MikroTik

This is a new link using a RB333 ROS 3.3 with SR5

Going to a WRAP board with NMP8601 ROS 2.9.27

Tx/Rx Signal is -81/78dBm Rx/Tx CCQ is 69/80 % SNR 23 dB.

Seeing about 3% packet loss across the link with almost NO LOAD.

The WRAP has another conenection to an SR5 another WRAP ROS 2.9.27 which is similar in margin and quality with 0 % packet loss running sustained 6 Mbps

Am I missing something on the RB333 or is the link just not good ?

Thanks,
S

Spent 3 hours working on this link today. The RX/TX CCQ is 90%/75% and it still dropping packets. I don't think this is a link problem. It goes for hours at the same CCQ with 2Mbps load in both directions with no issue, then randomly drops packets. The CCQ remains pretty much the same the whole time.

Changing the Hardware retries incrementally from 0 to 15 makes NO difference.

We have other links that are RB532 + SR5 to WRAP + NMP links with lower SNR, and worse CCQ that do NOT drop packets.

This looks like a "special" feature of ROS 3.3 or the RB333

Anyone ???

This same RB333 has a SR2 on a 15 db VPol omni. Last night I installed two test clients on the SR2.

#One is a RB411 with Compex WLM54G in a 15 db Rootenna. Range is about 190 meters to the Omni antenna - it's about 110m from the base of the Tower, so it's basically aiming straight Up.

#Two is a WRAP with Compex WLM54G in a 19 db Rootenna. Range is about 1,500 meters to the Omni.

The RB411 is -60, CCQ is 90%/25% Avg. Ping time is 51ms to the AP IP !!!

The WRAP is -73, CCQ is 65%/75%. Avg Ping time is 17ms to the AP IP !

Both Clients have similar Packet loss symptoms as the Backhaul Link.

This is Smokeping to the RB333 5.8 Backhaul IP Notice that the link is clean for the past 1 1/2 hours, even though there have been NO CHANGES to it since Last night (12 hours ago)

This is Smokeping to the WRAP Eth0 interface that Links to the RB333 across the backhaul - the WRAP is on the near end of the link (closer to the NOC). So even though nothing has changed the Backhaul is once again running clean - for now. It's done this more than once. Eventually it starts dropping again. The Test CPE's are not clean - ping times are terrible and packet loss is 5% or more.

The only thing different about this AP than any other that we've done is the RB333, ROS 3.3 and we have inline Surge Suppressors on the AP End of the LMR. Anywhere else that we used Surge Suppressors they were installed on the Antenna End of the LMR.

This really doesn't look like a Link or alignment issue. What are the chances that it's related to the Surge Suppressors ? Pretty slim I think...

Comments PLEASE

Hi Ian;

I have often put the arrestor at the ap. When I showed this to Dan lately he said ...ya....but its more often at the antenna end..... He did not think it would be a
problem. If you need a second hand on the other end of a link sometimes, call me.

Paul