openpbs/doc/man1/pbs_resources.7B

.\" Copyright (C) 1994-2018 Altair Engineering, Inc.
.\" For more information, contact Altair at www.altair.com.
.\"
.\" This file is part of the PBS Professional ("PBS Pro") software.
.\"
.\" Open Source License Information:
.\"
.\" PBS Pro is free software. You can redistribute it and/or modify it under the
.\" terms of the GNU Affero General Public License as published by the Free
.\" Software Foundation, either version 3 of the License, or (at your option) any
.\" later version.
.\"
.\" PBS Pro is distributed in the hope that it will be useful, but WITHOUT ANY
.\" WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
.\" FOR A PARTICULAR PURPOSE.
.\" See the GNU Affero General Public License for more details.
.\"
.\" You should have received a copy of the GNU Affero General Public License
.\" along with this program.  If not, see <http://www.gnu.org/licenses/>.
.\"
.\" Commercial License Information:
.\"
.\" For a copy of the commercial license terms and conditions,
.\" go to: (http://www.pbspro.com/UserArea/agreement.html)
.\" or contact the Altair Legal Department.
.\"
.\" Altair’s dual-license business model allows companies, individuals, and
.\" organizations to create proprietary derivative works of PBS Pro and
.\" distribute them - whether embedded or bundled with other software -
.\" under a commercial license agreement.
.\"
.\" Use of Altair’s trademarks, including but not limited to "PBS™",
.\" "PBS Professional®", and "PBS Pro™" and Altair’s logos is subject to Altair's
.\" trademark licensing policies.
.\"
.TH pbs_resources 7B "18 April 2018" Local "PBS Professional"
.SH NAME
.B pbs_resources 
\- computational resources for PBS jobs


.SH DESCRIPTION
PBS provides computational resources for jobs, limits on using 
resources, and control over how
jobs are placed on the vnodes from which resources may be allocated for
a job.

PBS provides built-in resources and custom resources for some systems, 
and allows the administrator to define
custom resources.  The administrator can specify which resources are
available on a given vnode, as well as at the queue or server level
(e.g. walltime).  Resources can be "stretched" across vnodes.
See the 
.I qmgr(8B)
man page and the 
.I PBS Professional Administrator's Guide.


Resources defined at the queue or server level apply to an entire job.
If resources are defined at the host level, they apply only to the part of
the job running on that host.

For information on defining resources, see the 
.I PBS Professional Administrator's Guide.

.B Allocating Resources to Jobs
.br
Resources are allocated to jobs when jobs explicitly request them, 
when hooks assign them, and when defaults are assigned by PBS.  
Resources are explicitly requested in this order of precedence:
.RS 3
Through a
.B qalter
operation
.br
Via the 
.B qsub
command line
.br
Via PBS job script directives  
.RE 
The administrator writes
any hooks that assign resources to jobs.  Default resources can be
specified by the administrator (in order of precedence) for
.B qsub
arguments, 
queues, the server, and vnodes.  

.B Limits on Resource Usage
.br
A resource allocated to a job, whether explicitly requested or
assigned via hooks or defaults, places a limit on the amount of that
resource a job can use.  This limit applies to how much the job can
use on each vnode and to how much the whole job can use.

.B Placing Jobs on Vnodes
.br
Jobs are placed on vnodes according to their explicit placement request, 
or according to default placement rules.
The explicit placement request can be specified (in order of precedence) using 
.B qalter, qsub, 
and PBS job script directives.
Default placement rules can be specified for queues and the server, 
and rules for default placement take effect if no other placement
specifications exist.  
A job's placement request is specified in its 
.I place statement.

.B Old Syntax
.br
A job submitted with the old node or resource specification syntax
will be converted to the new select and place syntax.  If the job is
submitted with
.I -lnodes= 
or 
.I -lncpus=
it will be converted to 
.I -l select=
and 
.I -l place=.  See 
.B BACKWARD COMPATIBILITY.  
Jobs cannot use both new and old syntax for resource requests.

.B Allocating Chunks and Job-wide Resources
.br
Job resource requests are defined either at the host level in 
.I chunks
specified in a 
.I selection statement,
or as job-wide resources.  
.br
Job-wide format: 
.RS 3
.nf
.I qsub ... -l <resource name>=<value>
.RE
Chunk format:
.RS 3
.I -l select=<chunks>
.fi
.RE
The only resources that can be
requested in chunks are host-level resources, such as 
.I mem 
and 
.I ncpus.
The only resources that can be in a job-wide request are server-level
or queue-level resources, such as 
.I walltime.  

