Provided by: slurm-llnl_2.6.5-1_amd64 
NAME
slurm.conf - Slurm configuration file
DESCRIPTION
slurm.conf is an ASCII file which describes general SLURM configuration information, the nodes to be
managed, information about how those nodes are grouped into partitions, and various scheduling parameters
associated with those partitions. This file should be consistent across all nodes in the cluster.
The file location can be modified at system build time using the DEFAULT_SLURM_CONF parameter or at
execution time by setting the SLURM_CONF environment variable. The SLURM daemons also allow you to
override both the built-in and environment-provided location using the "-f" option on the command line.
The contents of the file are case insensitive except for the names of nodes and partitions. Any text
following a "#" in the configuration file is treated as a comment through the end of that line. Changes
to the configuration file take effect upon restart of SLURM daemons, daemon receipt of the SIGHUP signal,
or execution of the command "scontrol reconfigure" unless otherwise noted.
If a line begins with the word "Include" followed by whitespace and then a file name, that file will be
included inline with the current configuration file.
Note on file permissions:
The slurm.conf file must be readable by all users of SLURM, since it is used by many of the SLURM
commands. Other files that are defined in the slurm.conf file, such as log files and job accounting
files, may need to be created/owned by the user "SlurmUser" to be successfully accessed. Use the "chown"
and "chmod" commands to set the ownership and permissions appropriately. See the section FILE AND
DIRECTORY PERMISSIONS for information about the various files and directories used by SLURM.
PARAMETERS
The overall configuration parameters available include:
AccountingStorageBackupHost
The name of the backup machine hosting the accounting storage database. If used with the
accounting_storage/slurmdbd plugin, this is where the backup slurmdbd would be running. Only used
for database type storage plugins, ignored otherwise.
AccountingStorageEnforce
This controls what level of association-based enforcement to impose on job submissions. Valid
options are any combination of associations, limits, nojobs, nosteps, qos, safe, and wckeys, or
all for all things.
If limits, qos, or wckeys are set, associations will automatically be set.
If wckeys is set, TrackWCKey will automatically be set.
If safe is set, limits and associations will automatically be set.
If nojobs is set nosteps will automatically be set.
By enforcing Associations no new job is allowed to run unless a corresponding association exists
in the system. If limits are enforced users can be limited by association to whatever job size or
run time limits are defined.
If nojobs is set Slurm will not account for any jobs or steps on the system, like wise if nosteps
is set Slurm will not account for any steps ran limits will still be enforced.
If safe is enforced a job will only be launched against an association or qos that has a
GrpCPUMins limit set if the job will be able to run to completion. Without this option set, jobs
will be launched as long as their usage hasn't reached the cpu-minutes limit which can lead to
jobs being launched but then killed when the limit is reached.
With qos and/or wckeys enforced jobs will not be scheduled unless a valid qos and/or workload
characterization key is specified.
When AccountingStorageEnforce is changed, a restart of the slurmctld daemon is required (not just
a "scontrol reconfig").
AccountingStorageHost
The name of the machine hosting the accounting storage database. Only used for database type
storage plugins, ignored otherwise. Also see DefaultStorageHost.
AccountingStorageLoc
The fully qualified file name where accounting records are written when the AccountingStorageType
is "accounting_storage/filetxt" or else the name of the database where accounting records are
stored when the AccountingStorageType is a database. Also see DefaultStorageLoc.
AccountingStoragePass
The password used to gain access to the database to store the accounting data. Only used for
database type storage plugins, ignored otherwise. In the case of SLURM DBD (Database Daemon) with
MUNGE authentication this can be configured to use a MUNGE daemon specifically configured to
provide authentication between clusters while the default MUNGE daemon provides authentication
within a cluster. In that case, AccountingStoragePass should specify the named port to be used
for communications with the alternate MUNGE daemon (e.g. "/var/run/munge/global.socket.2"). The
default value is NULL. Also see DefaultStoragePass.
AccountingStoragePort
The listening port of the accounting storage database server. Only used for database type storage
plugins, ignored otherwise. Also see DefaultStoragePort.
AccountingStorageType
The accounting storage mechanism type. Acceptable values at present include
"accounting_storage/filetxt", "accounting_storage/mysql", "accounting_storage/none",
"accounting_storage/pgsql", and "accounting_storage/slurmdbd". The "accounting_storage/filetxt"
value indicates that accounting records will be written to the file specified by the
AccountingStorageLoc parameter. The "accounting_storage/mysql" value indicates that accounting
records will be written to a MySQL database specified by the AccountingStorageLoc parameter. The
"accounting_storage/pgsql" value indicates that accounting records will be written to a PostgreSQL
database specified by the AccountingStorageLoc parameter. The "accounting_storage/slurmdbd" value
indicates that accounting records will be written to the SLURM DBD, which manages an underlying
MySQL or PostgreSQL database. See "man slurmdbd" for more information. The default value is
"accounting_storage/none" and indicates that account records are not maintained. Note: The
filetxt plugin records only a limited subset of accounting information and will prevent some sacct
options from proper operation. Note: The PostgreSQL plugin is not complete and should not be used
if wanting to use associations. It will however work with basic accounting of jobs and job steps.
If interested in completing, please email slurm-dev@schedmd.com. Also see DefaultStorageType.
AccountingStorageUser
The user account for accessing the accounting storage database. Only used for database type
storage plugins, ignored otherwise. Also see DefaultStorageUser.
AccountingStoreJobComment
If set to "YES" then include the job's comment field in the job complete message sent to the
Accounting Storage database. The default is "YES".
AcctGatherNodeFreq
The AcctGather plugins sampling interval for node accounting. For AcctGather plugin values of
none, this parameter is ignored. For all other values this parameter is the number of seconds
between node accounting samples. For the acct_gather_energy/rapl plugin, set a value less than 300
because the counters may overflow beyond this rate. The default value is zero. This value
disables accounting sampling for nodes. Note: The accounting sampling interval for jobs is
determined by the value of JobAcctGatherFrequency.
AcctGatherEnergyType
Identifies the plugin to be used for energy consumption accounting. The jobacct_gather plugin and
slurmd daemon call this plugin to collect energy consumption data for jobs and nodes. The
collection of energy consumption data takes place on node level, hence only in case of exclusive
job allocation the energy consumption measurements will reflect the jobs real consumption. In case
of node sharing between jobs the reported consumed energy per job (through sstat or sacct) will
not reflect the real energy consumed by the jobs.
Configurable values at present are:
acct_gather_energy/none
No energy consumption data is collected.
acct_gather_energy/ipmi
Energy consumption data is collected from the Baseboard Management Controller
(BMC) using the Intelligent Platform Management Interface (IPMI).
acct_gather_energy/rapl
Energy consumption data is collected from hardware sensors using the Running
Average Power Limit (RAPL) mechanism. Note that enabling RAPL may require the
execution of the command "sudo modprobe msr".
AcctGatherInfinibandType
Identifies the plugin to be used for infiniband network traffic accounting. The plugin is
activated only when profiling on hdf5 files is activated and the user asks for network data
collection for jobs through --profile=Network (or =All). The collection of network traffic data
takes place on node level, hence only in case of exclusive job allocation the collected values
will reflect the jobs real traffic. All network traffic data are logged on hdf5 files per job on
each node. No storage on the Slurm database takes place.
Configurable values at present are:
acct_gather_infiniband/none
No infiniband network data are collected.
acct_gather_infiniband/ofed
Infiniband network traffic data are collected from the hardware monitoring
counters of Infiniband devices through the OFED library.
AcctGatherFilesystemType
Identifies the plugin to be used for filesystem traffic accounting. The plugin is activated only
when profiling on hdf5 files is activated and the user asks for filesystem data collection for
jobs through --profile=Network (or =All). The collection of filesystem traffic data takes place on
node level, hence only in case of exclusive job allocation the collected values will reflect the
jobs real traffic. All filesystem traffic data are logged on hdf5 files per job on each node. No
storage on the Slurm database takes place.
Configurable values at present are:
acct_gather_filesystem/none
No filesystem data are collected.
acct_gather_filesystem/lustre
Lustre filesystem traffic data are collected from the counters found in
/proc/fs/lustre/.
AcctGatherProfileType
Identifies the plugin to be used for detailed job profiling. The jobacct_gather plugin and slurmd
daemon call this plugin to collect detailed data such as I/O counts, memory usage, or energy
consumption for jobs and nodes. There are interfaces in this plugin to collect data as step start
and completion, task start and completion, and at the account gather frequency. The data collected
at the node level is related to jobs only in case of exclusive job allocation.
Configurable values at present are:
acct_gather_profile/none
No profile data is collected.
acct_gather_profile/hdf5
This enables the HDF5 plugin. The directory where the profile files are stored
and which values are collected are configured in the acct_gather.conf file.
AuthType
The authentication method for communications between SLURM components. Acceptable values at
present include "auth/none", "auth/authd", and "auth/munge". The default value is "auth/munge".
"auth/none" includes the UID in each communication, but it is not verified. This may be fine for
testing purposes, but do not use "auth/none" if you desire any security. "auth/authd" indicates
that Brett Chun's authd is to be used (see "http://www.theether.org/authd/" for more information.
Note that authd is no longer actively supported). "auth/munge" indicates that LLNL's MUNGE is to
be used (this is the best supported authentication mechanism for SLURM, see
"http://munge.googlecode.com/" for more information). All SLURM daemons and commands must be
terminated prior to changing the value of AuthType and later restarted (SLURM jobs can be
preserved).
BackupAddr
The name that BackupController should be referred to in establishing a communications path. This
name will be used as an argument to the gethostbyname() function for identification. For example,
"elx0000" might be used to designate the Ethernet address for node "lx0000". By default the
BackupAddr will be identical in value to BackupController.
BackupController
The name of the machine where SLURM control functions are to be executed in the event that
ControlMachine fails. This node may also be used as a compute server if so desired. It will come
into service as a controller only upon the failure of ControlMachine and will revert to a
"standby" mode when the ControlMachine becomes available once again. This should be a node name
without the full domain name. I.e., the hostname returned by the gethostname() function cut at
the first dot (e.g. use "tux001" rather than "tux001.my.com"). While not essential, it is
recommended that you specify a backup controller. See the RELOCATING CONTROLLERS section if you
change this.
