routerboard, system, and wireless. it’s still a bit sluggish running winbox, but pings and throughput are good.
I had the same config, except I also needed ppp enabled for my pppoe authentication. Maybe that was just enough to put it over the edge, but if it is that close, I don’t want to bother. I wish version 3 weren’t so cpu intensive.
I put 3.0RC10 on a RB112 I use for my own Internet acces, so far, no problem, but I’m not doing anything fancy (just routing + NAT/DHCP). The CPU usage in resource monitor spikes to 100% every few seconds, but throughput / pings don’t seem affected. 3.0 still seems like it has a long way to go before it’s ready for the field though, I’d be very hesitant to put it anywhere a climb is required.
I’m doing PtmP using RB333 + rc10 as base station , RB133c + rc10 as clients , distances between 2Km to 11Km , clients are station-wds bridge , no NAT no DHCP no firewall , ping is normal avg=2 ms ..
same issues here! got two 133c working on field; also when i use RC the performance is awful ! 512kbps internet is working like dial-up! even the signal is btwn -60 -70 ; when i downgrade to 2.9 everything is fine!