SRX

i see via web ui high cpu utilization , then compare with cli there is 286.38% flowd_octeon_hm.

what is this(286.38% flowd_octeon_hm)?

what i have to do?

netadmin@SRX-> show system processes extensive      
last pid: 46682;  load averages:  5.91,  5.77,  5.90  up 4+16:56:45    16:16:28
140 processes: 26 running, 103 sleeping, 11 waiting

Mem: 246M Active, 38M Inact, 564M Wired, 61M Cache, 112M Buf, 61M Free
Swap:


  PID USERNAME  THR PRI NICE   SIZE    RES STATE  C   TIME   WCPU COMMAND
 1099 root        7  76    0   491M 48276K RUN    0 355.0H 286.38% flowd_octeon_hm
 1131 nobody      6  76    0 11972K  6012K ucond  0  22:24 43.46% httpd
46681 root        1 122    0  9260K  3888K RUN    0   0:01  6.74% httpd
   22 root        1 171   52     0K    16K RUN    0  89.3H  0.00% idle: cpu0
   24 root        1 -20 -139     0K    16K RUN    0  84:21  0.00% swi7: clock
 1125 root        1  76    0  4312K  1792K select 0  16:26  0.00% license-check
 1127 root        1   4    0 16256K  9012K kqread 0  14:22  0.00% eswd
    5 root        1 -84    0     0K    16K rtfifo 0  14:01  0.00% rtfifo_kern_recv
   23 root        1 -40 -159     0K    16K WAIT   0  11:29  0.00% swi2: net
 1135 root        1  76    0     0K    16K select 0   9:34  0.00% peer proxy
 1124 root        1   4    0  7076K  2828K kqread 0   9:33  0.00% mcsnoopd
   20 root        1 171   52     0K    16K RUN    2   6:59  0.00% idle: cpu2
   19 root        1 171   52     0K    16K RUN    3   6:49  0.00% idle: cpu3
 1095 root        1  76    0  7892K  2540K select 0   6:22  0.00% ppmd
 1120 root        1  76    0 10028K  4428K select 0   6:17  0.00% rtlogd

Hi,

take a look at the post from rkim http://forums.juniper.net/t5/SRX-Services-Gateway/2-cluster-SRX240H-CPU-flowd-octeon-hm/td-p/54692 . He describes a more accurate way to see the cpu usage and why flowd_octeon_hm always shows such a high load.

Kind regards,

Sebastian

The SRX uses a multicore processor.  The flowd_octeon thread usually runs at high percentage in order to make sure that all the packets received are processed faster.  This is why flowd reads high always, however it does not mean that system is busy.

The actual CPU usage can be viewed below.

For Routing Engine (Control Plane) CPU usage:

  show chassis routing-engine

For Packet Forwarding Engine (Data Plane) CPU usage:

  show chassis forwarding

  show security monitoring fpc 0

Hi Hendry,

SRX has 3 CPU cores, since freebsd resource allocation mechanism ensures pre-allocation of cpu resources to important processes like flowed in order to provide near-realtime processing. flowd usually roams around 280%-290% due to preallocation of cpus. Use top -H to view in detail.

regards

Hi,

43% cpu utilizaed by httpd is not a usual behavior. What is the junos version you are running? you can restart the web-mangement and check the httpd usage.