Alcatel-Lucent 5620 Troubleshooting Manual page 124

Release 12.0 R4 | August 2014 | 3HE 08864 AAAD TQZZA Edition 01
10 — Troubleshooting the 5620 SAM platform
iii
Perform the following steps to view a CPU activity summary.
8
i
ii
CPU minf mjf xcal intr ithr
0 10 0 442 302
1 1 0 220 237
2 0 0 27 192
3 1 0 160 255
Heading
CPU
minf
mjf
xcal
intr
ithr
csw
icsw
migr
smtx
srw
syscl
usr
sys
(1 of 2)
10-8
The top 5620 SAM process in the CPU column should be the Java process.
However, the Java process should not be consuming too much CPU. Some
Oracle processes may also consume CPU, depending on the database load.
Press CTRL-C to stop the command.
Enter the following command:
# mpstat time ↵
where time is the interval, in seconds, between CPU polls; a value between 10 and 60 is
recommended
Review the command output. Code
output; Table 10-3 describes each output field.
Code 10-2: Solaris mpstat output example
csw icsw migr smtx
419 166 12 196
100 383 161 41
100 178 94 38
100 566 202 28
Table 10-3 Solaris mpstat field descriptions
Description (events per second unless noted)
Processor identification
Minor faults
Major faults
Interprocessor cross-calls
Interrupts
Interrupts as threads (not counting clock interrupts)
Context switches
When the csw number slowly increases and the platform is not I/O bound,
a mutex contention is indicated
Involuntary context switches
When the iscw number increases beyond 500, the system is considered to
be under heavy load
Thread migrations to another processor
Spins on mutexes (lock not acquired on first try)
if the smtx number increases sharply, for instance from 30 to 300, a system
resource bottleneck is indicated
Spins on readers/writer locks (lock not acquired on first try)
System calls
Percent user time
Percent system time
10-2 is an example of Solaris mpstat
srw syscl usr sys
0 775 95 5
95 0 450 96
44 0 100 99
162 0 1286 87
Alcatel-Lucent 5620 Service Aware Manager
wt idl
0 0 95
4 0 0
1 0 0
8 0 5
5620 SAM
Troubleshooting Guide