BatchStartTimeout
The maximum time (in seconds) that a batch job is permitted for launching before being considered
missing and releasing the allocation. The default value is 10 (seconds). Larger values may be
required if more time is required to execute the Prolog, load user environment variables (for Moab
spawned jobs), or if the slurmd daemon gets paged from memory.
CacheGroups
If set to 1, the slurmd daemon will cache /etc/groups entries. This can improve performance for
highly parallel jobs if NIS servers are used and unable to respond very quickly. The default
value is 0 to disable caching group data.
CheckpointType
The system-initiated checkpoint method to be used for user jobs. The slurmctld daemon must be
restarted for a change in CheckpointType to take effect. Supported values presently include:
checkpoint/aix for IBM AIX systems only
checkpoint/blcr Berkeley Lab Checkpoint Restart (BLCR). NOTE: If a file is found at sbin/scch
(relative to the SLURM installation location), it will be executed upon
completion of the checkpoint. This can be a script used for managing the
checkpoint files. NOTE: SLURM's BLCR logic only supports batch jobs.
checkpoint/none no checkpoint support (default)
checkpoint/ompi OpenMPI (version 1.3 or higher)
checkpoint/poe for use with IBM POE (Parallel Operating Environment) only
ClusterName
The name by which this SLURM managed cluster is known in the accounting database. This is needed
distinguish accounting records when multiple clusters report to the same database. Because of
limitations in some databases, any upper case letters in the name will be silently mapped to lower
case. In order to avoid confusion, it is recommended that the name be lower case.
CompleteWait
The time, in seconds, given for a job to remain in COMPLETING state before any additional jobs are
scheduled. If set to zero, pending jobs will be started as soon as possible. Since a COMPLETING
job's resources are released for use by other jobs as soon as the Epilog completes on each
individual node, this can result in very fragmented resource allocations. To provide jobs with
the minimum response time, a value of zero is recommended (no waiting). To minimize fragmentation
of resources, a value equal to KillWait plus two is recommended. In that case, setting KillWait
to a small value may be beneficial. The default value of CompleteWait is zero seconds. The value
may not exceed 65533.
ControlAddr
Name that ControlMachine should be referred to in establishing a communications path. This name
will be used as an argument to the gethostbyname() function for identification. For example,
"elx0000" might be used to designate the Ethernet address for node "lx0000". By default the
ControlAddr will be identical in value to ControlMachine.
ControlMachine
The short hostname of the machine where SLURM control functions are executed (i.e. the name
returned by the command "hostname -s", use "tux001" rather than "tux001.my.com"). This value must
be specified. In order to support some high availability architectures, multiple hostnames may be
listed with comma separators and one ControlAddr must be specified. The high availability system
must insure that the slurmctld daemon is running on only one of these hosts at a time. See the
RELOCATING CONTROLLERS section if you change this.
CryptoType
The cryptographic signature tool to be used in the creation of job step credentials. The
slurmctld daemon must be restarted for a change in CryptoType to take effect. Acceptable values
at present include "crypto/munge" and "crypto/openssl". The default value is "crypto/munge".
DebugFlags
Defines specific subsystems which should provide more detailed event logging. Multiple subsystems
can be specified with comma separators. Valid subsystems available today (with more to come)
include:
Backfill Backfill scheduler details
BGBlockAlgo BlueGene block selection details
BGBlockAlgoDeep BlueGene block selection, more details
BGBlockPick BlueGene block selection for jobs
BGBlockWires BlueGene block wiring (switch state details)
CPU_Bind CPU binding details for jobs and steps
Energy AcctGatherEnergy debug info
ExtSensors External Sensors debug info
FrontEnd Front end node details
Gres Generic resource details
Gang Gang scheduling details
NO_CONF_HASH Do not log when the slurm.conf files differs between SLURM daemons
Priority Job prioritization
Reservation Advanced reservations
SelectType Resource selection plugin
Steps Slurmctld resource allocation for job steps
Switch Switch plugin
Triggers Slurmctld triggers
Wiki Sched/wiki and wiki2 communications
ThreadID Prints the thread ID in the log file
DefMemPerCPU
Default real memory size available per allocated CPU in MegaBytes. Used to avoid over-subscribing
memory and causing paging. DefMemPerCPU would generally be used if individual processors are
allocated to jobs (SelectType=select/cons_res). The default value is 0 (unlimited). Also see
DefMemPerNode and MaxMemPerCPU. DefMemPerCPU and DefMemPerNode are mutually exclusive. NOTE:
Enforcement of memory limits currently requires enabling of accounting, which samples memory use
on a periodic basis (data need not be stored, just collected).
DefMemPerNode
Default real memory size available per allocated node in MegaBytes. Used to avoid
over-subscribing memory and causing paging. DefMemPerNode would generally be used if whole nodes
are allocated to jobs (SelectType=select/linear) and resources are shared (Shared=yes or
Shared=force). The default value is 0 (unlimited). Also see DefMemPerCPU and MaxMemPerNode.
DefMemPerCPU and DefMemPerNode are mutually exclusive. NOTE: Enforcement of memory limits
currently requires enabling of accounting, which samples memory use on a periodic basis (data need
not be stored, just collected).
DefaultStorageHost
The default name of the machine hosting the accounting storage and job completion databases. Only
used for database type storage plugins and when the AccountingStorageHost and JobCompHost have not
been defined.
DefaultStorageLoc
The fully qualified file name where accounting records and/or job completion records are written
when the DefaultStorageType is "filetxt" or the name of the database where accounting records
and/or job completion records are stored when the DefaultStorageType is a database. Also see
AccountingStorageLoc and JobCompLoc.
DefaultStoragePass
The password used to gain access to the database to store the accounting and job completion data.
Only used for database type storage plugins, ignored otherwise. Also see AccountingStoragePass
and JobCompPass.
DefaultStoragePort
The listening port of the accounting storage and/or job completion database server. Only used for
database type storage plugins, ignored otherwise. Also see AccountingStoragePort and JobCompPort.
DefaultStorageType
The accounting and job completion storage mechanism type. Acceptable values at present include
"filetxt", "mysql", "none", "pgsql", and "slurmdbd". The value "filetxt" indicates that records
will be written to a file. The value "mysql" indicates that accounting records will be written to
a mysql database. The default value is "none", which means that records are not maintained. The
value "pgsql" indicates that records will be written to a PostgreSQL database. The value
"slurmdbd" indicates that records will be written to the SLURM DBD, which maintains its own
database. See "man slurmdbd" for more information. Also see AccountingStorageType and
JobCompType.
DefaultStorageUser
The user account for accessing the accounting storage and/or job completion database. Only used
for database type storage plugins, ignored otherwise. Also see AccountingStorageUser and
JobCompUser.
DisableRootJobs
If set to "YES" then user root will be prevented from running any jobs. The default value is
"NO", meaning user root will be able to execute jobs. DisableRootJobs may also be set by
partition.
DynallocPort
Socket port used for MapReduce dynamic allocation communications. Used only by the
slurmctld/dynalloc plugin.
EnforcePartLimits
If set to "YES" then jobs which exceed a partition's size and/or time limits will be rejected at
submission time. If set to "NO" then the job will be accepted and remain queued until the
partition limits are altered. The default value is "NO". NOTE: If set, then a job's QOS can not
be used to exceed partition limits.
Epilog Fully qualified pathname of a script to execute as user root on every node when a user's job
completes (e.g. "/usr/local/slurm/epilog"). A glob pattern (See glob(7)) may also be used to run
more than one epilog script (e.g. "/etc/slurm/epilog.d/*"). The Epilog script or scripts may be
used to purge files, disable user login, etc. By default there is no epilog. See Prolog and
Epilog Scripts for more information.
EpilogMsgTime
The number of microseconds that the slurmctld daemon requires to process an epilog completion
message from the slurmd dameons. This parameter can be used to prevent a burst of epilog
completion messages from being sent at the same time which should help prevent lost messages and
improve throughput for large jobs. The default value is 2000 microseconds. For a 1000 node job,
this spreads the epilog completion messages out over two seconds.
EpilogSlurmctld
Fully qualified pathname of a program for the slurmctld to execute upon termination of a job
allocation (e.g. "/usr/local/slurm/epilog_controller"). The program executes as SlurmUser, which
gives it permission to drain nodes and requeue the job if a failure occurs or cancel the job if
appropriate. Exactly what the program does and how it accomplishes this is completely at the
discretion of the system administrator. Information about the job being initiated, it's allocated
nodes, etc. are passed to the program using environment variables. See Prolog and Epilog Scripts
for more information.
ExtSensorsFreq
The external sensors plugin sampling interval. If ExtSensorsType=ext_sensors/none, this parameter
is ignored. For all other values of ExtSensorsType, this parameter is the number of seconds
between external sensors samples for hardware components (nodes, switches, etc.) The default value
is zero. This value disables external sensors sampling. Note: This parameter does not affect
external sensors data collection for jobs/steps.
ExtSensorsType
Identifies the plugin to be used for external sensors data collection. Slurmctld calls this
plugin to collect external sensors data for jobs/steps and hardware components. In case of node
sharing between jobs the reported values per job/step (through sstat or sacct) may not be
accurate. See also "man ext_sensors.conf".
Configurable values at present are:
ext_sensors/none No external sensors data is collected.
ext_sensors/rrd External sensors data is collected from the RRD database.
FastSchedule
Controls how a node's configuration specifications in slurm.conf are used. If the number of node
configuration entries in the configuration file is significantly lower than the number of nodes,
setting FastSchedule to 1 will permit much faster scheduling decisions to be made. (The scheduler
can just check the values in a few configuration records instead of possibly thousands of node
records.) Note that on systems with hyper-threading, the processor count reported by the node
will be twice the actual processor count. Consider which value you want to be used for scheduling
purposes.
1 (default)
Consider the configuration of each node to be that specified in the slurm.conf configuration
file and any node with less than the configured resources will be set DOWN.
0 Base scheduling decisions upon the actual configuration of each individual node except that
the node's processor count in SLURM's configuration must match the actual hardware
configuration if SchedulerType=sched/gang or SelectType=select/cons_res are configured (both
of those plugins maintain resource allocation information using bitmaps for the cores in the
system and must remain static, while the node's memory and disk space can be established
later).
