Ingres 11.2 | UNIX Operating System Utilities

System Administrator Guide > System Administrator Guide > Using Monitoring and Tracing Tools > Operating System Utilities > UNIX Operating System Utilities

Was this helpful?

UNIX Operating System Utilities

You can use the following UNIX operating system utilities to monitor Ingres:

• ps

• iostat

• vmstat

For a full description of the UNIX options, see your UNIX documentation (or online help). A detailed description of the utilities can be found in the Command Reference Guide.

Note: Not all of the utilities are present on every UNIX system. Some are present only in a BSD or System V environment but not both.

This command provides virtual memory and cpu information on each active process submitted from your account. Here is a sample ps output:

PID   TT  STAT  TIME  SL  RE  PAGEIN  SIZE   RSS   LIM   %CPU   %MEM   COMMAND
xx06  p3   s   28:50  13  99   45886  3696   2856   xx    0.0   39.3   iidbms
xx94  p3   s    4:24   0  99    2899   720    344   xx    0.0    4.7   dmfrcp
xx09  p3   I    0.51  99  99    4488     4    184   xx    0.0    2.5   iislave
xx19  p3   I    0.57  99  99    5764    64     17   xx    0.0    2.4   iislave
xx96  p3   I    0.04  99  99    1852   696    160   xx    0.0    2.2   dmfacp

The display fields are as follows:

Field	Description
PID	Process ID field
TT	Controlling terminal
STAT	Process status Runnability of the process: Runnable (r), Stopped (t), Disk or other short-term wait (d), Sleeping (s), or Idle (i) Swap status: Swapped out (w) or Loaded in core (blank) Process priority change: Reduced (n), Increased (>) or No change (blank)
SL	Sleep time (seconds blocked)
RE	Residency time (seconds in core)
PAGEIN	Number of disk I/Os resulting from page references not in core
SIZE	Virtual process size
RSS	Resident set size
LIM	Soft memory limit (setrlimit), else “xx”
%CPU	CPU utilization (1 minute decaying average)
%MEM	Memory utilization
COMMAND	Process name

iostat

The iostat command returns information about I/O status. It lists statistics on current I/O activity for each disk device and system CPU utilization percentages. Here is a sample iostat output:

tty cpu
tin tout us ni sy id
1 18 19 0 3 78

/dev/*dsk/c0d*s*/dev/*dsk/cld*s*/dev/*dsk/c2d*s*/dev/*dsk/c4d*s*
bps sps msps bps sps msps bps sps msps bps sps msps
2 0.1 61.7 1 0.0 95.5 1 0.0 60.4 2 0.2 44.3

• The tin and tout display fields show the number of characters written to and from terminal devices.

• CPU information includes the % time spent in user mode (us), “niced” user mode (ni), system mode (sy), and idle (id).

• The disk I/O for each disk device shows the average number of blocks transferred per second (bps), average number of seeks per second (sps), and average time per seek in ms (msps).

vmstat

The vmstat (Virtual Memory Statistics) command returns virtual memory status information, including process states and paging activity. Here is a sample vmstat output:

procs memory page disk faults cpu
r b w avm fre di re rd pi po de z0 z1 z2 z3 in sy cs us sy id
1 0 0 2536 456 24 2 1 4 0 0 1 1 0 1 24 475 23 4 6 91
0 0 0 2748 356 24 0 0 0 0 0 2 0 0 3 26 323 29 1 5 95
0 0 0 1044 344 24 0 0 0 0 0 0 0 0 0 16 216 18 0 3 97
0 0 0 2288 344 24 0 0 0 0 0 0 0 0 0 19 334 27 1 4 95
1 0 0 2372 332 24 0 0 4 0 0 0 0 0 1 28 552 40 1 6 93

The procs columns define the process states: in run queue (r), blocked for resources (b), and runnable or short sleeper (w).

The memory columns show virtual and real memory status: avm is active virtual pages (belonging to processes active in approximately the last 20 seconds), fre is size of the free list, and di is the number of dirty pages.

The page columns show page faults and paging activity. These are expressed in units per second, averaged over 5 seconds: re are page reclaims, rd are page reclaims from the dirty list, and pi, po are pages paged in/out. The de field is anticipated short-term memory shortfall.

The disk columns list disk activity, showing the device name and operations per second.

Last modified date: 08/28/2024