 Home
History
Products
VIP
|
|
MONITOR WORKING_SET
Monitors processes in an OpenVMS cluster environment for their working
set values and associated values.
Format:
$ VIP MONITOR WORKING_SET
SHORTCUTS:
You can use the following short cut keys:
Ctrl/F (Force inswap) Forces the processes at the current page to
swap-in in combination with a monitor scan
and remains at the current page.
Ctrl/I (Interval) Forces a monitor scan and shows the next page.
Ctrl/N (New values) Forces a monitor scan and remains at the
current page.
Ctrl/R (Refresh) Forces a refresh and shows the next page.
QUALIFIERS:
/ACCOUNT
/ACCOUNT=(account-name[,...])
Selects only those processes for monitoring that have the same
account name as one of the specified account names. Wildcards are
allowed.
/ALARM
/ALARM=(alarm-keyword[,...])
Allows you to set an alarm on a condition specified in the /ITEM
qualifier. Choose one or a combination of the following alarm
keywords:
EVERY=m Specifies after how many interval
periods an alarm message will be
generated on the item specified with
the /ITEM qualifier. The default value
is 1.
IDENTIFICATION=(pid[,...]) Specifies a list of pid values of
which the associated processes receive
the error messages on their terminal.
LOG=file-name Allows you to write the alarm messages
into a log file preceeded by a time
stamp.
MAIL_ADDRESS=(addr[,...]) Specifies a list of mail addresses
(usernames) to receive the error
messages in an OpenVMS mail message.
OPERATOR Allows you to send an alarm message to
the operator console terminal.
POSITION=name Allows you to place the alarm message
at one of the following positions:
BOTTOM
All alarm messages will be placed
at the bottom line of the terminal
device. Each new alarm message will
overwrite the previous message.
BOTTOM_UP
All alarm messages generated in the
same interval period will be places
starting at the bottom line of the
terminal device. Each new alarm
message will be placed at the line
above the previous message.
NONE (Default)
All alarm messages will be placed
at the current cursor position.
TOP
All alarm messages will be placed
at the top line of the terminal
device. Each new alarm message will
overwrite the previous message.
TOP_DOWN
All alarm messages generated in the
same interval period will be placed
starting at the top line of the
terminal device. Each new alarm
message will be placed at the line
below the previous message.
PROCESS_NAME=(name[,...]) Specifies a list of process names to
to receive the error messages on their
terminal.
REPEAT=m Specifies the repeat count of the alarm
messages. Dismounting the disk device
stops the sending of alarm messages,
even if the repeat count has not been
reached.
TERMINAL=(name[,...]) Specifies a list of terminal devices to
receive the error messages.
USERNAME=(name[,...]) Specifies a list of user names to
receive the error messages on their
terminal.
/BATCH
Allows you to select the batch processes in the system. When used
with the /CLUSTER qualifier, the /BATCH qualifier selects all batch
processes in an OpenVMS cluster environment. When used with the
/NODENAME qualifier, the /BATCH qualifier selects all batch processes
of the specified nodes.
Using this qualifier gives the same result as using the qualifier
/MODE=BATCH.
/BEGINNING
/BEGINNING=time
Allows you to specify the time when the actual monitoring starts. You
can specify an absolute time, a delta time or one of the relative date
strings YESTERDAY, TODAY or TOMORROW.
/CLUSTER
Allows you to select processes on all nodes in an OpenVMS cluster
environment for display. This qualifier can not be used when the
/NODENAME qualifier is also specified.
/ENDING
/ENDING=time
Allows you to specify the time the monitoring is to end. You can
specify an absolute time, a delta time or the relative date string
TOMORROW. When the ending time is a delta time and the beginning time
is also specified, the actual ending time is the sum of the specified
beginning and ending time. The ending time must be in the future.
/EXCLUDE
/EXCLUDE[=(process-name[,...])]
Allows you to exclude processes for monitoring according to the
process name. Wildcards are allowed. If you enter the /EXCLUDE
qualifier without a value, the current process will be excluded.
