Log in to ask questions, share your expertise, or stay connected to content you value. Don’t have a login? Learn how to become a member.
I receive a lot of alarms for CPU usage above 70% .
I have this issue on two stacked EX 2300 switches:
Model: ex2300-48pJunos: 20.4R3-S2.6Model: ex2300-48tJunos: 20.4R3-S2.6
Could you please share your experience or guidance regarding the topic?
How can I identify what causes the issue?
I tried to check the PR numbers and release notes for issues but I didn't find anything promising for Junos: 20.4R3-S2.6 and EX 2300.
I ran into a similar issue on a stack of 4400's. Eventually it brought the link down and we had to manually reboot the stack and clear an err'ed port on the upstream switch.I was told to replace optics on the uplinks, replace cabling, and check for excessive port flapping.I also found the upstream side of the connection had a high number of CRC errors. I replaced the fiber jumpers in between and it seems to have calmed down.What are you running upstream? do you have a LAG running between upstream/downstream switches? Can you run the following on your ae links?show interfaces xe-#/#/# extensiveSee if you have a high number of carrier transitions in the output. Also check on the upstream and downstream switches for high number of input/output errors on the connected trunk ports.
Are you by any chance using JunoSpace or another management tool? I've found they tend to do large amounts of bulk SNMP queries which sends the CPU usage sky high at times.