2 Consider the configuration of each node to be that specified in the slurm.conf configuration
file and any node with less than the configured resources will not be set DOWN. This can be
useful for testing purposes.
FirstJobId
The job id to be used for the first submitted to SLURM without a specific requested value. Job id
values generated will incremented by 1 for each subsequent job. This may be used to provide a
meta-scheduler with a job id space which is disjoint from the interactive jobs. The default value
is 1. Also see MaxJobId
GetEnvTimeout
Used for Moab scheduled jobs only. Controls how long job should wait in seconds for loading the
user's environment before attempting to load it from a cache file. Applies when the srun or sbatch
--get-user-env option is used. If set to 0 then always load the user's environment from the cache
file. The default value is 2 seconds.
GresTypes
A comma delimited list of generic resources to be managed. These generic resources may have an
associated plugin available to provide additional functionality. No generic resources are managed
by default. Insure this parameter is consistent across all nodes in the cluster for proper
operation. The slurmctld daemon must be restarted for changes to this parameter to become
effective.
GroupUpdateForce
If set to a non-zero value, then information about which users are members of groups allowed to
use a partition will be updated periodically, even when there have been no changes to the
/etc/group file. Otherwise group member information will be updated periodically only after the
/etc/group file is updated The default value is 0. Also see the GroupUpdateTime parameter.
GroupUpdateTime
Controls how frequently information about which users are members of groups allowed to use a
partition will be updated. The time interval is given in seconds with a default value of 600
seconds and a maximum value of 4095 seconds. A value of zero will prevent periodic updating of
group membership information. Also see the GroupUpdateForce parameter.
HealthCheckInterval
The interval in seconds between executions of HealthCheckProgram. The default value is zero,
which disables execution.
HealthCheckNodeState
Identify what node states should execute the HealthCheckProgram. Multiple state values may be
specified with a comma separator. The default value is ANY to execute on nodes in any state.
ALLOC Run on nodes in the ALLOC state (all CPUs allocated).
ANY Run on nodes in any state.
IDLE Run on nodes in the IDLE state.
MIXED Run on nodes in the MIXED state (some CPUs idle and other CPUs allocated).
HealthCheckProgram
Fully qualified pathname of a script to execute as user root periodically on all compute nodes
that are not in the NOT_RESPONDING state. This program may be used to verify the node is fully
operational and DRAIN the node or send email if a problem is detected. Any action to be taken
must be explicitly performed by the program (e.g. execute "scontrol update NodeName=foo
State=drain Reason=tmp_file_system_full" to drain a node). The execution interval is controlled
using the HealthCheckInterval parameter. Note that the HealthCheckProgram will be executed at the
same time on all nodes to minimize its impact upon parallel programs. This program is will be
killed if it does not terminate normally within 60 seconds. By default, no program will be
executed.
InactiveLimit
The interval, in seconds, after which a non-responsive job allocation command (e.g. srun or
salloc) will result in the job being terminated. If the node on which the command is executed
fails or the command abnormally terminates, this will terminate its job allocation. This option
has no effect upon batch jobs. When setting a value, take into consideration that a debugger
using srun to launch an application may leave the srun command in a stopped state for extended
periods of time. This limit is ignored for jobs running in partitions with the RootOnly flag set
(the scheduler running as root will be responsible for the job). The default value is unlimited
(zero) and may not exceed 65533 seconds.
JobAcctGatherType
The job accounting mechanism type. Acceptable values at present include "jobacct_gather/aix" (for
AIX operating system), "jobacct_gather/linux" (for Linux operating system),
"jobacct_gather/cgroup" (experimental) and "jobacct_gather/none" (no accounting data collected).
The default value is "jobacct_gather/none". "jobacct_gather/cgroup" is an experimental plugin for
the Linux operating system that uses cgroups to collect accounting statistics. The plugin collects
the following statistics: From the cgroup memory subsystem: memory.usage_in_bytes (reported as
'pages') and rss from memory.stat (reported as 'rss'). From the cgroup cpuacct subsystem: user cpu
time and system cpu time. No value is provided by cgroups for virtual memory size ('vsize'). In
order to use the sstat tool, "jobacct_gather/aix", "jobacct_gather/linux", or
"jobacct_gather/cgroup" must be configured.
NOTE: Changing this configuration parameter changes the contents of the messages between Slurm
daemons. Any previously running job steps are managed by a slurmstepd daemon that will persist
through the lifetime of that job step and not change it's communication prototol. Only change this
configuration parameter when there are no running job steps.
JobAcctGatherFrequency
The job accounting and profiling sampling intervals. The supported format is follows:
JobAcctGatherFrequency=<datatype>=<interval>
where <datatype>=<interval> specifies the task sampling interval for the
jobacct_gather plugin or a sampling interval for a profiling type by the
acct_gather_profile plugin. Multiple, comma-separated <datatype>=<interval> intervals
may be specified. Supported datatypes are as follows:
task=<interval>
where <interval> is the task sampling interval in seconds for the
jobacct_gather plugins and for task profiling by the acct_gather_profile
plugin.
energy=<interval>
where <interval> is the sampling interval in seconds for energy profiling using
the acct_gather_energy plugin
network=<interval>
where <interval> is the sampling interval in seconds for infiniband profiling
using the acct_gather_infiniband plugin.
filesystem=<interval>
where <interval> is the sampling interval in seconds for filesystem profiling
using the acct_gather_filesystem plugin.
The default value for task sampling interval
is 30 seconds. The default value for all other intervals is 0. An interval of 0 disables sampling
of the specified type. If the task sampling interval is 0, accounting information is collected
only at job termination (reducing SLURM interference with the job).
Smaller (non-zero) values have a greater impact upon job performance, but a value of 30 seconds is
not likely to be noticeable for applications having less than 10,000 tasks.
Users can independently override each interval on a per job basis using the --acctg-freq option
when submitting the job.
JobCheckpointDir
Specifies the default directory for storing or reading job checkpoint information. The data stored
here is only a few thousand bytes per job and includes information needed to resubmit the job
request, not job's memory image. The directory must be readable and writable by SlurmUser, but not
writable by regular users. The job memory images may be in a different location as specified by
--checkpoint-dir option at job submit time or scontrol's ImageDir option.
JobCompHost
The name of the machine hosting the job completion database. Only used for database type storage
plugins, ignored otherwise. Also see DefaultStorageHost.
JobCompLoc
The fully qualified file name where job completion records are written when the JobCompType is
"jobcomp/filetxt" or the database where job completion records are stored when the JobCompType is
a database. Also see DefaultStorageLoc.
JobCompPass
The password used to gain access to the database to store the job completion data. Only used for
database type storage plugins, ignored otherwise. Also see DefaultStoragePass.
JobCompPort
The listening port of the job completion database server. Only used for database type storage
plugins, ignored otherwise. Also see DefaultStoragePort.
JobCompType
The job completion logging mechanism type. Acceptable values at present include "jobcomp/none",
"jobcomp/filetxt", "jobcomp/mysql", "jobcomp/pgsql", and "jobcomp/script"". The default value is
"jobcomp/none", which means that upon job completion the record of the job is purged from the
system. If using the accounting infrastructure this plugin may not be of interest since the
information here is redundant. The value "jobcomp/filetxt" indicates that a record of the job
should be written to a text file specified by the JobCompLoc parameter. The value "jobcomp/mysql"
indicates that a record of the job should be written to a mysql database specified by the
JobCompLoc parameter. The value "jobcomp/pgsql" indicates that a record of the job should be
written to a PostgreSQL database specified by the JobCompLoc parameter. The value
"jobcomp/script" indicates that a script specified by the JobCompLoc parameter is to be executed
with environment variables indicating the job information.
JobCompUser
The user account for accessing the job completion database. Only used for database type storage
plugins, ignored otherwise. Also see DefaultStorageUser.
JobCredentialPrivateKey
Fully qualified pathname of a file containing a private key used for authentication by SLURM
daemons. This parameter is ignored if CryptoType=crypto/munge.
JobCredentialPublicCertificate
Fully qualified pathname of a file containing a public key used for authentication by SLURM
daemons. This parameter is ignored if CryptoType=crypto/munge.
JobFileAppend
This option controls what to do if a job's output or error file exist when the job is started. If
JobFileAppend is set to a value of 1, then append to the existing file. By default, any existing
file is truncated.
JobRequeue
This option controls what to do by default after a node failure. If JobRequeue is set to a value
of 1, then any batch job running on the failed node will be requeued for execution on different
nodes. If JobRequeue is set to a value of 0, then any job running on the failed node will be
terminated. Use the sbatch --no-requeue or --requeue option to change the default behavior for
individual jobs. The default value is 1.
JobSubmitPlugins
A comma delimited list of job submission plugins to be used. The specified plugins will be
executed in the order listed. These are intended to be site-specific plugins which can be used to
set default job parameters and/or logging events. Sample plugins available in the distribution
include "all_partitions", "cnode", "defaults", "logging", "lua", and "partition". For examples of
use, see the SLURM code in "src/plugins/job_submit" and "contribs/lua/job_submit*.lua" then modify
the code to satisfy your needs. SLURM can be configured to use multiple job_submit plugins if
desired, however the lua plugin will only execute one lua script named "job_submit.lua" and
located in default script directory (typically the subdirectory "etc" of the installation
directory). No job submission plugins are used by default.
KeepAliveTime
Specifies how long sockets communications used between the srun command and its slurmstepd process
are kept alive after disconnect. Longer values can be used to improve reliability of
communications in the event of network failures. The default value leaves the system default
value. The value may not exceed 65533.
KillOnBadExit
If set to 1, the job will be terminated immediately when one of the processes is crashed or
aborted. With the default value of 0, if one of the processes is crashed or aborted the other
processes will continue to run. The user can override this configuration parameter by using srun's
-K, --kill-on-bad-exit.
KillWait
The interval, in seconds, given to a job's processes between the SIGTERM and SIGKILL signals upon
reaching its time limit. If the job fails to terminate gracefully in the interval specified, it
will be forcibly terminated. The default value is 30 seconds. The value may not exceed 65533.
LaunchType
Identifies the mechanism to be used to launch application tasks. Acceptable values include
"launch/aprun" for use with Cray systems with ALPS, "launch/poe" for use with IBM Parallel
Environment (PE), "launch/runjob" for use with IBM BlueGene/Q systems, and "launch/slurm" for all
other systems. The default value is "launch/aprun" for Cray systems, "launch/poe" for systems
with the IBM NRT library installed, "launch/runjob" for IBM BlueGene/Q systems, and "launch/slurm"
for all other systems.
