04-25-2014 06:32 AM
I have a high cpu usage when SNMP is running. If I kill the SNMP process all is working properlly.
Does anybody give me any clue about what will be wrong with this BROCADE 300?
Fabric OS: v6.4.1a
top - 16:27:03 up 980 days, 6:11, 1 user, load average: 0.29, 0.15, 0.13
Tasks: 80 total, 1 running, 79 sleeping, 0 stopped, 0 zombie
Cpu(s): 64.0%us, 25.4%sy, 0.0%ni, 9.9%id, 0.0%wa, 0.0%hi, 0.7%si, 0.0%st
Mem: 504344k total, 299708k used, 204636k free, 21184k buffers
Swap: 0k total, 0k used, 0k free, 124004k cached
09-16-2014 10:23 AM - edited 09-16-2014 10:26 AM
i am also having this problem on three VDX 6720 switches running 4.1.2a. i currently have a support case open to see if there is a bug or what is causing this process to hang like that...
Switch# show process cpu
Load Average: One minute: 2.33; Five minutes: 2.34; Fifteen minutes: 2.36
Total CPU Utilization:
Current: 100.00% (user procs:25.00%, system-kernel:75.00%, iowait:0.00%)
Active Processes Lifetime Statistic:
Legend(State): D - uninterruptible sleep R - running S - interruptible sleep
T - stopped W - paging X - dead Z - defunct zombie
PID Process CPU% State Started
3473 snmpd 72.80 S 03:30:58 Jul 13, 2014
5600 sshd 1.30 S 11:25:11 Sep 16, 2014
4752 EAN_RX_43008080 1.20 S 03:32:25 Jul 13, 2014
3551 HSLK_TH_PKT 0.60 S 03:31:08 Jul 13, 2014
4750 EAN_TM_43008080 0.50 S 03:32:25 Jul 13, 2014
09-16-2014 01:56 PM
i just wanted to post this so others know both what causes the issue and how to kill the proccess. There is a known bug in the current version 4.1.2a that if UDP ports 8000 and 50000 are scanned with a tool such as nmap it can lock up the snmp process. this goes for any ip address / layer 3 address on the device. to kill the process simply log in as root using ssh, telnet, or console cable and run "kill -9 <pid>" to kill the process and free up the CPU. hope this helps others out there.