/FORCE_INSWAP
/FORCE_INSWAP[=item-value]
Controls whether outswapped processes are forced to swap-in in order
to retrieve process information. Please use with care, because it may
have impact on all processes running on the system. Choose one of the
following item values:
STARTUP When specified processes are forced to swap-in only at
the start of this monitor command before information is
retrieved to select processes for monitoring. This is the
default value.
REFRESH When specified processes are forced to swap-in when a
refresh is performed, whether by the use of the shortcut
key (Ctrl/R) or automatically.
INTERVAL When specified the processes on the next page are forced
to swap-in at every monitor interval just before the
process information is being displayed.
/IDENTIFICATION
/IDENTIFICATION=(pid[,...])
Allows you to select processes for monitoring according to the
process identification (PID). The PID is assigned by the system when
the process is created. When you specify a PID, you can omit the
leading zeros. This qualifier can not be used when the /PROCESS_NAME
qualifier is also specified.
/IMAGE_NAME
/IMAGE_NAME=(imagename[,...])
Allows you to select processes for monitoring that are currently
executing an image that matches the specified image name. Wildcards
are allowed. When you specify the image name as "ANY", all processes
that are currently executing an image are selected. When you specify
the image name as "NONE", all processes that are currently not
executing an image are selected. Note that also device names and
directory names are included in the image name.
/INTERVAL
/INTERVAL=seconds
/NOINTERVAL
Allows you to specify the interval time between two monitor scans in
seconds. The default interval time is 3 seconds. When you specify
the /NOINTERVAL qualifier, the first page will be shown and there
will be no monitoring besides the use of the short cut keys.
/ITEM
/ITEM=item-value
Allows you to specify an item value to signal special conditions.
When the specified condition occurs, this will be indicated on the
screen. When the /ALARM qualifier is specified also, an alarm message
will be generated. Choose one or a combination of the following item
values:
ATTRIBUTE=(attr[,...]) Specifies the way to signal a condition on
the screen. Choose one or a combination of
the following attributes:
BLINK Lets your text blink.
BOLD Displays your text in bold.
REVERSE Displays your text in reverse
video.
UNDERLINE Underlines your text.
The default value is BOLD.
DELETED When specified, you will be notified when a
process is deleted.
OUTSWAPPED When specified, you will be notified when a
process is swapped out.
PAGE_FAULTS=n When specified, you will be notified when
the number of page faults exceeds the
specified number of page faults.
PHYSICAL_MEMORY=n When specified, you will be notified when
the total number of pages currently in the
working set exceeds the specified size.
These number of pages includes both private
and global pages.
SUSPENDED When specified, you will be notified when a
process is suspended.
WS_SIZE=n When specified, you will be notified when
the current working set size exceeds the
specified size.
/JOBTYPE
/JOBTYPE=(keyword[,...])
Allows you to select processes according to the execution mode of the
process at the root of the job tree. Choose one of the following
keywords:
DETACHED
NETWORK
BATCH
LOCAL
DIALUP
REMOTE
/MODE
/MODE=(keyword[,...])
Allows you to select processes according to the process execution
mode. Choose one of the following keywords:
BATCH
INTERACTIVE
NETWORK
OTHER
/NETWORK
Allows you to select all network processes in the system. When used
with the /CLUSTER qualifier, the /NETWORK qualifier selects all
network processes in an OpenVMS cluster environment. When used with
the /NODENAME qualifier, the /NETWORK qualifier selects all network
processes of the specified nodes.
Using this qualifier gives the same result as using the qualifier
/MODE=NETWORK.
/NODENAME
/NODENAME[=(name[,...])]
Selects only processes on the specified node or nodes. Wildcards are
allowed. If you enter the /NODENAME qualifier without a value, the
qualifier selects only the processes on the local node. This
qualifier can not be used when the /CLUSTER qualifier is also
specified.
/PROCESS_NAME
/PROCESS_NAME=(name[,...])