Licenses
Specification of licenses (or other resources available on all nodes of the cluster) which can be
allocated to jobs. License names can optionally be followed by a colon and count with a default
count of one. Multiple license names should be comma separated (e.g. "Licenses=foo:4,bar").
Note that SLURM prevents jobs from being scheduled if their required license specification is not
available. SLURM does not prevent jobs from using licenses that are not explicitly listed in the
job submission specification.
MailProg
Fully qualified pathname to the program used to send email per user request. The default value is
"/usr/bin/mail".
MaxArraySize
The maximum job array size. The maximum job array task index value will be one less than
MaxArraySize to allow for an index value of zero. Configure MaxArraySize to 0 in order to disable
job array use. The value may not exceed 65533. Default value is 1001.
MaxJobCount
The maximum number of jobs SLURM can have in its active database at one time. Set the values of
MaxJobCount and MinJobAge to insure the slurmctld daemon does not exhaust its memory or other
resources. Once this limit is reached, requests to submit additional jobs will fail. The default
value is 10000 jobs. Performance can suffer with more than a couple hundred thousand jobs.
Setting per MaxSubmitJobs per user is generally valuable to prevent a single user from filling the
system with jobs. This is accomplished using Slurm's database and configuring enforcement of
resource limits. This value may not be reset via "scontrol reconfig". It only takes effect upon
restart of the slurmctld daemon.
MaxJobId
The maximum job id to be used for jobs submitted to SLURM without a specific requested value. Job
id values generated will incremented by 1 for each subsequent job. This may be used to provide a
meta-scheduler with a job id space which is disjoint from the interactive jobs. Once MaxJobId is
reached, the next job will be assigned FirstJobId. The default value is 4294901760 (0xffff0000).
Also see FirstJobId.
MaxMemPerCPU
Maximum real memory size available per allocated CPU in MegaBytes. Used to avoid over-subscribing
memory and causing paging. MaxMemPerCPU would generally be used if individual processors are
allocated to jobs (SelectType=select/cons_res). The default value is 0 (unlimited). Also see
DefMemPerCPU and MaxMemPerNode. MaxMemPerCPU and MaxMemPerNode are mutually exclusive. NOTE:
Enforcement of memory limits currently requires enabling of accounting, which samples memory use
on a periodic basis (data need not be stored, just collected).
MaxMemPerNode
Maximum real memory size available per allocated node in MegaBytes. Used to avoid
over-subscribing memory and causing paging. MaxMemPerNode would generally be used if whole nodes
are allocated to jobs (SelectType=select/linear) and resources are shared (Shared=yes or
Shared=force). The default value is 0 (unlimited). Also see DefMemPerNode and MaxMemPerCPU.
MaxMemPerCPU and MaxMemPerNode are mutually exclusive. NOTE: Enforcement of memory limits
currently requires enabling of accounting, which samples memory use on a periodic basis (data need
not be stored, just collected).
MaxStepCount
The maximum number of steps that any job can initiate. This parameter is intended to limit the
effect of bad batch scripts. The default value is 40000 steps.
MaxTasksPerNode
Maximum number of tasks SLURM will allow a job step to spawn on a single node. The default
MaxTasksPerNode is 128. May not exceed 65533.
MessageTimeout
Time permitted for a round-trip communication to complete in seconds. Default value is 10 seconds.
For systems with shared nodes, the slurmd daemon could be paged out and necessitate higher values.
MinJobAge
The minimum age of a completed job before its record is purged from SLURM's active database. Set
the values of MaxJobCount and
to insure the slurmctld daemon does not exhaust its memory or other resources. The default value
is 300 seconds. A value of zero prevents any job record purging. In order to eliminate some
possible race conditions, the minimum non-zero value for MinJobAge recommended is 2. May not
exceed 65533.
MpiDefault
Identifies the default type of MPI to be used. Srun may override this configuration parameter in
any case. Currently supported versions include: lam, mpich1_p4, mpich1_shmem, mpichgm, mpichmx,
mvapich, none (default, which works for many other versions of MPI) and openmpi. pmi2, More
information about MPI use is available here <http://slurm.schedmd.com/mpi_guide.html>.
MpiParams
MPI parameters. Used to identify ports used by OpenMPI only and the input format is
"ports=12000-12999" to identify a range of communication ports to be used.
OverTimeLimit
Number of minutes by which a job can exceed its time limit before being canceled. The configured
job time limit is treated as a soft limit. Adding OverTimeLimit to the soft limit provides a hard
limit, at which point the job is canceled. This is particularly useful for backfill scheduling,
which bases upon each job's soft time limit. The default value is zero. May not exceed exceed
65533 minutes. A value of "UNLIMITED" is also supported.
PluginDir
Identifies the places in which to look for SLURM plugins. This is a colon-separated list of
directories, like the PATH environment variable. The default value is "/usr/local/lib/slurm".
PlugStackConfig
Location of the config file for SLURM stackable plugins that use the Stackable Plugin Architecture
for Node job (K)control (SPANK). This provides support for a highly configurable set of plugins
to be called before and/or after execution of each task spawned as part of a user's job step.
Default location is "plugstack.conf" in the same directory as the system slurm.conf. For more
information on SPANK plugins, see the spank(8) manual.
PreemptMode
Enables gang scheduling and/or controls the mechanism used to preempt jobs. When the PreemptType
parameter is set to enable preemption, the PreemptMode selects the default mechanism used to
preempt the lower priority jobs for the cluster. PreemptMode may be specified on a per partition
basis to over ride this default value if PreemptType=preempt/partition_prio, but a valid default
PreemptMode value must be specified for the cluster as a whole when preemption is enabled. The
GANG option is used to enable gang scheduling independent of whether preemption is enabled (the
PreemptType setting). The GANG option can be specified in addition to a PreemptMode setting with
the two options comma separated. The SUSPEND option requires that gang scheduling be enabled
(i.e, "PreemptMode=SUSPEND,GANG").
OFF is the default value and disables job preemption and gang scheduling. This is the
only option compatible with SchedulerType=sched/wiki or SchedulerType=sched/wiki2
(used by Maui and Moab respectively, which provide their own job preemption
functionality).
CANCEL always cancel the job.
CHECKPOINT preempts jobs by checkpointing them (if possible) or canceling them.
GANG enables gang scheduling (time slicing) of jobs in the same partition. NOTE: Gang
scheduling is performed independently for each partition, so configuring partitions
with overlapping nodes and gang scheduling is generally not recommended.
REQUEUE preempts jobs by requeuing them (if possible) or canceling them.
SUSPEND preempts jobs by suspending them. A suspended job will resume execution once the high
priority job preempting it completes. The SUSPEND may only be used with the GANG
option (the gang scheduler module performs the job resume operation) and with
PreemptType=preempt/partition_prio (the logic to suspend and resume jobs current only
has the data structures to support partitions).
PreemptType
This specifies the plugin used to identify which jobs can be preempted in order to start a pending
job.
preempt/none
Job preemption is disabled. This is the default.
preempt/partition_prio
Job preemption is based upon partition priority. Jobs in higher priority partitions
(queues) may preempt jobs from lower priority partitions. This is not compatible with
PreemptMode=OFF.
preempt/qos
Job preemption rules are specified by Quality Of Service (QOS) specifications in the SLURM
database a database. This is not compatible with PreemptMode=OFF or PreemptMode=SUSPEND
(i.e. preempted jobs must be removed from the resources).
PriorityDecayHalfLife
This controls how long prior resource use is considered in determining how over- or under-serviced
an association is (user, bank account and cluster) in determining job priority. If set to 0 no
decay will be applied. This is helpful if you want to enforce hard time limits per association.
If set to 0 PriorityUsageResetPeriod must be set to some interval. Applicable only if
PriorityType=priority/multifactor. The unit is a time string (i.e. min, hr:min:00,
days-hr:min:00, or days-hr). The default value is 7-0 (7 days).
PriorityCalcPeriod
The period of time in minutes in which the half-life decay will be re-calculated. Applicable only
if PriorityType=priority/multifactor. The default value is 5 (minutes).
PriorityFavorSmall
Specifies that small jobs should be given preferential scheduling priority. Applicable only if
PriorityType=priority/multifactor. Supported values are "YES" and "NO". The default value is
"NO".
PriorityFlags
Flags to modify priority behavior Applicable only if PriorityType=priority/multifactor. The
keywords below have no associated value (e.g.
"PriorityFlags=ACCRUE_ALWAYS,SMALL_RELATIVE_TO_TIME").
ACCRUE_ALWAYS If set, priority age factor will be increased despite job dependencies or holds.
SMALL_RELATIVE_TO_TIME
If set, the job's size component will be based upon not the job size alone, but
the job's size divided by it's time limit.
TICKET_BASED If set, priority will be calculated based on the ticket system.
PriorityMaxAge
Specifies the job age which will be given the maximum age factor in computing priority. For
example, a value of 30 minutes would result in all jobs over 30 minutes old would get the same
age-based priority. Applicable only if PriorityType=priority/multifactor. The unit is a time
string (i.e. min, hr:min:00, days-hr:min:00, or days-hr). The default value is 7-0 (7 days).
PriorityUsageResetPeriod
At this interval the usage of associations will be reset to 0. This is used if you want to
enforce hard limits of time usage per association. If PriorityDecayHalfLife is set to be 0 no
decay will happen and this is the only way to reset the usage accumulated by running jobs. By
default this is turned off and it is advised to use the PriorityDecayHalfLife option to avoid not
having anything running on your cluster, but if your schema is set up to only allow certain
amounts of time on your system this is the way to do it. Applicable only if
PriorityType=priority/multifactor.
NONE Never clear historic usage. The default value.
NOW Clear the historic usage now. Executed at startup and reconfiguration time.
DAILY Cleared every day at midnight.
WEEKLY Cleared every week on Sunday at time 00:00.
MONTHLY Cleared on the first day of each month at time 00:00.
QUARTERLY Cleared on the first day of each quarter at time 00:00.
YEARLY Cleared on the first day of each year at time 00:00.
