/AFFINITY (Alpha/Integrity servers only)
/NOAFFINITY
Allows bits in the kernel thread affinity mask to be set or cleared
individually, in groups, or all at once. This qualifier is mutually
exclusive with the /CAPABILITY qualifier.
Note
The SET PROCESS/[NO]AFFINITY command fails if none of the specified
CPUs has the capabilities required by the process.
|
The /NOAFFINITY qualifier clears all affinity bits currently set in the
current or permanent affinity masks, based on the setting of the
/PERMANENT qualifier. Specifying the /AFFINITY qualifier has no direct
effect, but merely indicates the target of the operations specified by
the following secondary parameters:
/SET=(n[,...])
|
Sets affinity for currently active CPUs defined by the CPU IDs
n, where
n has the range of 0 to 31.
|
/CLEAR=(
n[,...])
|
Clears affinity for currently active CPUs defined by the position values
n, where
n has the range of 0 to 31.
|
/PERMANENT
|
Performs the operation on the permanent affinity mask as well as the
current affinity mask, making the changes valid for the life of the
kernel thread. (The default behavior is to affect only the affinity
mask for the running image.)
|
The secondary qualifiers can all be used at once as long as the
affinity bits defined in the /SET and /CLEAR parameters do not overlap.
The privileges required to execute the SET PROCESS/AFFINITY command
match those required by the $PROCESS_AFFINITY system service. ALTPRI is
the base privilege required to make any modifications, and the only
privilege required to modify the current owner's kernel thread.
Modifications within the same UIC group require GROUP privilege.
Modifications to any unrelated kernel thread require WORLD privilege.
As with the other SET PROCESS qualifiers, the bit operations occur on
the current process if no /IDENTIFICATION qualifier or explicit process
name parameter is specified. Specifying a process name does not imply
that all kernel threads associated with the process are affected; the
SET PROCESS command affects only the initial kernel thread of a
multithreaded process.
/AUTO_UNSHELVE
/NOAUTO_UNSHELVE
Controls whether the process automatically unshelves files.
Note that the /NOAUTO_UNSHELVE qualifier does not work across a
cluster. It can be issued only for a process on the same node,
including as the default case, the process from which the command is
issued.
/CAPABILITY
/NOCAPABILITY
Allows bits in the process user capability mask to be set or cleared
individually, in groups, or all at once. This qualifier is mutually
exclusive with the /AFFINITY qualifier.
Note
The SET PROCESS/[NO]CAPABILITY command fails if there is no CPU with
the required set of capabilities.
|
The /NOCAPABILITY qualifier clears all user capability bits currently
set in the current or permanent capability masks, based on the setting
of the /PERMANENT qualifier. Specifying the /CAPABILITY qualifier has
no direct effect, but merely indicates the target of the operations
specified by the following secondary qualifiers:
/SET=(
n[,...])
|
Sets all user capabilities defined by the position values
n, where
n has the range of 1 to 16.
|
/CLEAR=(
n[,...])
|
Clears all user capabilities defined by the position values
n, where
n has the range of 1 to 16.
|
/PERMANENT
|
Performs the operation on the permanent user capability mask as well as
the current user capability mask, making the changes valid for the life
of the kernel thread. (The default behavior is to affect only the
capabilities mask for the running image.)
|
The secondary qualifiers can all be used at once as long as the user
capability bits defined in the /SET and /CLEAR parameters do not
overlap.
The privileges required to execute the SET PROCESS/CAPABILITY command
match those required by the $PROCESS_CAPABILITIES system service.
ALTPRI is the base privilege required to make any modifications, and
the only privilege required to modify the current owner's kernel
thread. Modifications within the same UIC group require GROUP
privilege. Modifications to any unrelated kernel thread require WORLD
privilege.
As with the other SET PROCESS qualifiers, the bit operations occur on
the current process if no /IDENTIFICATION qualifier or explicit process
name parameter is specified. Specifying a process name does not imply
that all kernel threads associated with the process are affected; the
SET PROCESS command affects only the initial kernel thread of a
multithreaded process.
/CASE_LOOKUP=keyword
The valid keywords for this qualifier are BLIND and SENSITIVE.
HP strongly recommends that you use caution when enabling case
sensitivity in your processes.
See the Guide to OpenVMS File Applications for additional information.
/CLEAR=(n[,...])
Clears all user capabilities or CPU affinities defined by the position
values n specified by the /CAPABILITY or the /AFFINITY
qualifier.
