Performance

 $ 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
    3     1     0  2814429  1073396   13    7    28   20     2    0    65   1388  10409   946  25  3 72
    3     1     0  2814429  1073319   12    2     0    0     0    0     0    539  12408   311   1  0 99
    3     1     0  2814429  1073319   12    0     9    0     0    0     0    533  12371   323   2  2 97
    3     1     0  2814429  1073319    7    0     5    0     0    0     0    539  15867   365   1  1 97

The fields for vmstat's display are

 procs   Number of kernel threads in each of the three following states
   r    run queue
   b    blocked kernel threads waiting for IO
   w    idle processes that have been swapped at some time

 Memory Report on usage of virtual and real Memory
   avm active virtual pages in 4KB
   free size of the free list in 4KB

 Page
   re  pages reclaimed from free list
   pi  pages paged in from filesystem or swap device to Memory
   po  pages paged out to filesystem or swap device from Memory
   fr  pages that have been freed or destroyed
   de  pages freed after writes
   sr  Pages scanned by clock algorithm per second

 CPU   Percentage usage of CPU
  us   user time
  sy   system time
  id   idle time

SWAPINFO: Prints information about device and file system paging space.

 swapinfo 
 -m  displays values in Mb instead of Kb
 -t add totals line
 -a show all device paging area

 swapinfo -tam
             Mb      Mb      Mb   PCT  START/      Mb
 TYPE      AVAIL    USED    FREE  USED   LIMIT RESERVE  PRI  NAME
 dev       20480   15578    4902   76%       0       -    1  /dev/vg00/lvol2
 reserve       -    4863   -4863
 memory    16349    6417    9932   39%
 total     36829   26858    9971   73%       -       0    -

top

To find out the process which are using more memory

 # top -f top.out -n 10000 -d 1
 # sort -n -k7,7 top.out 
 # sort -n -k7,7 top.out | awk '{if ($7~/M$/) {print $0}}'

glance

GlancePlus is a powerful and easy-to-use online performance diagnostic tool for HP-UX systems. It is distributed in two forms: "gpm", a Motif-based program and "glance", a character mode program.

The most use full sub command of glance when used on character based program are:

 a   CPU By Processor                                  
 c   CPU Report                                        
 d   Disk Report                                       
 g   Process List                                      
 i   IO By File System                                 
 l   Network By Interface                              
 m   Memory Report                                     
 n   NFS By System 
 t   System Tables Report
 u   IO By Disk                                        
 v   IO By Logical Volume                              
 w   Swap Space                                        
 U   IO By HBA Card                                    
 V   Logical System List                               
 j   Adjust intervel                                  
 ?   Commands Menu

sar
-L report activity for each active lunpath. If no lunpath is active, then a blank line is printed

# sar -L 2 2

HP-UX aembwqd1 B.11.31 U ia64    06/12/09

12:42:29                          lunpath   %busy   avque     r/s     w/s  blks/s  avwait  avserv
                                             %age     num     num     num     num    msec    msec
12:42:31               disk127_lunpath103    1.00    0.50       0       1       9    0.00    5.70
12:42:33                 disk101_lunpath3    0.50    0.50       0       3       6    0.00    6.28
                        disk101_lunpath53    0.50    0.50       0       3       6    0.00    7.67
                       disk127_lunpath103    2.50    0.50       0       4      56    0.00   33.09

Average                disk127_lunpath103    1.75    0.50       0       2      32    0.00   27.00
Average                  disk101_lunpath3    0.25    0.50       0       2       3    0.00    6.28
Average                 disk101_lunpath53    0.25    0.50       0       2       3    0.00    7.67 

Tuning

kctune: kctune is the administrative command for HP-UX kernel tunable parameters. It gives information about tunable parameters and their values, and makes changes to tunable values.

To see all tunables and their current values:

  $ kctune

To see which tunables have new values being held until next boot:

  $ kctune -D

To To see the tunables with only non default value

  $ kctune -S

To see verbose information about a tunable:

  $ kctune -v tunablename

To set a tunable value on the running system:

  $ kctune tunable=12

To set a tunable value to be used when the system reboots:

  $ kctune -h tunable=12

To increase a tunable's value by 100:

  $ kctune tunable+=100