PriorityType
This specifies the plugin to be used in establishing a job's scheduling priority. Supported values
are "priority/basic" (jobs are prioritized by order of arrival, also suitable for sched/wiki and
sched/wiki2), "priority/multifactor" (jobs are prioritized based upon size, age, fair-share of
allocation, etc). Also see PriorityFlags for configuration options. The default value is
"priority/basic".
PriorityWeightAge
An integer value that sets the degree to which the queue wait time component contributes to the
job's priority. Applicable only if PriorityType=priority/multifactor. The default value is 0.
PriorityWeightFairshare
An integer value that sets the degree to which the fair-share component contributes to the job's
priority. Applicable only if PriorityType=priority/multifactor. The default value is 0.
PriorityWeightJobSize
An integer value that sets the degree to which the job size component contributes to the job's
priority. Applicable only if PriorityType=priority/multifactor. The default value is 0.
PriorityWeightPartition
An integer value that sets the degree to which the node partition component contributes to the
job's priority. Applicable only if PriorityType=priority/multifactor. The default value is 0.
PriorityWeightQOS
An integer value that sets the degree to which the Quality Of Service component contributes to the
job's priority. Applicable only if PriorityType=priority/multifactor. The default value is 0.
PrivateData
This controls what type of information is hidden from regular users. By default, all information
is visible to all users. User SlurmUser and root can always view all information. Multiple
values may be specified with a comma separator. Acceptable values include:
accounts
(NON-SLURMDBD ACCOUNTING ONLY) Prevents users from viewing any account definitions unless
they are coordinators of them.
jobs Prevents users from viewing jobs or job steps belonging to other users. (NON-SLURMDBD
ACCOUNTING ONLY) Prevents users from viewing job records belonging to other users unless
they are coordinators of the association running the job when using sacct.
nodes Prevents users from viewing node state information.
partitions
Prevents users from viewing partition state information.
reservations
Prevents regular users from viewing reservations.
usage Prevents users from viewing usage of any other user, this applies to sshare. (NON-SLURMDBD
ACCOUNTING ONLY) Prevents users from viewing usage of any other user, this applies to
sreport.
users (NON-SLURMDBD ACCOUNTING ONLY) Prevents users from viewing information of any user other
than themselves, this also makes it so users can only see associations they deal with.
Coordinators can see associations of all users they are coordinator of, but can only see
themselves when listing users.
ProctrackType
Identifies the plugin to be used for process tracking. The slurmd daemon uses this mechanism to
identify all processes which are children of processes it spawns for a user job. The slurmd
daemon must be restarted for a change in ProctrackType to take effect. NOTE:
"proctrack/linuxproc" and "proctrack/pgid" can fail to identify all processes associated with a
job since processes can become a child of the init process (when the parent process terminates) or
change their process group. To reliably track all processes, one of the other mechanisms
utilizing kernel modifications is preferable. NOTE: "proctrack/linuxproc" is not compatible with
"switch/elan." Acceptable values at present include:
proctrack/aix which uses an AIX kernel extension and is the default for AIX systems
proctrack/cgroup which uses linux cgroups to constrain and track processes. NOTE: see "man
cgroup.conf" for configuration details NOTE: This plugin writes to disk often
and can impact performance. If you are running lots of short running jobs
(less than a couple of seconds) this plugin slows down performance
dramatically. It should probably be avoided in an HTC environment.
proctrack/linuxproc which uses linux process tree using parent process IDs
proctrack/lua which uses a site-specific LUA script to track processes
proctrack/sgi_job which uses SGI's Process Aggregates (PAGG) kernel module, see
http://oss.sgi.com/projects/pagg/ for more information
proctrack/pgid which uses process group IDs and is the default for all other systems
Prolog Fully qualified pathname of a program for the slurmd to execute whenever it is asked to run a job
step from a new job allocation (e.g. "/usr/local/slurm/prolog"). A glob pattern (See glob(7))
may also be used to specify more than one program to run (e.g. "/etc/slurm/prolog.d/*"). The
slurmd executes the prolog before starting the first job step. The prolog script or scripts may
be used to purge files, enable user login, etc. By default there is no prolog. Any configured
script is expected to complete execution quickly (in less time than MessageTimeout). If the
prolog fails (returns a non-zero exit code), this will result in the node being set to a DOWN
state and the job requeued to executed on another node. See Prolog and Epilog Scripts for more
information.
PrologSlurmctld
Fully qualified pathname of a program for the slurmctld daemon to execute before granting a new
job allocation (e.g. "/usr/local/slurm/prolog_controller"). The program executes as SlurmUser on
the same node where the slurmctld daemon executes, giving it permission to drain nodes and requeue
the job if a failure occurs or cancel the job if appropriate. The program can be used to reboot
nodes or perform other work to prepare resources for use. Exactly what the program does and how
it accomplishes this is completely at the discretion of the system administrator. Information
about the job being initiated, it's allocated nodes, etc. are passed to the program using
environment variables. While this program is running, the nodes associated with the job will be
have a POWER_UP/CONFIGURING flag set in their state, which can be readily viewed. The slurmctld
daemon will wait indefinitely for this program to complete. Once the program completes with an
exit code of zero, the nodes will be considered ready for use and the program will be started. If
some node can not be made available for use, the program should drain the node (typically using
the scontrol command) and terminate with a non-zero exit code. A non-zero exit code will result
in the job being requeued (where possible) or killed. Note that only batch jobs can be requeued.
See Prolog and Epilog Scripts for more information.
PropagatePrioProcess
Controls the scheduling priority (nice value) of user spawned tasks.
0 The tasks will inherit the scheduling priority from the slurm daemon. This is the default
value.
1 The tasks will inherit the scheduling priority of the command used to submit them (e.g. srun
or sbatch). Unless the job is submitted by user root, the tasks will have a scheduling
priority no higher than the slurm daemon spawning them.
2 The tasks will inherit the scheduling priority of the command used to submit them (e.g. srun
or sbatch) with the restriction that their nice value will always be one higher than the
slurm daemon (i.e. the tasks scheduling priority will be lower than the slurm daemon).
PropagateResourceLimits
A list of comma separated resource limit names. The slurmd daemon uses these names to obtain the
associated (soft) limit values from the users process environment on the submit node. These
limits are then propagated and applied to the jobs that will run on the compute nodes. This
parameter can be useful when system limits vary among nodes. Any resource limits that do not
appear in the list are not propagated. However, the user can override this by specifying which
resource limits to propagate with the srun commands "--propagate" option. If neither of the
'propagate resource limit' parameters are specified, then the default action is to propagate all
limits. Only one of the parameters, either PropagateResourceLimits or
PropagateResourceLimitsExcept, may be specified. The following limit names are supported by SLURM
(although some options may not be supported on some systems):
ALL All limits listed below
NONE No limits listed below
AS The maximum address space for a process
CORE The maximum size of core file
CPU The maximum amount of CPU time
DATA The maximum size of a process's data segment
FSIZE The maximum size of files created. Note that if the user sets FSIZE to less than the
current size of the slurmd.log, job launches will fail with a 'File size limit exceeded'
error.
MEMLOCK The maximum size that may be locked into memory
NOFILE The maximum number of open files
NPROC The maximum number of processes available
RSS The maximum resident set size
STACK The maximum stack size
PropagateResourceLimitsExcept
A list of comma separated resource limit names. By default, all resource limits will be
propagated, (as described by the PropagateResourceLimits parameter), except for the limits
appearing in this list. The user can override this by specifying which resource limits to
propagate with the srun commands "--propagate" option. See PropagateResourceLimits above for a
list of valid limit names.
RebootProgram
Program to be executed on each compute node to reboot it. Invoked on each node once it becomes
idle after the command "scontrol reboot_nodes" is executed by an authorized user. After being
rebooting, the node is returned to normal use.
ReconfigFlags
Flags to control various actions that may be taken when an "scontrol reconfig" command is issued.
Currently the only option defined is:
KeepPartInfo If set, an "scontrol reconfig" command will maintain the in-memory value of
partition "state" and other parameters that may have been dynamically updated by
"scontrol update". Partition information in the slurm.conf file will be merged
with in-memory data. This flag supersedes the KeepPartState flag.
KeepPartState If set, an "scontrol reconfig" command will preserve only the current "state"
value of in-memory partitions and will reset all other parameters of the
partitions that may have been dynamically updated by "scontrol update" to the
values from the slurm.conf file. Partition information in the slurm.conf file
will be merged with in-memory data.
The default for the above flags is not set, and the "scontrol reconfig" will rebuild the partition
information using only the definitions in the slurm.conf file.
ResumeProgram
SLURM supports a mechanism to reduce power consumption on nodes that remain idle for an extended
period of time. This is typically accomplished by reducing voltage and frequency or powering the
node down. ResumeProgram is the program that will be executed when a node in power save mode is
assigned work to perform. For reasons of reliability, ResumeProgram may execute more than once
for a node when the slurmctld daemon crashes and is restarted. If ResumeProgram is unable to
restore a node to service, it should requeue any node associated with the node and set the node
state to DRAIN. The program executes as SlurmUser. The argument to the program will be the names
of nodes to be removed from power savings mode (using SLURM's hostlist expression format). By
default no program is run. Related configuration options include ResumeTimeout, ResumeRate,
SuspendRate, SuspendTime, SuspendTimeout, SuspendProgram, SuspendExcNodes, and SuspendExcParts.
More information is available at the SLURM web site ( http://slurm.schedmd.com/power_save.html ).
ResumeRate
The rate at which nodes in power save mode are returned to normal operation by ResumeProgram. The
value is number of nodes per minute and it can be used to prevent power surges if a large number
of nodes in power save mode are assigned work at the same time (e.g. a large job starts). A value
of zero results in no limits being imposed. The default value is 300 nodes per minute. Related
configuration options include ResumeTimeout, ResumeProgram, SuspendRate, SuspendTime,
SuspendTimeout, SuspendProgram, SuspendExcNodes, and SuspendExcParts.
ResumeTimeout
Maximum time permitted (in second) between when a node is resume request is issued and when the
node is actually available for use. Nodes which fail to respond in this time frame may be marked
DOWN and the jobs scheduled on the node requeued. The default value is 60 seconds. Related
configuration options include ResumeProgram, ResumeRate, SuspendRate, SuspendTime, SuspendTimeout,
SuspendProgram, SuspendExcNodes and SuspendExcParts. More information is available at the SLURM
web site ( http://slurm.schedmd.com/power_save.html ).