/DUMP[=NOW] [/ID=pid] [process-name]
/NODUMP (default)
Causes the contents of the address space to be written to the file
named (image-name).DMP in the current default directory (where the
image name is the same as the file name) when an image terminates
because of an unhandled error. To specify a target process, indicate
either the process ID or the process name.
Using the NOW option (available on Alpha and Integrity server systems
only) causes the targeted process to dump as soon as possible. This is
especially useful for hung processes.
You can then analyze the dump with the ANALYZE/PROCESS_DUMP utility,
the Debugger, or the System Dump Analyzer (SDA).
/GLOBAL
Modifies the global cell SCH$GL_DEFAULT_PROCESS_CAP. This global cell
is used to initialize the user capability mask of processes when they
are started.
/IDENTIFICATION=pid
Requires GROUP or WORLD privilege for processes other than your
own.
Specifies the process identification (PID) value of the kernel thread
for which characteristics are to be changed. The target process must be
on the same node as the process from which the command is issued. The
/IDENTIFICATION qualifier overrides the process-name parameter.
The PID is assigned by the system when the process is created. When you
specify a PID, you can omit the leading zeros.
If you use the /IDENTIFICATION qualifier, the process-name
parameter is ignored.
Note
The /IDENTIFICATION qualifier allows the SET PROCESS command to affect
individual kernel thread PIDs; because each thread is a separate
runnable entity, this command treats them as discrete entities in terms
of affinities and capabilities.
|
/KERNEL_THREAD_LIMIT=n
Changes the limit of the number of kernel threads that can be created
in the process. If the value specified is higher than the SYSGEN
parameter MULTITHREAD, an error message is returned. Specifying the
value 0 makes the process use the systemwide limit set by the
MULTITHREAD parameter.
/NAME=string
Changes the name of the current process to a string of 1 to 15
characters.
/PARSE_STYLE=(keyword)
Allows a user to indicate how commands are to be parsed.
Setting a particular parse style tells DCL how it should handle command
syntax. User programs can also examine the state of this setting if
they need to use different parse rules.
The valid keywords for this qualifier are TRADITIONAL and EXTENDED.
These keywords are mutually exclusive and cannot be negated.
If the /PARSE_STYLE qualifier is not specified, the default is the
TRADITIONAL format.
Keyword |
Explanation |
TRADITIONAL (default)
|
Indicates that commands should be examined using the former (prior to
Version 7.2) rules for DCL syntax.
|
EXTENDED (Alpha/Integrity servers only)
|
Indicates that commands should be examined using a syntax that allows
ODS-5 file specifications.
|
The main differences for DCL when EXTENDED parse rules are in effect
are:
- Arguments to foreign commands are case preserved. You can get the
command string by calling LIB$GET_FOREIGN.
C/C++ programs that use
the argc/argv mechanism will have unquoted arguments in lowercase
unless the C Run-Time Library logical DECC$ARGV_PARSE_STYLE is set to
ENABLE. When DECC$ARGV_PARSE_STYLE is enabled, case is preserved in
command line arguments when the process is set up for extended DCL
parsing using /PARSE_STYLE=EXTENDED.
- Some characters that were previously treated as token delimiters
are no longer delimiters. The pound sign (#), circumflex (^), and
question mark (?) fall into this category.
- A circumflex (^) is an escape character, which can be used to
indicate that the next character in the command string is to be treated
as if it were quoted, thereby losing its syntactic significance.
EXTENDED parsing also modifies DCL's rules for parsing a parameter or
qualifier that is defined as a file specification in a command's
definition:
- File specifications will not be in uppercase.
- Any number of commas (,) may be placed between directory delimiters
([] and <>).
- Directory file ID's (DIDs) can be included in the file
specification.
- Any number of periods (.) or semi-colons (;) may be included in the
file specification.
See the HP OpenVMS System Manager's Manual, Volume 1: Essentials for more information.
/PERMANENT
Performs the operation on the permanent affinity or capability mask as
well as the current mask, making the changes valid for the life of the
kernel thread. (The default behavior is to affect the current mask only
for the running image.) The /PERMANENT qualifier is only valid in
conjunction with the /CAPABILITY or the /AFFINITY qualifier.
/PRIORITY=n
Requires ALTPRI (alter priority) privilege to set the priority
higher than the base priority of the specified process.
Changes the priority for the specified kernel thread. If you do not
have the ALTPRI privilege, the value you specify is compared to your
current base priority, and the lower value is always used.
/PRIVILEGES=(privilege[,...])
