Good afternoon everyone.
I’m from Brazil and I had some problems with a 4011iGS+.
I was using it in version 6.43.16 running PPPoE, CGNAT, SNMP, some VLANs and OSPF.
This equipment was connected to a CRS112-8G-4S switch in version 6.43.16 with all bridged ports connected to my BGP (CCR1036-12G-4S) in version 6.44.5 running only OSPF, eBGP and some VLANs.
Those equipments were connected by optic fiber just a few miles away, but the VLAN that communicated the RB-4011 with the BGP was losing a lot of packets.
When checking the interface there was no problem, performing OSPF communication directly on the interface caused all packet loss to cease.
When I downgraded from version 6.44.5 to 6.43.16 the losses ceased and there were no more problems.
Seems we’ve got about same issue.
Migrated from 6.43.10 > 6.44.5 > 6.45.3 with config saving. On CCR1036 router with some vlans and eBGP had about 90% packetloss, with zero errors on interfaces and it was depending from source - from 3 sources was 70-90% packet loss, from another one - zero packetloss, on 3rd day situation changed and we’ve got 90% packetloss everywhere. I’ve got 3 days of debugging this magic.
How it was cleared:
We’ve wiped out configuration and started it configuring from factory default settings step-by-step with as minimal as possible, i cannot say about 6.44.5, but now running on 6.45.3 - fine, old config was reproduced exactly same at was, with 1 change - i’ve did not renamed default port interface names leaved sfp-sfpplus1, sfp-sfpplus2 as it is and make a comment instead of renaming. Running about 3 days - without packet loss.
I’m not sure that problem was exact with port renaming, or some artifact was still in config, but need to move ccr to production in a few days and just do not have time to recheck again this issue.