ResvEpilog
Fully qualified pathname of a program for the slurmctld to execute when a reservation ends. The
program can be used to cancel jobs, modify partition configuration, etc. The reservation named
will be passed as an argument to the program. By default there is no epilog.
ResvOverRun
Describes how long a job already running in a reservation should be permitted to execute after the
end time of the reservation has been reached. The time period is specified in minutes and the
default value is 0 (kill the job immediately). The value may not exceed 65533 minutes, although a
value of "UNLIMITED" is supported to permit a job to run indefinitely after its reservation is
terminated.
ResvProlog
Fully qualified pathname of a program for the slurmctld to execute when a reservation begins. The
program can be used to cancel jobs, modify partition configuration, etc. The reservation named
will be passed as an argument to the program. By default there is no prolog.
ReturnToService
Controls when a DOWN node will be returned to service. The default value is 0. Supported values
include
0 A node will remain in the DOWN state until a system administrator explicitly changes its state
(even if the slurmd daemon registers and resumes communications).
1 A DOWN node will become available for use upon registration with a valid configuration only if
it was set DOWN due to being non-responsive. If the node was set DOWN for any other reason
(low memory, prolog failure, epilog failure, unexpected reboot, etc.), its state will not
automatically be changed.
2 A DOWN node will become available for use upon registration with a valid configuration. The
node could have been set DOWN for any reason. (Disabled on Cray systems.)
SallocDefaultCommand
Normally, salloc(1) will run the user's default shell when a command to execute is not specified
on the salloc command line. If SallocDefaultCommand is specified, salloc will instead run the
configured command. The command is passed to '/bin/sh -c', so shell metacharacters are allowed,
and commands with multiple arguments should be quoted. For instance:
SallocDefaultCommand = "$SHELL"
would run the shell in the user's $SHELL environment variable. and
SallocDefaultCommand = "srun -n1 -N1 --mem-per-cpu=0 --pty --preserve-env --mpi=none $SHELL"
would run spawn the user's default shell on the allocated resources, but not consume any of the
CPU or memory resources, configure it as a pseudo-terminal, and preserve all of the job's
environment variables (i.e. and not over-write them with the job step's allocation information).
SchedulerParameters
The interpretation of this parameter varies by SchedulerType. Multiple options may be comma
separated.
default_queue_depth=#
The default number of jobs to attempt scheduling (i.e. the queue depth) when a running job
completes or other routine actions occur. The full queue will be tested on a less frequent
basis. The default value is 100. In the case of large clusters (more than 1000 nodes),
configuring a relatively small value may be desirable. Specifying a large value (say 1000
or higher) can be expected to result in poor system responsiveness since this scheduling
logic will not release locks for other events to occur. It would be better to let the
backfill scheduler process a larger number of jobs (see max_job_bf, bf_continue and other
options here for more information).
defer Setting this option will avoid attempting to schedule each job individually at job submit
time, but defer it until a later time when scheduling multiple jobs simultaneously may be
possible. This option may improve system responsiveness when large numbers of jobs (many
hundreds) are submitted at the same time, but it will delay the initiation time of
individual jobs. Also see default_queue_depth above.
bf_continue
The backfill scheduler periodically releases locks in order to permit other operations to
proceed rather than blocking all activity for what could be an extended period of time.
Setting this option will cause the backfill scheduler to continue processing pending jobs
from its original job list after releasing locks even if job or node state changes. This
can result in lower priority jobs from being backfill scheduled instead of newly arrived
higher priority jobs, but will permit more queued jobs to be considered for backfill
scheduling.
bf_interval=#
The number of seconds between iterations. Higher values result in less overhead and better
responsiveness. The default value is 30 seconds. This option applies only to
SchedulerType=sched/backfill.
bf_max_job_part=#
The maximum number of jobs per partition to attempt backfill scheduling for, not counting
jobs which cannot be started due to an association resource limit. This can be especially
helpful for systems with large numbers of partitions and jobs. The default value is 0,
which means no limit. This option applies only to SchedulerType=sched/backfill.
bf_max_job_user=#
The maximum number of jobs per user to attempt backfill scheduling for, not counting jobs
which cannot be started due to an association resource limit. One can set this limit to
prevent users from flooding the backfill queue with jobs that cannot start and that prevent
jobs from other users to start. This is similar to the MAXIJOB limit in Maui. The default
value is 0, which means no limit. This option applies only to
SchedulerType=sched/backfill.
bf_resolution=#
The number of seconds in the resolution of data maintained about when jobs begin and end.
Higher values result in less overhead and better responsiveness. The default value is 60
seconds. This option applies only to SchedulerType=sched/backfill.
bf_window=#
The number of minutes into the future to look when considering jobs to schedule. Higher
values result in more overhead and less responsiveness. The default value is 1440 minutes
(one day). A value at least as long as the highest allowed time limit is generally
advisable to prevent job starvation. In order limit the amount of data managed by the
backfill scheduler, if the value of bf_window is increased, then it is generally advisable
to also increase bf_resolution. if This option applies only to
SchedulerType=sched/backfill.
max_job_bf=#
The maximum number of jobs to attempt backfill scheduling for (i.e. the queue depth).
Higher values result in more overhead and less responsiveness. Until an attempt is made to
backfill schedule a job, its expected initiation time value will not be set. The default
value is 50. This option applies only to SchedulerType=sched/backfill.
max_depend_depth=#
Maximum number of jobs to test for a circular job dependency. Stop testing after this
number of job dependencies have been tested. The default value is 10 jobs.
max_switch_wait=#
Maximum number of seconds that a job can delay execution waiting for the specified desired
switch count. The default value is 300 seconds.
SchedulerPort
The port number on which slurmctld should listen for connection requests. This value is only used
by the Maui Scheduler (see SchedulerType). The default value is 7321.
SchedulerRootFilter
Identifies whether or not RootOnly partitions should be filtered from any external scheduling
activities. If set to 0, then RootOnly partitions are treated like any other partition. If set to
1, then RootOnly partitions are exempt from any external scheduling activities. The default value
is 1. Currently only used by the built-in backfill scheduling module "sched/backfill" (see
SchedulerType).
SchedulerTimeSlice
Number of seconds in each time slice when gang scheduling is enabled (PreemptMode=GANG). The
value must be between 5 seconds and 65533 seconds. The default value is 30 seconds.
SchedulerType
Identifies the type of scheduler to be used. Note the slurmctld daemon must be restarted for a
change in scheduler type to become effective (reconfiguring a running daemon has no effect for
this parameter). The scontrol command can be used to manually change job priorities if desired.
Acceptable values include:
sched/backfill
For a backfill scheduling module to augment the default FIFO scheduling. Backfill
scheduling will initiate lower-priority jobs if doing so does not delay the expected
initiation time of any higher priority job. Effectiveness of backfill scheduling is
dependent upon users specifying job time limits, otherwise all jobs will have the same time
limit and backfilling is impossible. Note documentation for the SchedulerParameters option
above. This is the default configuration.
sched/builtin
This is the FIFO scheduler which initiates jobs in priority order. If any job in the
partition can not be scheduled, no lower priority job in that partition will be scheduled.
An exception is made for jobs can not run due to partition constraints (e.g. the time
limit) or down/drained nodes. In that case, lower priority jobs can be initiated and not
impact the higher priority job.
sched/gang
Defunct option. See PreemptType and PreemptMode options.
sched/hold
To hold all newly arriving jobs if a file "/etc/slurm.hold" exists otherwise use the
built-in FIFO scheduler
sched/wiki
For the Wiki interface to the Maui Scheduler
sched/wiki2
For the Wiki interface to the Moab Cluster Suite
SelectType
Identifies the type of resource selection algorithm to be used. Changing this value can only be
done by restarting the slurmctld daemon and will result in the loss of all job information
(running and pending) since the job state save format used by each plugin is different.
Acceptable values include
select/linear
for allocation of entire nodes assuming a one-dimensional array of nodes in which
sequentially ordered nodes are preferable. This is the default value for non-BlueGene
systems.
select/cons_res
The resources within a node are individually allocated as consumable resources. Note that
whole nodes can be allocated to jobs for selected partitions by using the Shared=Exclusive
option. See the partition Shared parameter for more information.
select/serial
for allocating resources to single CPU jobs only. Highly optimized for maximum throughput.
NOTE: SPANK environment variables are NOT propagated to the job's Epilog program.
select/bluegene
for a three-dimensional BlueGene system. The default value is "select/bluegene" for
BlueGene systems.
select/cray
for a Cray system. The default value is "select/cray" for all Cray systems.
SelectTypeParameters
The permitted values of SelectTypeParameters depend upon the configured value of SelectType.
SelectType=select/bluegene supports no SelectTypeParameters. The only supported option for
SelectType=select/linear are CR_ONE_TASK_PER_CORE and CR_Memory, which treats memory as a
consumable resource and prevents memory over subscription with job preemption or gang scheduling.
The following values are supported for SelectType=select/cons_res:
CR_ALLOCATE_FULL_SOCKET
Jobs are allocated whole sockets rather than individual cores. Must be used with CR_Socket
or CR_Socket_Memory option.
CR_CPU CPUs are consumable resources. Configure the number of CPUs on each node, which may be
equal to the count of cores or hyper-threads on the node depending upon the desired minimum
resource allocation. The node's Sockets, CoresPerSocket and ThreadsPerCore may also be
configured if desired and may result in job allocations which have improved locality.
CR_CPU_Memory
CPUs and memory are consumable resources. Configure the number of CPUs on each node, which
may be equal to the count of cores or hyper-threads on the node depending upon the desired
minimum resource allocation. The node's Sockets, CoresPerSocket and ThreadsPerCore may also
be configured if desired and may result in job allocations which have improved locality.
Setting a value for DefMemPerCPU is strongly recommended.
CR_Core
Cores are consumable resources. On nodes with hyper-threads, each thread is counted as a
CPU to satisfy a job's resource requirement, but multiple jobs are not allocated threads on
the same core. The count of CPUs allocated to a job may be rounded up to account for every
CPU on an allocated core.