Requires SETPRV (set privilege) privilege as an authorized
privilege to enable any privilege you do not have in your authorized
privilege mask.
Enables the following process privileges:
ACNT
|
ALLSPOOL
|
ALTPRI
|
AUDIT
|
BUGCHK
|
BYPASS
|
CMEXEC
|
CMKRNL
|
DETACH
|
DIAGNOSE
|
DOWNGRADE
|
EXQUOTA
|
GROUP
|
GRPNAM
|
GRPPRV
|
IMPORT
|
LOG_IO
|
MOUNT
|
NETMBX
|
OPER
|
PFNMAP
|
PHY_IO
|
PRMCEB
|
PRMGBL
|
PRMMBX
|
PSWAPM
|
READALL
|
SECURITY
|
SETPRV
|
SHARE
|
SHMEM
|
SYSGBL
|
SYSLCK
|
SYSNAM
|
SYSPRV
|
TMPMBX
|
UPGRADE
|
VOLPRO
|
WORLD
|
|
Use the SHOW PROCESS/PRIVILEGES command to determine what authorized
privileges are enabled currently.
/RAD=HOME=n
Changes the home resource affinity domain (RAD) of a process.
This command only works on the current process.
RAD is supported on AlphaServer GS series systems and starting from OpenVMS Version 8.4, support is extended to NUMA capable Integrity servers.
/RESOURCE_WAIT
/NORESOURCE_WAIT
Enables resource wait mode so that the process waits for resources to
become available. If you specify the /NORESOURCE_WAIT qualifier, the
process receives an error status code when system dynamic memory is not
available or when the process exceeds one of the following resource
quotas: direct I/O (DIOLM) limit, buffered I/O (BIOLM) limit, buffered
I/O byte (BYTLM) count limit (buffer space), timer queue quota, mailbox
buffer quota, or pipe quota.
Caution
Disabling resource waiting should be performed with caution, as doing
so can have unexpected effects on constituent sharable images and
runtime libraries. See the $SETRWM service in the HP OpenVMS System Services Reference Manual for
additional information.
|
/RESUME
Allows a process suspended by a previous SET PROCESS/SUSPEND command to
resume operation. The /RESUME qualifier is equivalent to the /NOSUSPEND
qualifier.
/SCHEDULING_CLASS="class_name"
/NOSCHEDULING_CLASS
Adds a process to, or removes a process from, a scheduling class. This
command does not modify the permanent class scheduling database file.
Therefore, the effect of this command can be viewed as temporary
placement into, or removal from, a scheduling class.
/SET=(n[,...])
Sets all user capabilities or CPU affinities defined by the position
values n specified by the /CAPABILITY or the /AFFINITY
qualifier.
/SSLOG=(STATE={ON | OFF | UNLOAD} [,COUNT=n] [,FLAGS=[NO]ARG])
Valid on Alpha and Integrity server systems only. Requires
CMEXEC, CMKRNL, or SETPRV privilege to log argument values. The SYSGEN
parameter SYSSER_LOGGING must be enabled or the command will fail.
Enables or disables system service logging, using a log file (named
SSLOG.DAT by default) to log data.
Keywords are as follows:
Keyword |
Explanation |
COUNT=
n
|
Specifies how many P2-space buffers to log. (Default: 2)
|
FLAGS=[NO]ARG
|
Specifies whether service argument values are to be logged. (Default:
ARG) ARG requires CMEXEC, CMKRNL, or SETPRV privilege.
|
STATE=state
|
Turns system service logging on or off. Possible states are:
ON
|
Enables system service logging.
|
OFF
|
Disables (turns off) system service logging; logging can still be
reenabled.
|
UNLOAD
|
Stops logging and closes the log file, which is named SSLOG.DAT by
default.
|
|
When enabling SSLOG for a process, you can specify the number of
buffers to be used for logging. Buffers are allocated in P2 space and
are charged against the process's paging file quota. Each buffer is
65,02410 bytes or FE0016 bytes. The buffer space
remains allocated and the quota is charged until the process is deleted.
Between the time when SSLOG is first enabled and when the log file is
closed, logging can be stopped and resumed.
Before you delete the process, you should stop all logging and close
the log file. The log file does not close automatically.
To analyze the log file, use the DCL command ANALYZE/SSLOG, which is
described in online help and in the System Service Logging chapter of
the HP OpenVMS System Analysis Tools Manual.
/SUSPEND[=SUPERVISOR]
/SUSPEND=KERNEL
/NOSUSPEND
Requires privileges as described in text.
