We are seeing an issue with RX Drops between (2) x86 systems. We have 2 of these routers connected via a direct cat5 cable (less than 1 meter) negotiated at 1GB FDX port speeds.

The RX Drop counters of the interfaces connecting the (2) routers are increasing on both routers. There are no TX or RX errors, only Drops. Anyone have ideas as to what would cause RX Drops on these 2 routers?

Try to disable simple queues if you have any, and try again.

I use the same Axiomtech systems and having RX drops as well.

Sometimes disabling simple queues help.

We are seeing an issue with RX Drops between (2) x86 systems. We have 2 of these routers connected via a direct cat5 cable (less than 1 meter) negotiated at 1GB FDX port speeds.

The RX Drop counters of the interfaces connecting the (2) routers are increasing on both routers. There are no TX or RX errors, only Drops. Anyone have ideas as to what would cause RX Drops on these 2 routers?

I am having the same issue.
2 x86 systems,
CPU load is low ( i have disabled all Mangle rules to test)
there are no Queues.

one x86 is our BGP router, connecting to another x86 thats normally our QOS (currently just has IP addresses as i have disabled everything in trial to fix issue) and the third is our x86 LNS.

I am not noticing any RX drops on the interfaces that connect to our Cisco Switch. Only those that connect between x86 routers.
there are 0 errors

the drops don't 'count up slowly' just randomly every few hours i get a 'total drop of all traffic' and get large numbers added to my RX drops on that interface, meanwhile its basically a total network outage for 3-4 seconds.

BGP Boz


QoS Box

We are having a vary similar problem. We upgraded from RoS 5.11 to 5.14 and on 2 particular VLANS we are getting TX Drops (not RX). We are running PPC package on RB1100 but we tried switching all the configuration to an RB1000 but we are seeing the same problem. Like you are saying, the count just bursts suddenly, then it stops.

Anyone else seeing the same affect?

we are still getting the issue. Thus to stop it we put a Cisco switch between the routers, fixed the issue. (messy and un-necessary use of extra equipment though)

try to use auto-neg on on the interface.

to many outages to try now;

but when we did first setup the units it was on auto-neg, i turned it off to see if it resolved the porblem (it didn't)

Hint: RouterOS v6.x has an updated versions of x86 Ethernet drivers

"Version 6.x is a very early development version,
please use only in testing environment and report any problems "

Unfortunately our Core and Border and LNS routers are a little too important to put 6.x just yet