Hi!
We are an ISP with a lot of RB1100AHX4 and the same problem. There are a lot of rx-drop e the CPU load is low.
I tried update and even prey.
Anybody can help me?
Thanks,
Hi!
We are an ISP with a lot of RB1100AHX4 and the same problem. There are a lot of rx-drop e the CPU load is low.
I tried update and even prey.
Anybody can help me?
Thanks,
i can’t say if rx-drop it is necessarily a problem or failure or not i think
when searching about this there is no there is no concrete or definitive explanation
I think that in the statistical sense, the number of discarded packets vs. the total packages must be checked to establish a percentage ratio to establish their relevance
we cannot attribute any fault that is being presented simply to find a non-zero error counter
I have multiple routers with rx-drop counters and working correctly
for example in a interface 2 days up i have 224.808 rx-drop counter vs a total of 4.776.741.456 rx packets in that interface
that is 0.005 percent only
keep an eye on that rx-drop counter monitoring it to see if it increases in correlation with some other event, fail, condition or situation , but without it we don’t have arguments to infer there is a problem
the best and professional approach to solve technical problems is to be objective to identify problems avoid playing guessing games
Yes. But ignoring warning signals is a bad idea too. Customer behind this interface might see slower connections. So digging into RX-Drops is a good idea imho, MIght be packet bursts overload packet buffers of the internal switch or flow control problems.
Switch statistics:
interface ethernet switch print stats
Digging into a RB1100AHx4 I see tx-pause packets on switch-cpu interface using “interface ethernet switch port print stats”. So the internal switches do flow-control to the cpu. I see this with 200MBit traffic and <4% load. I guess the buffers of the internal switches are small so it might start pushing back packets on small bursts and start dropping very soon.
This router does not show problems. So it might be negligable but may give some slowdowns customer notices?
May be MT could share some insights on the packet flow and possible optimizations.
We are seeing the same problem with RX drops on a RB1100AHx4. Although the error rate is very low, I am concerned the error rate may go up with traffic. Previous RB1100 AHx2 didn’t have this problem. Anybody have an update on this?
Thanks,