Hello,
I connected a device we use as a motor control to a Dell PowerConnect 7048 and it started to report to our Check_MK server some in-errors (usually around 4%) during network activities (no errors while the machine is idle), the device is a black box to me because I don't know what it does and how and it have no logs or reports.
Unfortunatly the device went in an unreachable state during a night shift and it was rebooted by the operators, I couldn't do any test on it during the problem so I actually have no idea about what happend.
Now they're asking me if it is possibile that the switch itself bring down the connection because of the reported errors. I know it could seem unlikely but the situation is a bit messy... There are 6 to 8 of these device, all in production (this one included) by a long time and no problem like this was ever reported, some of the devices are connected to unmanaged switch, the others to some old HP ProCurve switches (all monitored via SNMP by the same Check_MK server) and none of them reporterd any in-errors or such, this one was connected to an HP switch before we move it to the 7048 so everything seems to point to the switch as a possibile cause of the problem.
I've done some tests with a spare switch and a spare device and I was able to replicate the in-errors problem just with a plain nmap to the device (but with 1-2% error rate) but even after hours of "stress-test" the device was never unreachable.
What I'd like to know is:
1) it is possibile that the switch bring down the interface because of the errors? In case it could be shouldn't be notified somehow? (the switch's log showed nothing but the link-down/link-up event of the reboot)
2) is there any way I could try to "debug" the errors from inside the switch or my best shot at is to use wireshark on a mirrored port?
Thanks
