 |
» |
|
|
|
|
|
|
Examples
- The following example shows how to run a
full data collection and display the output
at the terminal using the standard interval
of 10 seconds:
# collect
This command is similar to output monitoring
commands such as vmstat, iostat,
or netstat and the command volstat.
-
The following command uses the -s option
to collect only process information in the
file sys.data. The -S option
specifies that the data is sorted by CPU
usage, and the -n option specifies
that the top ten processes are saved:
# collect -sp -S -n10 -f sys.data
Initializing (10.0 seconds)
The message "Initializing (10.0 seconds)" indicates
that data collection will be performed at
10-second intervals.
- The following command displays the data
collected in the preceding example by piping
the output to the
more command:
# collect
-p sys.data.cgz ¦ more
| ###################################################################### |
| OSF1 glop.ytx.tog.comT5.0
77.11 |
DEC1000 1/266MHz/256
MB |
| HOST............glop.ytx.tog.cm |
Started......<DY:MM:DT:HH:MM:SS:YR> |
| |
Seconds........943298217 |
| CPU FAMILY......21064
(EV4 core) |
CPU ID.........EV4.5
(21064) |
| CPU EXTENSIONS.. |
| PLATFORM NAME...DEC1000 |
CPU SPEED.......266
MHz |
| SWAP SIZE.......196MB |
Physical Mem...256MB |
| NUM CPUS........1 |
NUM DISKS......3 |
| NUM LANS........3 |
NUMFSYS.......4 |
| MAX MQUEUES.....64 |
NUM TAPES......0 |
| INTERVAL........10.00 |
PROC_INTERVAL..10.00 |
| UBCMAXPERCENT...100 |
UBCMINPERCENT..10 |
| MAXUSERS........256 |
MAXUPRC........64 |
| Delay_WBuffers..0 |
LSM Volumes....0 |
| ###################################################################### |
| #### RECORD1 (943298227:10)
(Mon Nov 22 14:17:07 1999) #### |
| # Process Statistics
(RSS & VSZ in KBytes) |
| # PID |
User |
%CPU |
RSS |
VSZ |
UsrTim |
SysTim |
IBk |
OBk |
Maj |
Min |
Command |
| 0 |
root |
1.7 |
12M |
342M |
0.00 |
0.00 |
0 |
0 |
0 |
0 |
kernel idle |
| 3275 |
root |
0.3 |
3.3M |
5.6M |
0.00 |
0.00 |
0 |
0 |
0 |
8 |
collect |
| 482 |
root |
0.0 |
2.6M |
6.3M |
0.00 |
0.00 |
0 |
0 |
0 |
0 |
insightd |
| 360 |
root |
0.00 |
2.0M |
4.4M |
0.00 |
0.00 |
0 |
0 |
0 |
0 |
automount |
| . |
| . |
| . |
Note that this sample output, and others on
this page, have been modified and compressed
to fit the page. Normal screen or printer output
will be wider.
- The following comand uses the
-e option
to exclude filesystem data and collects data
every second, except for process data, which
is collected every 5 seconds. The times are
set using the -i option.
# collect
-ef -i1,5 -f sys.data
Initializing (1.0 seconds) ... done.
Note that the time
has changed in the
initialization message.
- The following command prints only the header
section of a data file. That is, the information
bordered by the hash, or pound, symbol (#),
as shown in the sample output in Example 3:
# collect -sh -p sys.data
- The following command selects only the data
from the network subsystem and displays it
at the command prompt:
#collect -sn |
Initializing (10.0
seconds) ... done. |
#### RECORD 1 (943045470:0)
(Fri Nov 19 16:04:30 1999) #### |
# Network Statistics |
#Cnt |
Name |
Inpck |
InErr |
Outpck |
OutErr |
Coll |
IKB |
OKB |
%BW |
0 |
lo0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
1 |
sl0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
2 |
tu0 |
75 |
0 |
0 |
0 |
0 |
8 |
0 |
0 |
- The following command specifies only data
from the disk subsystem, and then only from
specific disks identified as
dsk0, dsk1,
and dsk8. The disk names are
determined by their device special file names
in the /dev/disk directory.
# collect
-sd -Ddsk0,dsk1,dsk8
Initializing (1.0 seconds) ... done.
The hwmgr command
can also be used
to find devices.
See the hwmgr(8)
reference page for
information on the
command options.
- Use the
-p option to convert
data files created using a previous version
of the Collect utility as follows:
collect -p /tmp/olddata.col -f
/tmp/oldconverted.col
Initializing (1.0 seconds) ... done.
More Collect product
details
|
|
|
|
|
|
|
|
|