Temporarily stops the process's activities. The process remains
suspended until another process resumes or deletes it. Use the
qualifiers /NOSUSPEND and /RESUME to resume a suspended process.
Specify either of the following keywords with the /SUSPEND qualifier to
produce different results:
Keyword |
Explanation |
SUPERVISOR (default)
|
Specifies that the named process is to be suspended to allow the
delivery of asynchronous system traps (ASTs) at EXEC or KERNEL mode.
Specifying this keyword is optional.
|
KERNEL
|
Specifies that the named process is to be suspended such that no ASTs
can be delivered. To specify the KERNEL keyword, you must be in either
KERNEL mode or EXEC mode, or have CMKRNL (change mode to kernel) and
CMEXEC (change mode to executive) privilege enabled. Note that this was
the default behavior of the SET PROCESS/SUSPEND command for versions of
OpenVMS prior to VMS Version 5.0.
|
Depending on the operation, the process from which you specify the
/SUSPEND qualifier requires privileges. You must have GROUP privilege
to suspend another process in the same group, unless that process has
the same user identification code (UIC). You must have WORLD privilege
to suspend any other process in the system.
When you enter the SET PROCESS/SUSPEND=KERNEL/ID= command in a cluster
environment, the KERNEL keyword is ignored if the target process and
the current process reside on different cluster nodes. As a result,
process suspension is handled as if you had specified the SUPERVISOR
keyword (the default).
Note that you can specify SET PROCESS/SUSPEND=KERNEL to override a
previous SET PROCESS/SUSPEND=SUPERVISOR. SET PROCESS/SUSPEND=SUPERVISOR
does not, however, override SET PROCESS/SUSPEND=KERNEL.
/SWAPPING (default)
/NOSWAPPING
Requires the PSWAPM (process swap) privilege to disable
swapping for your process.
Permits the process to be swapped. By default, a process that is not
currently executing can be removed from physical memory so that other
processes can execute. If you specify the /NOSWAPPING qualifier, the
process is not swapped out of the balance set when it is in a wait
state.
/SYMLINK=keyword
Controls the behavior of all directory wildcard searches used in the
RMS$SEARCH service. Note that these directory wildcards are not
overridden by command qualifiers or other explicit RMS characteristics.
Keyword options are:
Keyword |
Explanation |
NOWILDCARD
|
Indicates that symlinks are disabled during directory wildcard searches.
|
WILDCARD
|
Indicates that symlinks are enabled during wildcard searches.
|
NOELLIPSIS
|
Indicates that symlinks are matched for all wildcard fields except for
ellipsis.
|
ELLIPSIS
|
Equivalent to WILDCARD (included for command symmetry).
|
On the command line, you can set a flag to select the /SYMLINK
qualifier. It uses the keywords DEFAULT, [NO]WILDCARD, [NO]ELLIPSIS,
and [NO]TARGET. In addition, /SYMLINK and /NOSYMLINK without any
keywords are interpreted as /SYMLINK=NOTARGET and /SYMINK=TARGET,
respectively.
/TOKEN=keyword
Changes the maximum size of tokens (elements) in a DCL command. (A
token is any element in a command line that is bounded by spaces. For
example, the command COPY X.TXT Y.TXT contains three tokens.) The token
size is determined by the setting of bit 1 in the DCL_CTLFLAGS system
parameter. By default, the bit is clear, indicating that traditional
tokens (255 bytes) are being used. If the bit is set, extended tokens
(4000 bytes) are used. The corresponding keywords for /TOKEN are
TRADITIONAL and EXTENDED. You can use SHOW PROCESS/TOKEN to determine
the current token size setting.
/UNITS[=keyword]
Specifies whether the amount of disk space reported by certain
utilities is to be displayed in blocks or bytes. Keyword options are:
Keyword |
Explanation |
BLOCKS
|
Displays disk space in blocks.
|
BYTES
|
Displays disk space in bytes.
|
Blocks is the default until /UNITS is set to BYTES. If you specify
/UNITS with no keyword, disk space is reported in blocks.
Displays that are affected by changing the value of /UNITS include
output from certain forms of the following commands: COPY, DELETE,
DIRECTORY, PURGE, SHOW DEVICE, SHOW MEMORY, and SHOW QUOTA. Note that
input to these commands can be specified only in blocks. The DIRECTORY,
SHOW DEVICES, and SHOW MEMORY commands have a qualifier that lets you
override the default SET PROCESS/UNITS setting for a single command.