.IP " " 3
.RS
.B Requesting Resources in Chunks
.br
A 
.I chunk 
declares the value of each resource in a set of resources
which are to be allocated as a unit to a job.  All of a chunk 
must be taken from a single vnode.  
A 
.I chunk
request is a host-level request, and it must be for a host-level resource.
A
.I chunk
is the smallest 
set of resources that will be allocated to a job.  It is one or 
more 
.I <resource name>=<value>
statements separated by a colon, e.g.:
.RS 3
ncpus=2:mem=10GB:host=Host1
.br
ncpus=1:mem=20GB:arch=linux
.RE

.I Chunks 
are described in a 
.I selection statement, 
which specifies how many of each kind of chunk.
A selection statement is of the form:
.RS 3
.I -l select=[<N>:]<chunk>[+[<N>:]<chunk> ...]
.RE
If 
.I N 
is not specified, it is taken to be 1.
.br
No spaces are allowed between chunks.

Example of multiple chunks in a 
.I selection statement:
.RS 3
-l select=2:ncpus=1:mem=10GB+3:ncpus=2:mem=8GB
.RE

.B Requesting Job-wide Resources
.br
A job-wide resource request is for resource(s) at the server or queue
level.  This resource must be a server-level or queue-level resource.
A job-wide resource is designed to be used by the entire job, and is
available to the complex, not just one execution host.  Job-wide
resources are requested outside of a
.I selection statement, 
in this form:
.RS 3
.I -l <resource name>=<value>[,<resource name>=<value> ...]
.RE
where 
.I resource name
is either a consumable resource or a time-based resource such
as
.I walltime.

Example of job-wide resources: walltime

.B Do not mix old style resource or node specification with the new select and place statements.
Do not use one in a job script and the other on the command line.
This will result in an error.

See the qsub(1B) man page for a detailed description of how to request 
resources and place jobs on vnodes.

.B Applying Resource Defaults
.br
When a default resource is defined, it is applied to a job when that
job does not explicitly request the resource.  Jobs get default
resources, both
.I job-wide 
and per-
.I chunk
with the following order of precedence, from 
.RS 3
Default
.B qsub
arguments
.br	
Default queue resources
.br	
Default server resources
.RE

For each 
.I chunk
in the job's selection statement, first queue chunk defaults are
applied, then server chunk defaults are applied.  If the chunk does not
contain a resource defined in the defaults, the default is
added.  The chunk defaults are specified with "default_chunk.<resource name>".

For example, if the queue in which the job is enqueued has the
following defaults defined: 
.RS 3
default_chunk.ncpus=1
.br
default_chunk.mem=2gb
.RE
a job submitted with this selection statement:
.RS 3
select=2:ncpus=4+1:mem=9gb 
.RE
will have this specification after the 
.I default_chunk 
elements are applied:
.RS 3
select=2:ncpus=4:mem=2gb+1:ncpus=1:mem=9gb 
.RE
In the above, mem=2gb and ncpus=1 are inherited from 
.I default_chunk.

If a default job-wide resource is defined which is not specified in a
job's resource request, it is added to the resource request.
Queue defaults are applied first, then server defaults are applied
to any remaining gaps in the resource request.

.B Specifying Default Resources
.br
The administrator can specify default resources on the server and
queue.  These resources can be job-wide or apply to chunks.  Job-wide
resources are specified via
.I resources_default 
on the server or queue, and chunk
resources are specified via 
.I default_chunk 
on the server or queue.

The administrator can specify default resources to be added to
any qsub arguments via the server's 
.I default_qsub_arguments
attribute.

.B Specifying Default Placement
.br
The administrator can specify default job placement by setting 
a value for 
.I resources.default 
at the queue or server. 

See the qmgr(8B) man page for how to set default resources.

.B How Default Resources Work When 
.B Moving Jobs Between Queues
.br
If the job is moved from the current queue to a new queue, any default
resources in the job's resource list are removed.  This
includes a select specification and place directive generated by the
rules for conversion from the old syntax.  If a job's resource is
unset (undefined) and there exists a default value at the new queue or server,
that default value is applied to the job's resource list.  If either
.I select 
or 
.I place
is missing from the job's new resource list, it will 
be automatically generated,
using any newly-inherited default values.

