Error description:
We discovered serious connectivity problems when two (new/out of the box) CRS326-24S+2Q+RM are connected directly by copper and (new/out of the box) S-RJ01 modules.
We also recognized a rising FCS error count on the involved interfaces.
Setup:
switch0 (see config attached) is directly connected with switch1 (see config attached); both with S-RJ01 modules on the interfaces sfp-sfpplus2.
After a more or less short period of time, pings from switch0 to switch1 will timeout, then again be successful, then again timeout… so on (see .png attached).
config_switch1.rsc (758 Bytes)
FCS errors rises by time on the involved interfaces. Link Down count also rises, so does Rx Error Event count (see .png attached).
We further connected a ThinkPad to sfp-sfpplus24 (with S-RJ01) on switch0 and pinged switch1, same outcome (see .png attached).
Further information:
We recognized the errors in our productive environment on both redundant networks. In the one network we switched to S-RJ10 modules and the error disappeared.
In the other network we switched to LX modules and the error disappeared.
We have the RouterOS version 6.48.6 (Long-term) running on all involved devices. We reproduced the error on the table with new CRS326, new S-RJ01 modules, new copper cables.
Always the same outcome.
config_switch0.rsc (832 Bytes)