In the network we use Cisco SG350X-24-K9 V02 switches (FW: 2.5.0.90). After 495 days they rebooted spontaneously, or their uptime was reset, but there is no mention of the reboot in the log (and also no interruption of data). After about 3 weeks, the first of the switches started showing 100% CPU utilization. The change occurred in leaps and bounds from about 3-5% to 100%. I read data from the switches via SNMP, but even disabling this service did not reduce the CPU utilization. The switch has 6 active ports. Data flow is about 4Mbps. No routing. After a week or so, the same situation repeated itself on next switch. These are in a stack of 4 and again the processor has been running at 100% since then. I would like to find a way to find out what the CPU is utilizing and could prevent these conditions.
PS: The great thing about these SG350X switches is that even with 100% CPU utilization, the CLI and web interface are still available. With the SG500, we could also get the CPU to 100% utilization, but in addition, it was no longer possible to connect via telnet or the web, and when the device was disconnected from the port, the device would not reconnect.