Example:
.RS 3
Given the following set of queue and server default values:
.IP Server 3
resources_default.ncpus=1
.IP "Queue QA" 3
resources_default.ncpus=2
.br
default_chunk.mem=2gb
.IP "Queue QB" 3
default_chunk.mem=1gb
.br
no default for ncpus
.LP

The following illustrate the equivalent select specification for jobs
submitted into queue QA and then moved to (or submitted directly to)
queue QB:
.IP "qsub -l ncpus=1 -lmem=4gb" 3
In QA:  select=1:ncpus=1:mem=4gb  -  no defaults need be applied
.br
In QB:  select=1:ncpus=1:mem=4gb  -  no defaults need be applied

.IP "qsub -l ncpus=1" 3
In QA: select=1:ncpus=1:mem=2gb
.br
In QB: select=1:ncpus=1:mem=1gb

.IP "qsub -lmem=4gb" 3
In QA: select=1:ncpus=2:mem=4gb
.br
In QB: select=1:ncpus=1:mem=4gb

.IP "qsub -l nodes=4" 3
In QA: select=4:ncpus=1:mem=2gb
.br
In QB: select=4:mem=1gb
.IP "qsub -l mem=16gb -l nodes=4" 3
In QA: select=4:ncpus=1:mem=4gb
.br
In QB: select=4:ncpus=1:mem=4gb
.RE
.RE
.LP

.B Limits on Resource Usage
.br
Each chunk's
per-chunk limits determine how much of any resource can be used in
that chunk.
Per-chunk resource usage limits are established by per-chunk
resources, both from explicit requests and from defaults.  

.I Job-wide resource limits 
set a limit for per-job resource usage.  
.I Job-wide resource limits 
are established both by requesting job-wide resources and 
by summing per-chunk consumable resources.  
.I Job-wide resource limits 
from sums of all chunks, including defaults, override those from
job-wide defaults and resource requests.  Limits include both
explicitly requested resources and default resources.

If a job-wide resource limit exceeds queue or server restrictions, it
will not be put in the queue or accepted by the server.  If, while
running, a job exceeds its limit for a consumable or time-based
resource, it will be terminated.  See the
.I PBS Professional Administrator's Guide.

.B Controlling Placement of Jobs
.br
Jobs are placed on vnodes according to their 
.I place
statements.  The 
.I place 
statement is specified, in order of precedence, via:
.RS 3
Explicit placement request in 
.B qalter
.br
Explicit placement request in
.B qsub
.br
Explicit placement request in PBS job script directives
.br
Default 
.B qsub
place statement
.br
Queue default placement rules
.br
Server default placement rules
.br
Built-in default conversion and placement rules
.RE

The 
.I place
statement may be not be used without the 
.I select
statement.  

For a detailed description of the 
.I place 
statement, see the qsub(1B) man page, or the
.I PBS Professional User's Guide.

Note that vnodes can have sharing attributes that override
job placement requests.  See the
.I pbs_node_attributes(7B)
man page.
.LP
.RS 3
.B Default Placement
.br
If, after all defaults have been applied to a resource request that
contains a selection statement, there is no place statement, then
.I arrangement
is set to 
.I free.
Default 
.I sharing
is 
.I shared.

If the job's place statement does not contain 
.I group=resource,
then a grouping defined at the queue level may be used, 
or a grouping defined at the server level if there is
none at the queue level.

.B Placement of Jobs Submitted 
.B with Old Syntax
.br
A job submitted with a 
node (
.I -lnodes=
) or resource (
.I -lncpus=
) specification will be converted to select and place,
according to the rules described below in 
.B BACKWARD COMPATIBILITY.
.RE

.B Boolean Resources
.br 
A Boolean resource can be either true or false.  A resource request
for a Boolean specifies the required value for the Boolean resource.  For
example, if some vnodes have
.I green=true 
and some have
.I red=true,
a selection statement for two vnodes, each with one CPU, all green and
no red, would be:
.RS 3
-l select=2:green=true:red=false:ncpus=1
.RE

.B Consumable Resources 
.br
Consumable resources are those whose use by a job reduces the amount
available to other concurrent jobs, e.g. memory 
.I (mem), 
CPUs (ncpus) and licenses.  
Non-consumable resources include time-based resources such as 
.I walltime 
and CPU time
.I (cput), 
and string-value resources such as architecture
.I (arch).

.B Custom Resources
.br
Custom resources are site-defined and site-dependent.  The
administrator defines custom resources.  These are typically used for
licenses and scratch space.  PBS provides custom resources
specifically for Cray systems.  See the
.I PBS Professional Administrator's Guide.

