load average

[Warren_9]

on a rp3410 machine, the load average always more than 5 but the %idle is 95%+..
on the other 3410 the load average is around 1 when 100% idle.

any idea?

[Ravi_8]

Hi,

which tool did you used to get this info?

Glance plus gives more clarity about the processes which are hogging more CPU

[T G Manikandan]

your load average is high with the idle CPU being 95%

check

the wio% using sar -u

Also check your disk io using iostat and vmstat

When the CPU is idle and the load average is up then there might be a high io in the system.

[Warren_9]

both top and w showed the same value,
and vmstat user sys and idle values are also same as top

sar -u:
wio% is 0

Cpu states:
LOAD USER NICE SYS IDLE BLOCK SWAIT INTR SSYS
4.51 4.2% 0.0% 1.2% 94.6% 0.0% 0.0% 0.0% 0.0%

.... strange!!

[Warren_9]

Since web application is running on the server, many httpd and java are running.

From the "top", there are 66 processes running, will this trigger the problem?

[T G Manikandan]

how about your iostat and vmstat!!

[Warren_9]

vmstat 2
procs memory page faults cpu
r b w avm free re at pi po fr de sr in sy cs us sy id
1 1 0 1640645 23964 1999 57 8 0 0 0 1 31 24873 916 40 17 43
1 1 0 1640645 24346 144 14 0 0 0 0 0 124 5412 562 0 0 100
1 1 0 1640645 24346 121 14 0 0 0 0 0 121 4905 541 4 4 91
1 1 0 1605508 24346 78 8 0 0 0 0 0 123 3764 541 0 0 100
1 1 0 1605508 24346 50 4 0 0 0 0 0 124 3024 534 3 0 97
# iostat 2

device bps sps msps

c2t0d0 0 0.0 1.0
c3t2d0 0 0.0 1.0
c12t0d0 0 0.0 1.0
c12t0d1 0 0.0 1.0

c2t0d0 38 10.6 1.0
c3t2d0 32 6.0 1.0
c12t0d0 1 0.0 1.0
c12t0d1 42 11.1 1.0

c2t0d0 27 7.9 1.0
c3t2d0 16 4.2 1.0
c12t0d0 1 0.0 1.0
c12t0d1 37 5.3 1.0

c12t0d0 1 0.0 1.0
c12t0d1 156 17.5 1.0

c2t0d0 14 3.5 1.0
c3t2d0 9 2.5 1.0
c12t0d0 1 0.0 1.0
c12t0d1 16 3.5 1.0

c2t0d0 3 0.5 1.0
c3t2d0 3 0.5 1.0
c12t0d0 1 0.0 1.0
c12t0d1 14 1.5 1.0

[T G Manikandan]

please provide the output of sar -d

[Warren_9]

sar -d

see attachment