CR_Core_Memory
Cores and memory are consumable resources. On nodes with hyper-threads, each thread is
counted as a CPU to satisfy a job's resource requirement, but multiple jobs are not
allocated threads on the same core. The count of CPUs allocated to a job may be rounded up
to account for every CPU on an allocated core. Setting a value for DefMemPerCPU is
strongly recommended.
CR_ONE_TASK_PER_CORE
Allocate one task per core by default. Without this option, by default one task will be
allocated per thread on nodes with more than one ThreadsPerCore configured.
CR_CORE_DEFAULT_DIST_BLOCK
Allocate cores within a node using block distribution by default. This is a
pseudo-best-fit algorithm that minimizes the number of boards and minimizes the number of
sockets (within minimum boards) used for the allocation. This default behavior can be
overridden specifying a particular "-m" parameter with srun/salloc/sbatch. Without this
option, cores will be allocated cyclicly across the sockets.
CR_Socket
Sockets are consumable resources. On nodes with multiple cores, each core or thread is
counted as a CPU to satisfy a job's resource requirement, but multiple jobs are not
allocated resources on the same socket. The count of CPUs allocated to a job may be
rounded up to account for every CPU on an allocated socket.
CR_Socket_Memory
Memory and sockets are consumable resources. On nodes with multiple cores, each core or
thread is counted as a CPU to satisfy a job's resource requirement, but multiple jobs are
not allocated resources on the same socket. The count of CPUs allocated to a job may be
rounded up to account for every CPU on an allocated socket. Setting a value for
DefMemPerCPU is strongly recommended.
CR_Memory
Memory is a consumable resource. NOTE: This implies Shared=YES or Shared=FORCE for all
partitions. Setting a value for DefMemPerCPU is strongly recommended.
SlurmUser
The name of the user that the slurmctld daemon executes as. For security purposes, a user other
than "root" is recommended. This user must exist on all nodes of the cluster for authentication
of communications between SLURM components. The default value is "root".
SlurmdUser
The name of the user that the slurmd daemon executes as. This user must exist on all nodes of the
cluster for authentication of communications between SLURM components. The default value is
"root".
SlurmctldDebug
The level of detail to provide slurmctld daemon's logs. The default value is info. If the
slurmctld daemon is initiated with -v or --verbose options, that debug level will be preserve or
restored upon reconfiguration.
quiet Log nothing
fatal Log only fatal errors
error Log only errors
info Log errors and general informational messages
verbose Log errors and verbose informational messages
debug Log errors and verbose informational messages and debugging messages
debug2 Log errors and verbose informational messages and more debugging messages
debug3 Log errors and verbose informational messages and even more debugging messages
debug4 Log errors and verbose informational messages and even more debugging messages
debug5 Log errors and verbose informational messages and even more debugging messages
SlurmctldLogFile
Fully qualified pathname of a file into which the slurmctld daemon's logs are written. The
default value is none (performs logging via syslog).
See the section LOGGING if a pathname is specified.
SlurmctldPidFile
Fully qualified pathname of a file into which the slurmctld daemon may write its process id. This
may be used for automated signal processing. The default value is "/var/run/slurmctld.pid".
SlurmctldPlugstack
A comma delimited list of SLURM controller plugins to be started when the daemon begins and
terminated when it ends. Only the plugin's init and fini functions are called.
SlurmctldPort
The port number that the SLURM controller, slurmctld, listens to for work. The default value is
SLURMCTLD_PORT as established at system build time. If none is explicitly specified, it will be
set to 6817. SlurmctldPort may also be configured to support a range of port numbers in order to
accept larger bursts of incoming messages by specifying two numbers separated by a dash (e.g.
SlurmctldPort=6817-6818). NOTE: Either slurmctld and slurmd daemons must not execute on the same
nodes or the values of SlurmctldPort and SlurmdPort must be different.
SlurmctldTimeout
The interval, in seconds, that the backup controller waits for the primary controller to respond
before assuming control. The default value is 120 seconds. May not exceed 65533.
SlurmdDebug
The level of detail to provide slurmd daemon's logs. The default value is info.
quiet Log nothing
fatal Log only fatal errors
error Log only errors
info Log errors and general informational messages
verbose Log errors and verbose informational messages
debug Log errors and verbose informational messages and debugging messages
debug2 Log errors and verbose informational messages and more debugging messages
debug3 Log errors and verbose informational messages and even more debugging messages
debug4 Log errors and verbose informational messages and even more debugging messages
debug5 Log errors and verbose informational messages and even more debugging messages
SlurmdLogFile
Fully qualified pathname of a file into which the slurmd daemon's logs are written. The default
value is none (performs logging via syslog). Any "%h" within the name is replaced with the
hostname on which the slurmd is running. Any "%n" within the name is replaced with the SLURM node
name on which the slurmd is running.
See the section LOGGING if a pathname is specified.
SlurmdPidFile
Fully qualified pathname of a file into which the slurmd daemon may write its process id. This
may be used for automated signal processing. Any "%h" within the name is replaced with the
hostname on which the slurmd is running. Any "%n" within the name is replaced with the SLURM node
name on which the slurmd is running. The default value is "/var/run/slurmd.pid".
SlurmdPort
The port number that the SLURM compute node daemon, slurmd, listens to for work. The default value
is SLURMD_PORT as established at system build time. If none is explicitly specified, its value
will be 6818. NOTE: Either slurmctld and slurmd daemons must not execute on the same nodes or the
values of SlurmctldPort and SlurmdPort must be different.
SlurmdSpoolDir
Fully qualified pathname of a directory into which the slurmd daemon's state information and batch
job script information are written. This must be a common pathname for all nodes, but should
represent a directory which is local to each node (reference a local file system). The default
value is "/var/spool/slurmd". Any "%h" within the name is replaced with the hostname on which the
slurmd is running. Any "%n" within the name is replaced with the SLURM node name on which the
slurmd is running.
SlurmdTimeout
The interval, in seconds, that the SLURM controller waits for slurmd to respond before configuring
that node's state to DOWN. A value of zero indicates the node will not be tested by slurmctld to
confirm the state of slurmd, the node will not be automatically set to a DOWN state indicating a
non-responsive slurmd, and some other tool will take responsibility for monitoring the state of
each compute node and its slurmd daemon. SLURM's hierarchical communication mechanism is used to
ping the slurmd daemons in order to minimize system noise and overhead. The default value is 300
seconds. The value may not exceed 65533 seconds.
SlurmSchedLogFile
Fully qualified pathname of the scheduling event logging file. The syntax of this parameter is
the same as for SlurmctldLogFile. In order to configure scheduler logging, set both the
SlurmSchedLogFile and SlurmSchedLogLevel parameters.
SlurmSchedLogLevel
The initial level of scheduling event logging, similar to the SlurmctlDebug parameter used to
control the initial level of slurmctld logging. Valid values for SlurmSchedLogLevel are "0"
(scheduler logging disabled) and "1" (scheduler logging enabled). If this parameter is omitted,
the value defaults to "0" (disabled). In order to configure scheduler logging, set both the
SlurmSchedLogFile and SlurmSchedLogLevel parameters. The scheduler logging level can be changed
dynamically using scontrol.
SrunEpilog
Fully qualified pathname of an executable to be run by srun following the completion of a job
step. The command line arguments for the executable will be the command and arguments of the job
step. This configuration parameter may be overridden by srun's --epilog parameter. Note that
while the other "Epilog" executables (e.g., TaskEpilog) are run by slurmd on the compute nodes
where the tasks are executed, the SrunEpilog runs on the node where the "srun" is executing.
SrunProlog
Fully qualified pathname of an executable to be run by srun prior to the launch of a job step.
The command line arguments for the executable will be the command and arguments of the job step.
This configuration parameter may be overridden by srun's --prolog parameter. Note that while the
other "Prolog" executables (e.g., TaskProlog) are run by slurmd on the compute nodes where the
tasks are executed, the SrunProlog runs on the node where the "srun" is executing.
StateSaveLocation
Fully qualified pathname of a directory into which the SLURM controller, slurmctld, saves its
state (e.g. "/usr/local/slurm/checkpoint"). SLURM state will saved here to recover from system
failures. SlurmUser must be able to create files in this directory. If you have a
BackupController configured, this location should be readable and writable by both systems. Since
all running and pending job information is stored here, the use of a reliable file system (e.g.
RAID) is recommended. The default value is "/var/spool". If any slurm daemons terminate
abnormally, their core files will also be written into this directory.
SuspendExcNodes
Specifies the nodes which are to not be placed in power save mode, even if the node remains idle
for an extended period of time. Use SLURM's hostlist expression to identify nodes. By default no
nodes are excluded. Related configuration options include ResumeTimeout, ResumeProgram,
ResumeRate, SuspendProgram, SuspendRate, SuspendTime, SuspendTimeout, and SuspendExcParts.
SuspendExcParts
Specifies the partitions whose nodes are to not be placed in power save mode, even if the node
remains idle for an extended period of time. Multiple partitions can be identified and separated
by commas. By default no nodes are excluded. Related configuration options include
ResumeTimeout, ResumeProgram, ResumeRate, SuspendProgram, SuspendRate, SuspendTime SuspendTimeout,
and SuspendExcNodes.
SuspendProgram
SuspendProgram is the program that will be executed when a node remains idle for an extended
period of time. This program is expected to place the node into some power save mode. This can
be used to reduce the frequency and voltage of a node or completely power the node off. The
program executes as SlurmUser. The argument to the program will be the names of nodes to be
placed into power savings mode (using SLURM's hostlist expression format). By default, no program
is run. Related configuration options include ResumeTimeout, ResumeProgram, ResumeRate,
SuspendRate, SuspendTime, SuspendTimeout, SuspendExcNodes, and SuspendExcParts.
SuspendRate
The rate at which nodes are place into power save mode by SuspendProgram. The value is number of
nodes per minute and it can be used to prevent a large drop in power power consumption (e.g. after
a large job completes). A value of zero results in no limits being imposed. The default value is
60 nodes per minute. Related configuration options include ResumeTimeout, ResumeProgram,
ResumeRate, SuspendProgram, SuspendTime, SuspendTimeout, SuspendExcNodes, and SuspendExcParts.
SuspendTime
Nodes which remain idle for this number of seconds will be placed into power save mode by
SuspendProgram. A value of -1 disables power save mode and is the default. Related configuration
options include ResumeTimeout, ResumeProgram, ResumeRate, SuspendProgram, SuspendRate,
SuspendTimeout, SuspendExcNodes, and SuspendExcParts.