A job requesting a floating license must specify it outside of a
selection statement, as a job-wide resource limit.  A job requesting a
node-locked license must specify it inside a selection statement in a
chunk.  See your system administrator.  Refer to the
.I PBS Professional User's Guide.

Custom resources can be created to be invisible or read-only for
unprivileged users.  See the pbsnodes(8B), pbs_rstat(1B), pbs_rsub(1B),
qalter(1B), qselect(1B), qstat(1B), and qmgr(8B) man pages.
These restricted resources cannot be requested by a job via the 
.B qsub
command, regardless of privilege.  

.B Behavior of Unset Resources
.br
An unset resource is undefined.  An unset numerical resource
(i.e. float, long, size, or time) at the host level behaves as if its
value is zero, but at the server or queue level it behaves as if it
were infinite.  An unset string or string array resource at the
server, queue or vnode level cannot be matched by a job's resource
request.  An unset Boolean resource at a server, queue, or vnode
behaves as if that resource is set to "false".

.SH Resources Built Into PBS

.IP accelerator 8
Indicates whether this vnode is associated with an accelerator.
Used for requesting accelerators.
.br
On Cray, this resource exists only when there is at least one associated
accelerator. 
.br 
Behavior:
.RS
.IP True 3
On Cray, this is set to 
.I True 
when there is at least one associated accelerator whose state is UP.  
.IP False 3
On Cray, set to 
.I False 
when all associated accelerators are in state DOWN.  
.RE
.IP
Host-level.  Can be requested only inside of a select statement.  
.br
Not consumable.
.br
Format: 
.I Boolean
.br
Python type: 
.I bool
.br
Default: 
.I False

.IP accelerator_memory 8
Indicates amount of memory for accelerator(s) associated with this
vnode.  
.br
On Cray, PBS sets this resource only on vnodes with at
least one accelerator with state = UP.  For Cray, PBS sets this
resource on the 0th NUMA node (the vnode with PBScrayseg=0), and the
resource is shared by other vnodes on the compute node.
.br
For example, on vnodeA_2_0: 
.br
.nf
   resources_available.accelerator_memory=4196mb 
On vnodeA_2_1: 
   resources_available.accelerator_memory=@vnodeA_2_0
.fi
.br
A scheduler rounds all resources of type 
.I size
up to the nearest kb.
.br
Host-level.  Can be requested only inside of a select statement.  
.br
Consumable.  
.br
Format: 
.I Size
.br
Python type: 
.I pbs.size
.br
Default: No default

.IP accelerator_model 8
Indicates model of accelerator(s) associated with this vnode.
.br
On Cray, PBS sets this resource only on vnodes with at
least one accelerator with state = UP.  
.br
Host-level.  Can be requested only inside of a select statement.
.br
Non-consumable.  
.br
Format: 
.I String
.br
Python type: 
.I str
.br
Default: No default

.IP aoe 8
List of AOEs (Application Operating Environments) 
that can be instantiated on this vnode.  Case-sensitive.  
An AOE is the environment that results from provisioning a vnode.
Each job can request at most one AOE.  Cannot be set on server's host.
.br
Valid values: Allowable values are site-dependent.  
.br
Host-level.  Can be requested only inside of a select statement.  
.br
Non-consumable.
.br
Type: 
.I String_array
.br
Python type: 
.I str
.br
Default: No default

.IP arch 8
System architecture.  One architecture
can be defined for a vnode.  One architecture can be requested per
vnode.  
.br
Valid values: 
.RS 11
Allowable values and effect on job placement are site-dependent.  
.RE
.IP
Host-level.  Can be requested only inside of a select statement.
.br
Non-consumable.  
.br
Type: 
.I String
.br
Python type: 
.I str
.br
Default: No default

.IP cput 8
Amount of CPU time used by the job for all processes on all
vnodes.  Establishes a job-wide resource limit.  
.br
Job-wide.  Can be requested only outside of a select statement.
.br
Non-consumable.
.br
Type: 
.I Duration
.br
Python type: 
.I pbs.duration
.br
Default: No default

.IP energy 8
The energy used by a job.  Set by PBS.  
.br
Consumable.
.br
Format: 
.I Float
.br
Units: 
.I kWh
.br
Default: No default