Allows you to select processes for monitoring according to the
process name. Wildcards are allowed. This qualifier can not be used
when the /IDENTIFICATION qualifier is also specified.
/REFRESH
/REFRESH=(item-value[,...])
Allows you to refresh the list of processes that matches the
specified selection criteria. Choose one or combination of the
following item values:
DELETED Specifies that refresh must take place in the next
interval scan after a process has been deleted.
INTERVAL=n Specifies that refresh must take place after n interval
scans.
LIST=n Specifies that refresh must take place after the total
process list has been scanned n times.
By default VIP scans the system or OpenVMS cluster only once for
processes matching the specified selection criteria. These processes
will be monitored every interval time. This is done this way to
minimize the load of VIP on the system. You must be aware that when
you do not specify the /REFRESH qualifier you will not monitor newly
created processes or processes that match the selection criterea in a
later stage. For an actual list of processes every interval please
specify /REFRESH=INTERVAL=1.
When the refresh function scans all processes again to rebuild the
process list, it might be possible that this action takes more time
then one interval period. In this case VIP skips as many interval
periods until the calculated interval time is a time in the future.
To force a refresh by hand please refer to SHORTCUTS.
/SUBPROCESS
Allows you to select all subprocesses in the system. When used with
the /CLUSTER qualifier, the /SUBPROCESS qualifier selects all
subprocesses in an OpenVMS cluster environment. When used with the
/NODENAME qualifier, the /SUBPROCESS qualifier selects all
subprocesses of the specified nodes.
/USERNAME
/USERNAME=(user-name[,...])
Allows you to select processes for monitoring according to the
username of the process. Wildcards are allowed.
EXAMPLES:
1. In this example all processes are monitored that are currently
executing an image of which its name contains the word "SCAN":
$ VIP MONITOR WORKING_SET /IMAGE=*SCAN* -
/NODE=MSSTST
Node : MSSDEV/MSSTST VIP - MONITOR WORKING SET 8-SEP-1996 11:01:49
Page : 1/1 of 2 WS WS WS WS Physical Page
Pid Process Name Extent Quota Default Size Memory faults
-------- --------------- ------ ----- ------- ------ -------- ------
27E0450C JONES_EXE 6000 2048 1024 4024 2913 3635 S
27E0450D SCANNER_LOG 6000 2048 1024 1274 826 108073
The header information in this example includes the following:
o Node name(s)
One or two 1- to 6-character strings used to identify the local node
and remote node.
If the processes on the page that is currently displayed are running
at the local node, only the name of the local node will be displayed.
If the processes on the page that is currently displayed are running
a remote node in an OpenVMS cluster environment two nodenames will
be displayed. The first node indicates the local node, the second
node indicated the remote node where the processes are currently
running. The two node names are separated by a slash character (/).
o Page
An indicator that contains three numeric values.
If the total number of selected processes exceeds the number of
lines on the screen, the list of processes will be displayed in
multiple pages starting with the first page. Every interval the next
page will be displayed. When all pages are displayed, the sequence
starts again.
The first value indicates the page number that is currently beeing
displayed. The second value indicates the total number of pages.
The third value indicates the total number of selected processes.
o Current time
A 23-character string used to show the current time.
The body information in this example includes the following:
o A list of data items for every process and updated every interval
which contains:
- Process identification code
A 32-bit binary value that uniquely identifies a process.
- Process name
A 1- to 15-character string used to identify a process.
- Working set extent
Treshold value to which WS Size can be adjusted.
- Working set quota
Treshhold value to which WS Size can be adjusted.
- Working set default
Default working set size, which is reestablished at each image
activation.
- Working set size
Current size of the working set. When the number of pages actually
allocated (pages in working set) reaches this treshold, subsequent
page faults will cause page replacements.
- Physical memory
The number of pages currently in the working set. These number of
pages includes both private and global pages.
- Page faults
Total number of page faults that have occurred since process
creation.
- Process indicator
Letter B indicates a batch job; letter S indicates a subprocess;
letter N indicates a network process.
|
