Hi everyone - we seem to be having continual issues with our 1072 routers we are using and im looking for any advice to stabilise fix some of these issues.

Lately aside from frequent watchdog reboots, high cpu spikes we are seeing this weird issue on two of our CCR1072's so far:
An event occurs after no change or obvious trigger, very minimal traffic flow.
sfp+2 interface goes down with module plugged in. Reboot and firmware change no fix. Unplug / replug sfp all OK. Other end of link showed link was up whole way through - no traffic was passing.
sfp+1 interface was receiving traffic but no TX traffic was evident over link. Interface could ping local address either end but no useful transfer of traffic was occurring. Tried rebooting, upgrading etc only thing that fixed this was to change the ip address on each end of sfp1+ link to a new range and everything worked straight away. Checked for IP conflict nothing evident. Looks like an interface driver / ARP issue.

Sfp+ 3,4 were working fine - being able to access and check things
Sfp+ in router while this was occurring.

This issue of one way traffic and having to change IP has now occurred on two routers in completely seperate parts of our network.

Once comment to note is that with watchdog off when the router crashes we are unable to access via CLI and sfp+1 and 2 lights are on, from memory 1 is green and 2 is yellow even with no router response or modules plugged in.

We are only running low traffic levels, simple ospf and bgp. We have tried all sorts of firmwares - seems 6.39 rc54 has been the most stable for reboots so far? CPU speed, traffic mon, FW version all checked as per forum suggestions.

Thanks for any advice to settle these things down.

For the watchdog reboots, please make sure that you are running your router at 1000Mhz and not 1200Mhz.

We had a lot of random watchdog reboots when the cpu was set to 1200Mhz