.IP eoe 8
Stands for "Energy Operational Environment".  When set on a vnode in
.I resources_available.eoe, 
contains the list of available power profiles.  When set for a job
in
.I Resource_List.eoe,
can contain at most one power profile.  (A job can request only one
power profile.)  
.br
Non-consumable.
.br
Format: 
.I String_array
.br
Python type:
.I str
.br
Default value for 
.I resources_available.eoe: 
unset
.br
Default value for 
.I Resource_List.eoe:
no default

.IP exec_vnode 8
The vnodes that PBS estimates this job will use.  Cannot
be requested for a job; used for reporting only. Read-only.  
.br
Type: 
.I String
.br
Python type: 
.I str
.br
Default: No default

.IP file 8
Size of any single file that may be created by the job.
.br
The scheduler rounds all resources of type 
.I size
up to the nearest kb.
.br
Job-wide.  Can be requested only outside of a select statement.
.br
Type: 
.I Size
.br
Python type: 
.I pbs.size
.br
Default: No default

.IP hbmem 8
High-bandwidth memory.  Available only on some architectures such as 
Xeon Phi KNL.
.br
Valid values: Greater than or equal to zero.
.br
Host-level.  
.br
Format:
.I Size
.br
Python type:
.I pbs.size
.br
Default: No default

.IP host 8
Name of execution host.  Cannot be changed.  Site-dependent.
.br
Can be requested only inside of a select statement.  
.br
Behavior:
.RS 11
Automatically set to the short form of the hostname in the 
.I Mom 
attribute. 
.br
On Cray compute node, set to 
.I <mpp_host>_<nid>.
.RE
.IP
Type: 
.I String
.br
Python type: 
.I str

.IP max_walltime 8
Maximum walltime allowed for a shrink-to-fit job.  Job's actual
walltime is between 
.I max_walltime 
and 
.I min_walltime.  
PBS sets 
.I walltime
for a shrink-to-fit job.  If 
.I max_walltime 
is specifed, 
.I min_walltime
must also be specified.  
Cannot be used for 
.I resources_min 
or 
.I resources_max.
Cannot be set on job arrays or reservations.  
.br
Valid values: Must be greater than or equal to
.I min_walltime.  
.br
Can be requested only outside of a select statement.  
.br
Non-consumable.
.br
Format: 
.I Duration
.br
Python type: 
.I pbs.duration
.br
Default: 
.I 5 years

.IP mem 8
Amount of physical memory i.e. workingset allocated to 
the job, either job-wide or host-level.  
.br
The scheduler rounds all resources of type 
.I size
up to the nearest kb.
.br
Can be requested only inside of a select statement. 
.br
Consumable.
.br
Format: 
.I Size
.br
Python type: 
.I pbs.size
.br
Default: No default

.IP min_walltime 8
Minimum walltime allowed for a shrink-to-fit job.  When 
.I min_walltime
is specified, job is a shrink-to-fit job.  If this attribute is set,
PBS sets the job 
.I walltime.  
Job's actual 
.I walltime 
is between
.I max_walltime 
and 
.I min_walltime.  
Cannot be used for 
.I resources_min 
or 
.I resources_max.
Cannot be set on job arrays or reservations.  
.br
Valid values: Must be less than or equal to
.I max_walltime.  
.br
Can be requested only outside of a select statement.  
.br
Non-consumable.
.br
Type: 
.I Duration
.br
Python type: 
.I pbs.duration
.br
Default: No default

.IP mpiprocs 8
Number of MPI processes for this chunk.  Cannot use sum from chunks
as job-wide limit.
.br
The number of lines in PBS_NODEFILE is the sum of the values
of 
.I mpiprocs
for all chunks requested by the job.  For each chunk with 
.I mpiprocs=P, 
the host name for that chunk is written to the PBS_NODEFILE
.I P
times.  
.br
Host-level.  Can be requested only inside of a select statement.
.br
Format: 
.I Integer
.br
Python type: 
.I int
.br
Default: 
.RS 11
.nf
If ncpus > 0
\ \ \ 1
Otherwise
\ \ \ 0
.fi
.RE
.IP

.IP naccelerators 8
Number of accelerators on the host.  PBS sets this resource to the
number of accelerators with state = 
.I UP.
.br
On Cray, PBS sets this resource only on vnodes whose hosts have at 
least one accelerator with state = 
.I UP.  
For Cray, PBS sets this resource on the 0th NUMA
node (the vnode with PBScrayseg=0), and the resource is shared by
other vnodes on the compute node.
.nf
For example, on vnodeA_2_0:
   resources_available.naccelerators=1
On vnodeA_2_1:
   resources_available.naccelerators=@vnodeA_2_0
.fi
Host-level.  Can be requested only inside of a select statement.
.br
Consumable.
.br
Format: 
.I Integer
.br
Python type: 
.I int
.br
Default: No default

