RX Drops

[DesertDweller]

I have a DC-26 that runs fine for a random amount of time, then starts spiking latency up as high as 500-2000ms. When this happens the port details show the Rx Drop counter going up. The only thing that fixes it is a reboot. I am running the latest firmware and it is connected to a Mikrotik RB4011. We have some port based VLANS, but it is an otherwise very basic configuration.

What could be happening here?

David

[DesertDweller]

I think most of the drops were coming from VLAN1 not being untagged on the port, but this still doesn't explain why ping speeds spike the way they do.

[DesertDweller]

To clarify, latency to the switch is fine, but anything connected to it spikes.

[Stephen]

Is flow control enabled? If it is, try disabling it.

[DesertDweller]

Flow control is not enabled.

[Stephen]

Can you post your config tab, status tab, ports tab, and vlans tab? Flow control is where we always start with something like this, but if it goes past that then we need more info. Also, if you could get the logs for about the time the latency start to occur it would also be helpful.

[AsiNetwork]

we got the same problem, is not netonix switch they are splendit fast.
We tried changing switch, cables, sfp modules.

With around 5 tower the common thing was the mikrotik RB4011, i looks that is has some kind of bug software I have seen the issue on 6.44+ software but the firsts ones that we bought had 6.43 dont have this issue.
none packet loss, ping to Directed connected devices promt 0 ms. with netonix,
We changed the routers to CCR1009 and problem gone.

Try it

[Stephen]

Thanks for the post AsiNetwork, I will circulate this information around to anyone else who might need it.