I’ve got two point to point links on a Canopy BackHaul-20 equipment. Let’s designate them as “LINK1” and “LINK2”.
BH20(Master) <-----Bridging only 1 MAC, 5Mbit/s of traffic------> BH20(Slave) <—(Pinging from this side)
BH20(Slave) <------Bridging ~600 MACs, 5Mbit/s of traffic-------> BH20(Master) <----(Pinging from this side)
These apply to both of the links:
- The signal quality is great (1. -45dBm) (2. -49dBm)
- The jitter is 0~1 on the Master device (Sessions page) and 7~8 on the Slave
- The downlink data percentage is set to 50%
- 2X rate enabled (and working!)
- HW scheduling enabled
The problem is: If i ping the “other end” of LINK1 (that is bridging only 1 MAC) I got excellent results: 6~20ms even with 4.5Mbit/s of continuous traffic load… but, if i ping the “other end” of LINK2 (that is bridging approximately 600 MACs) the results are disappointing. The response time starts from 50-60ms and it can reach 150-250 milliseconds for a longer period of time (5-15 minutes).
What seems to be the problem here? The “Bridge Table” page states, that the Canopy equipment is capable of bridging more than 4000 MACs… Has anyone experienced a similar behavior on a Canopy-BH link?
Do you see the same lag when you ping something behind the BH bridging the 600 MAC? If you do not see this variance on the second radio, then it would not appear to be impacting throughput.
What are your frequency settings and is there a CMM in place?
I have the same question as wtkirk - are these results from pinging the radios or devices behind the radios? Have you tried doing the same ttests with hosts behind the radios? I have experienced very high ping return times on a number of Canopy radios on my network at various times, yet throughput has never been affected.
My theory - perhaps the processor in the radio has prioritized forwarding packets from hosts behind it instead of answering requests destined for itself. Would you rather the radio ignore your customers traffic to answer a ping from you?
the radios operate in the 5.7GHz range.
The first link is set to 5800MHz, the second is 5840MHz.
No CMM, but they’re connected with a sync cable…
Yes! The lag was noticeable on the next device, behind the radios, too…
I hope you’re right with your theory…
All right… I have to admit… I might be too fast on posting this question; I simply panicked when I saw those response times. :oops:
Anyway, thank you all, for your time and answers…