.IP nchunk 8
Number of chunks requested between plus symbols in a
select statement.  For example, if the select statement is 
.br
\ \ \ -lselect=4:ncpus=2+12:ncpus=8 
.br
the value of nchunk for the first part is 4, and
for the second part it is 12.  The 
.I nchunk 
resource cannot be named in
a select statement; it can only be specified as a number preceding the
colon, as in the above example.  When the number is omitted, 
.I nchunk 
is 
.I 1.
.br
This resource can be used to specify the default
number of chunks at the server or queue 
Example:  
.br
\ \ \ set queue myqueue default_chunk.nchunk=2
.br
This resource cannot be used in server and queue 
.I resources_min 
and 
.I resources_max.
.br
Non-consumable.
.br
Format: 
.I Integer
.br
Python type: 
.I int
.br
Default value: 
.I 1

.IP ncpus 8
Number of processors.
.br
Can be requested only inside of a select statement.
.br
Consumable.  
.br
Format: 
.I Integer
.br
Python type: 
.I int
.br
Default: No default

.IP nice 8
Nice value with which the job is to be run.  Host-dependent.
.br
Can be requested only outside of a select statement.
.br
Format: 
.I Integer
.br
Python type: 
.I int
.br
Default: No default

.IP nodect 8
.B Deprecated.    
Number of chunks in resource request from selection
directive, or number of nodes requested from node specification.
Otherwise defaults to value of 1.  Can be requested only outside 
of a select statement.  Read-only.
.br
Format:
.I Integer
.br
Python type:
.I int
.br
Default: 
.I 1

.IP nodes
.B Deprecated.
.br
Number of hosts requested.  
.br
Format:
.I Integer  
.br
See 
.B BACKWARD COMPATIBILITY.

.IP ompthreads 8
Number of OpenMP threads for this chunk.  
.br
Cannot use sum from chunks as job-wide limit.  
.br
For the MPI process with rank 0, the environment variables NCPUS and  
OMP_NUM_THREADS are set to the value of 
.I ompthreads.
For other MPI processes, behavior is dependent on MPI implementation.
See the 
.I PBS Professional Administrator's Guide.
.br
Can be requested only inside of a select statement.
.br
Non-consumable.  
.br
Format:
.I Integer
.br
Python type:
.I int
.br
Default: Value of
.I ncpus 

.IP PBScrayhost 8
Used to differentiate a Cray system, containing
ALPS, login nodes running PBS MoMs, and compute nodes, from a separate
Cray system with a separate ALPS.  
.br
Non-consumable.  
.br
Format: 
.I String
.br
Python type:
.I str
.br
Default: Value of 
.I mpp_host 
for this system

.IP "PBScraylabel_<label name>" 8
Tracks labels applied to compute nodes.  
.br
For each label on a compute node, PBS creates a custom resource 
whose name is a concatenation of
.I PBScraylabel_ 
and the name of the label.  
Name format: PBScraylabel_<label name>
.br
For example, if the label name is 
.I Blue,
the name of this resource is 
.I PBScraylabel_Blue.
.br
Behavior: 
.RS 11
PBS sets the value of the resource to 
.I True 
on all vnodes representing the compute node.
.RE
.IP
Format: 
.I Boolean
.br
Python type:
.I bool
.br
Default: No default

.IP PBScraynid 8
Tracks the node ID of the associated compute node.  All vnodes
representing a particular compute node share a value
for 
.I PBScraynid.  
.br 
Non-consumable.  
.br 
Format: 
.I String 
.br 
Python type:
.I str
.br
Default: Value of 
.I node_id 
for this compute node

.IP PBScrayorder 8
Tracks the order in which compute nodes are listed in the Cray
inventory.  All vnodes associated with a particular compute node share
a value for 
.I PBScrayorder.  
.br
Behavior:
.RS 11
Vnodes for the first compute node listed are assigned a value of 
.I 1 
for 
.I PBScrayorder.  
.br
The vnodes for each subsequent compute node listed are assigned a value
one greater than the previous value.  
.RE
.IP 
Do not use this resource in a resource request.
.br
Non-consumable.  
.br
Format: 
.I Integer 
.br
Python type: 
.I int
.br
Default: No default

