We have a CCR running version 6.4 that suddenly just stop responding to all L3 traffic.
We cannot even access using mac address using winbox. Only can console in.
There is power to the CCR. After a reboot, everything goes back to normal.
Please help check and advise urgently as this is a production unit.
We had a CCR with 4.5 which suddenly dropped all Ethernetports
but luckily was accessible with the SFP port for remote
reboot. This CCR had an uptime of 4 weeks before this
happened.
2 other CCRs work for weeks now without a problem.
We’ve stopped deploying CCR for the moment until 6.x matures.
We want to replace some RB1000 but I guess they stay for another
6 months.
Edit: 6.5 not 4.5
We have fixed this issue in v6.7, please try that. It’s also written in the changelog
This?
fixed port flapping on CCR - could happen when having other than only-hardware-queue interface queue.
Note that having other interface queue than only-hardware-queue dramatically reduces performace, so should be avoided if possible;
This was no flapping. The ports were disconnected for 10minutes until we rebooted.
The queuing for this interfaces is “only-hardware-queue”.
The ethernet fix affects more than this, the changelog included the most popular problem. The optimization was more generic
We have now seen this happen on two CCR1036-12G-4S unit running 6.9 (will upgrade them when we get a chance).
In both cases: The units had been working fine for weeks, then all copper ports went down at the same time, while the SFPs continued to work. A reboot brought everything back in both cases.
If it happens again, I’ve told the guys to grab a supout before rebooting.
Any ideas?
–Eric