with 6.3 we expect some problem of duplex on ethernet on plateforme 2011 and 1100AHx2.
if the duplex is manualy setup. the MKT not go to full duplex. and go to half duplex and the bad thinks with it...

back to 6.2 resolve this issue.

you are talking about ethernet interfaces? At what rate you have them?

We're seeing the same problems: in rb1100ahx2 some ethernets simply won't go to full duplex (even if auto is disabled).

recycling the interface(disa>ena) won't help.

We're seeing the same problems: in rb1100ahx2 some ethernets simply won't go to full duplex (even if auto is disabled).

recycling the interface(disa>ena) won't help.

We are seeing the same on RB1100ahx2. Port 11/12 seem to be fine but ports on the switch chips are unusable.

if you want to run interface as 1GBps - you have to have auto-negotiation enabled. There are other parameters that are negotiated for GBit link, so there is no surprise that interface is not working at full speed with that disabled.

I have the problem in 10Mbits & 100Mbits...

good option is to check other cables and/or introduce some switch between the devices, then you can check where is the problem. Also, good idea is to ground device properly, so that port is not damaged due to static discharge.

This specific ethernet is in a 15cm stack shielded cable to the fiber company router.
No way it's the snr in the cables, or EMI, cause it's fit in a #14 steel outdoor rack. The ethernet is locked ad 100FDX, but gets stuck at 100HDX in 6.3 firmware(HOW? both devices are locked at 100FDX!!).
Solution: went back to 6.2 and everything is working properly.

Ohh maan... Does it happen only with 1100AHx2? Did anyone tested with CCR1036? I have dynamic invalid interface issue on 6.2 with CCR and planned to upgrade to 6.3 this week. Ohhh... Too bad

This specific ethernet is in a 15cm stack shielded cable to the fiber company router.
No way it's the snr in the cables, or EMI, cause it's fit in a #14 steel outdoor rack. The ethernet is locked ad 100FDX, but gets stuck at 100HDX in 6.3 firmware(HOW? both devices are locked at 100FDX!!).
Solution: went back to 6.2 and everything is working properly.

Although officially not required, some manufacturers recommend a minimum cable length of 1 meter.. You can try using a longer cable.

Why searching about the cable tested OK with Flupke.
was running ok from 5.24, 5.25, 6.0, 6.1, 6.2 and now with 6.3 the problem appear at this time.
I reverese to an old version and the problem disappear...

The problem is clear...and reproducable...

now check 6.4 =)

Chupaka,

thanks

6.4 problem solved
now it is ok..

nothing about that in the release note

6.4 problem solved ... now it is ok ...

nothing about that in the release note

I think I'm gonna roll the dice ... v6.4 treat me right!

Not sure if it's the same issue, but I'm experiencing an auto negotiation problem between a CCR and Cisco 2960S, upon boot up. Sometimes it negotiates, and sometimes there's no traffic. At the moment I have the Cisco side set for 1000 full, which functions.