.IP PBScrayseg 8
Not used. 
.br
Format: 
.I String
.br
Default: No default

.IP pcput 8
Amount of CPU time allocated to any single process in the job.
Establishes a per-process resource limit.  
.br
Can be requested only outside of a select statement.
.br
Non-consumable.
.br
Format:
.I Duration
.br
Python type:
.I pbs.duration
.br
Default: No default

.IP pmem 8
Amount of physical memory (workingset) for use by any single
process of the job.  Establishes a
per-process resource limit.  
.br
The scheduler rounds all resources of type
.I size
up to the nearest kb.
.br
Can be requested only outside of a select statement.
.br
Non-consumable.
.br
Format:
.I Size
.br
Python type:
.I pbs.size
.br
Default: No default

.IP preempt_targets
List of resources and/or queues.  Jobs requesting those resources or
in those queues are preemption targets.  
.br
Job-wide.  Can be requested only outside of a select statement.  
.br
Non-consumable.  
.br
Format:
.I String_array
.br
Syntax:
.RS 11
preempt_targets="Queue=<queue name>[,Queue=<queue name>],
Resource_List.<resource>=<value>[,Resource_List.<resource>=<value>]" 
.RE
.IP
or 
.RS 11
preempt_targets=None
.RE
.IP 
Keywords "queue" and "none" are case-insensitive.  You can
list multiple comma-separated targets.  
.br
Python type: 
.I str
.br
Default: No default

.IP pvmem 8
Amount of virtual memory for use by any single process in the
job.  Establishes a per-process resource limit.  
.br
The scheduler rounds all resources of type
.I size
up to the nearest kb.
.br
Can be requested only outside of a select statement.
.br
Non-consumable.  
.br
Format:
.I Size
.br
Python type:
.I pbs.size
.br
Default: No default

.IP site 8
Arbitrary string resource.  
.br
Can be requested only outside of a select statement.  
.br
Non-consumable.  
.br
Format:
.I String
.br  
Python type:
.I str
.br
Default: No default

.IP software 8
Site-specific software specification.  
.br
Values: Allowable values and effect on job placement are site-dependent.  
.br
Can be requested only outside of a select statement.
.br
Format:
.I String
.br
Python type:
.I pbs.software
.br
Default: No default

.IP soft_walltime 8
Soft limit on walltime.  Similar to 
.I walltime, 
but cannot be requested by unprivileged users, and job is not killed
if it exceeds its 
.I soft_walltime.  
A job's 
.I soft_walltime 
cannot exceed its 
.I walltime.  
Can be set by Manager only.
.br
Format:
.I Duration
.br
Python type:
.I pbs.duration
.br
Default: No default

.IP start_time 8
The estimated start time for this job.  Cannot be requested
for a job; used for reporting only.  Appears only in job's 
.I estimated
attribute.  Read-only.
.br
Host-level.
.br
Consumable.
.br
Format:
.I Integer
.br
Python type:
.I int
.br
Default: No default

.IP vmem 8
Amount of virtual memory for use by all concurrent processes in
the job.  Establishes a per-chunk resource limit.  
.br
The scheduler rounds all resources of type
.I size
up to the nearest kb
.br
Can be requested only inside of a select statement.
.br
Consumable.
.br
Format:
.I Size
.br
Python type:
.I pbs.size
.br
Default: No default

.IP vnode 8
Name of virtual node (vnode) on which to execute.  Site-dependent.  
See the 
.I pbs_node_attributes(7B) 
man page.
.br
Can be requested only inside of a select statement.
.br
Format:
.I String
.br
Python type:
.I str
.br
Default: No default

.IP vntype 8
The type of the vnode.  
Automatically set by PBS to one of two specific values for Cray vnodes.  
Has no meaning for non-Cray vnodes.  
.br
Automatically assigned values for Cray vnodes:
.RS 11
.IP cray_compute 3
This vnode represents part of a compute node.
.IP cray_login 3
This vnode represents a login node.
.RE
.IP
Can be requested only inside of a select statement.
.br
Non-consumable. 
.br
Format: 
.I String_array
.br
Python type:
.I str
.br
Default:  No default

.IP walltime 8
Amount of wall-clock time.  Establishes a job-wide resource limit. 
.br
Actual elapsed time may differ from 
.I walltime 
during Daylight Savings transitions.
.br 
Can be requested only outside of a select statement.
.br
Non-consumable.  
.br
Format:
.I Duration
.br
Python type:
.I pbs.duration
.br
Default: 
.I 5 years

.SH BACKWARD COMPATIBILITY
.B Conversion to Select and Place
.br
For backward compatibility, a legal node specification or resource
specification is converted into selection and placement directives.

.B Node Specification Conversion
.br
Node specification format:
.IP " " 3
-lnodes=[<N>:<spec list> | <spec list>]
.br
        [[+<N>:<spec list> | +<spec list>] ...]
.br
        [#<suffix> ...][-lncpus=Z]
.LP
where:
.IP " " 3
.I spec list  
has syntax:  
.I <spec>[:<spec> ...]
.br
.I spec 
is any of:  
.I <hostname> | <property> | ncpus=X | cpp=X | ppn=P 
.br
.I suffix 
is any of: 
.I <property> | excl | shared
.br
.I N
and 
.I P 
are positive integers
.br
.I X 
and 
.I Z 
are non-negative integers
.LP
The node specification is converted into selection and placement directives as follows:
.IP " " 3
Each
.I spec list 
is converted into one chunk, 
so that 
.I <N>:<spec list> 
is converted into 
.I N 
chunks.

If 
.I spec 
is 
.I hostname
:
.br
The chunk will include 
.I host=<hostname>

If 
.I spec 
matches any vnode's 
.I resources_available.host 
value:
.br
The chunk will include 
.I host=<hostname>

If 
.I spec 
is 
.I <property>
:
.br
The chunk will include 
.I <property>=true
.br
where
.I property 
must be a site-defined host-level Boolean resource.

If 
.I spec 
is 
.I ncpus=X 
or 
.I cpp=X
:
.br 
The chunk will include 
.I ncpus=X

If no 
.I spec 
is 
.I ncpus=X 
and no 
.I spec 
is 
.I cpp=X
:
.br 
The chunk will include 
.I ncpus=1

If 
.I spec 
is 
.I ppn=P
:
.br
The chunk will include 
.I mpiprocs=P
.br
Example: 
    -lnodes=4:ppn=2 
.br
is converted into 
.br
    -lselect=4:ncpus=2:mpiprocs=2

If 
.I -lncpus=Z 
is specified and no 
.I spec 
contains 
.I ncpus=X 
and no 
.I spec 
is 
.I ccp=X
:
.br
Every chunk will include 
.I ncpus=W,
.br
where 
.I W 
is 
.I Z 
divided by the total number of chunks.
.br
(Note:
.I W 
must be an integer; 
.I Z 
must be evenly divisible by the number of chunks.)

If 
.I property 
is a 
.I suffix 
:
.br
All chunks will include 
.I <property>=true

If 
.I excl 
is a 
.I suffix 
:
.br
The placement directive will be 
.I -lplace=scatter:excl

If 
.I shared 
is a 
.I suffix
:
.br
The placement directive will be 
.I -lplace=scatter:shared

If neither 
.I excl 
nor shared is a 
.I suffix
:
.br 
The placement directive will be 
.I -lplace=scatter
.LP
Example: 
.IP " " 3
-l nodes=3:green:ncpus=2:ppn=2+2:red
.LP
is converted to:
.IP " " 3
-l select=3:green=true:ncpus=4:mpiprocs=2+2:red=true:ncpus=1 
.br
-l place=scatter
.LP
Node specification syntax for requesting properties is 
.B deprecated.
The new Boolean resource syntax "<property>=<value>" is accepted only in a
selection directive.  It is erroneous to mix old and new syntax.

.B Resource Specification Conversion
.br
The resource specification is converted to select and place statements 
after any defaults have been applied.

Resource specification format:
.IP " " 3
-l<resource>=<value>[:<resource>=<value> ...]
.LP
The resource specification is converted to:
.IP " " 3
select=1[:<resource>=<value> ...]
.br
place=pack
.br
.LP
with one instance of 
.I <resource>=<value> 
for each of the following 
host-level resources in the resource request:
.IP " " 3
Built-in resources: 
.I ncpus 
| 
.I mem 
| 
.I vmem 
| 
.I arch 
| 
.I host

Site-defined host-level resources with flags including "h"

.SH SEE ALSO
The
.I PBS Professional Administrator's Guide,
the 
.I PBS Professional User's Guide,
pbs_node_attributes(7B),
pbs_rsub(1B),
qalter(1B), 
qmgr(8B), 
qstat(1B),
qsub(1B)