SuspendTimeout
Maximum time permitted (in second) between when a node suspend request is issued and when the node
shutdown. At that time the node must ready for a resume request to be issued as needed for new
work. The default value is 30 seconds. Related configuration options include ResumeProgram,
ResumeRate, ResumeTimeout, SuspendRate, SuspendTime, SuspendProgram, SuspendExcNodes and
SuspendExcParts. More information is available at the SLURM web site (
http://slurm.schedmd.com/power_save.html ).
SwitchType
Identifies the type of switch or interconnect used for application communications. Acceptable
values include "switch/none" for switches not requiring special processing for job launch or
termination (Myrinet, Ethernet, and InfiniBand) and "switch/nrt" for IBM's Network Resource Table
API. The default value is "switch/none". All SLURM daemons, commands and running jobs must be
restarted for a change in SwitchType to take effect. If running jobs exist at the time slurmctld
is restarted with a new value of SwitchType, records of all jobs in any state may be lost.
TaskEpilog
Fully qualified pathname of a program to be execute as the slurm job's owner after termination of
each task. See TaskProlog for execution order details.
TaskPlugin
Identifies the type of task launch plugin, typically used to provide resource management within a
node (e.g. pinning tasks to specific processors). More than one task plugin can be specified in a
comma separated list. The prefix of "task/" is optional. Acceptable values include:
task/affinity enables resource containment using CPUSETs. This enables the --cpu_bind and/or
--mem_bind srun options. If you use "task/affinity" and encounter problems, it may
be due to the variety of system calls used to implement task affinity on different
operating systems. If that is the case, you may want to install Portable Linux
Process Affinity (PLPA, see http://www.open-mpi.org/software/plpa), which is
supported by SLURM.
task/cgroup enables resource containment using Linux control cgroups. This enables the
--cpu_bind and/or --mem_bind srun options. NOTE: see "man cgroup.conf" for
configuration details. NOTE: This plugin writes to disk and can slightly impact
performance. If you are running lots of short running jobs (less than a couple of
seconds) this plugin slows down performance slightly. It should probably be
avoided in an HTC environment.
task/none for systems requiring no special handling of user tasks. Lacks support for the
--cpu_bind and/or --mem_bind srun options. The default value is "task/none".
TaskPluginParam
Optional parameters for the task plugin. Multiple options should be comma separated If None,
Sockets, Cores, Threads, and/or Verbose are specified, they will override the --cpu_bind option
specified by the user in the srun command. None, Sockets, Cores and Threads are mutually
exclusive and since they decrease scheduling flexibility are not generally recommended (select no
more than one of them). Cpusets and Sched are mutually exclusive (select only one of them).
Cores Always bind to cores. Overrides user options or automatic binding.
Cpusets Use cpusets to perform task affinity functions. By default, Sched task binding is
performed.
None Perform no task binding. Overrides user options or automatic binding.
Sched Use sched_setaffinity or plpa_sched_setaffinity (if available) to bind tasks to
processors.
Sockets Always bind to sockets. Overrides user options or automatic binding.
Threads Always bind to threads. Overrides user options or automatic binding.
Verbose Verbosely report binding before tasks run. Overrides user options.
TaskProlog
Fully qualified pathname of a program to be execute as the slurm job's owner prior to initiation
of each task. Besides the normal environment variables, this has SLURM_TASK_PID available to
identify the process ID of the task being started. Standard output from this program can be used
to control the environment variables and output for the user program.
export NAME=value Will set environment variables for the task being spawned. Everything after
the equal sign to the end of the line will be used as the value for the
environment variable. Exporting of functions is not currently supported.
print ... Will cause that line (without the leading "print ") to be printed to the job's
standard output.
unset NAME Will clear environment variables for the task being spawned.
The order of task prolog/epilog execution is as follows:
1. pre_launch_priv()
Function in TaskPlugin
1. pre_launch() Function in TaskPlugin
2. TaskProlog System-wide per task program defined in slurm.conf
3. user prolog Job step specific task program defined using srun's --task-prolog option or
SLURM_TASK_PROLOG environment variable
4. Execute the job step's task
5. user epilog Job step specific task program defined using srun's --task-epilog option or
SLURM_TASK_EPILOG environment variable
6. TaskEpilog System-wide per task program defined in slurm.conf
7. post_term() Function in TaskPlugin
TmpFS Fully qualified pathname of the file system available to user jobs for temporary storage. This
parameter is used in establishing a node's TmpDisk space. The default value is "/tmp".
TopologyPlugin
Identifies the plugin to be used for determining the network topology and optimizing job
allocations to minimize network contention. See NETWORK TOPOLOGY below for details. Additional
plugins may be provided in the future which gather topology information directly from the network.
Acceptable values include:
topology/3d_torus default for Sun Constellation systems, best-fit logic over three-dimensional
topology
topology/node_rank orders nodes based upon information a node_rank field in the node record as
generated by a select plugin. SLURM performs a best-fit algorithm over those
ordered nodes
topology/none default for other systems, best-fit logic over one-dimensional topology
topology/tree used for a hierarchical network as described in a topology.conf file
TrackWCKey
Boolean yes or no. Used to set display and track of the Workload Characterization Key. Must be
set to track wckey usage.
TreeWidth
Slurmd daemons use a virtual tree network for communications. TreeWidth specifies the width of
the tree (i.e. the fanout). On architectures with a front end node running the slurmd daemon, the
value must always be equal to or greater than the number of front end nodes which eliminates the
need for message forwarding between the slurmd daemons. On other architectures the default value
is 50, meaning each slurmd daemon can communicate with up to 50 other slurmd daemons and over 2500
nodes can be contacted with two message hops. The default value will work well for most clusters.
Optimal system performance can typically be achieved if TreeWidth is set to the square root of the
number of nodes in the cluster for systems having no more than 2500 nodes or the cube root for
larger systems. The value may not exceed 65533.
UnkillableStepProgram
If the processes in a job step are determined to be unkillable for a period of time specified by
the UnkillableStepTimeout variable, the program specified by UnkillableStepProgram will be
executed. This program can be used to take special actions to clean up the unkillable processes
and/or notify computer administrators. The program will be run SlurmdUser (usually "root"). By
default no program is run.
UnkillableStepTimeout
The length of time, in seconds, that SLURM will wait before deciding that processes in a job step
are unkillable (after they have been signaled with SIGKILL) and execute UnkillableStepProgram as
described above. The default timeout value is 60 seconds.
UsePAM If set to 1, PAM (Pluggable Authentication Modules for Linux) will be enabled. PAM is used to
establish the upper bounds for resource limits. With PAM support enabled, local system
administrators can dynamically configure system resource limits. Changing the upper bound of a
resource limit will not alter the limits of running jobs, only jobs started after a change has
been made will pick up the new limits. The default value is 0 (not to enable PAM support).
Remember that PAM also needs to be configured to support SLURM as a service. For sites using
PAM's directory based configuration option, a configuration file named slurm should be created.
The module-type, control-flags, and module-path names that should be included in the file are:
auth required pam_localuser.so
auth required pam_shells.so
account required pam_unix.so
account required pam_access.so
session required pam_unix.so
For sites configuring PAM with a general configuration file, the appropriate lines (see above),
where slurm is the service-name, should be added.
VSizeFactor
Memory specifications in job requests apply to real memory size (also known as resident set size).
It is possible to enforce virtual memory limits for both jobs and job steps by limiting their
virtual memory to some percentage of their real memory allocation. The VSizeFactor parameter
specifies the job's or job step's virtual memory limit as a percentage of its real memory limit.
For example, if a job's real memory limit is 500MB and VSizeFactor is set to 101 then the job will
be killed if its real memory exceeds 500MB or its virtual memory exceeds 505MB (101 percent of the
real memory limit). The default valus is 0, which disables enforcement of virtual memory limits.
The value may not exceed 65533 percent.
WaitTime
Specifies how many seconds the srun command should by default wait after the first task terminates
before terminating all remaining tasks. The "--wait" option on the srun command line overrides
this value. The default value is 0, which disables this feature. May not exceed 65533 seconds.
The configuration of nodes (or machines) to be managed by SLURM is also specified in /etc/slurm.conf.
Changes in node configuration (e.g. adding nodes, changing their processor count, etc.) require
restarting both the slurmctld daemon and the slurmd daemons. All slurmd daemons must know each node in
the system to forward messages in support of hierarchical communications. Only the NodeName must be
supplied in the configuration file. All other node configuration information is optional. It is
advisable to establish baseline node configurations, especially if the cluster is heterogeneous. Nodes
which register to the system with less than the configured resources (e.g. too little memory), will be
placed in the "DOWN" state to avoid scheduling jobs on them. Establishing baseline configurations will
also speed SLURM's scheduling process by permitting it to compare job requirements against these
(relatively few) configuration parameters and possibly avoid having to check job requirements against
every individual node's configuration. The resources checked at node registration time are: CPUs,
RealMemory and TmpDisk. While baseline values for each of these can be established in the configuration
file, the actual values upon node registration are recorded and these actual values may be used for
scheduling purposes (depending upon the value of FastSchedule in the configuration file.
Default values can be specified with a record in which NodeName is "DEFAULT". The default entry values
will apply only to lines following it in the configuration file and the default values can be reset
multiple times in the configuration file with multiple entries where "NodeName=DEFAULT". Each line where
NodeName is "DEFAULT" will replace or add to previous default values and not a reinitialize the default
values. The "NodeName=" specification must be placed on every line describing the configuration of
nodes. A single node name can not appear as a NodeName value in more than one line (duplicate node name
records will be ignored). In fact, it is generally possible and desirable to define the configurations
of all nodes in only a few lines. This convention permits significant optimization in the scheduling of
larger clusters. In order to support the concept of jobs requiring consecutive nodes on some
architectures, node specifications should be place in this file in consecutive order. No single node
name may be listed more than once in the configuration file. Use "DownNodes=" to record the state of
nodes which are temporarily in a DOWN, DRAIN or FAILING state without altering permanent configuration
information. A job step's tasks are allocated to nodes in order the nodes appear in the configuration
file. There is presently no capability within SLURM to arbitrarily order a job step's tasks.
Multiple node names may be comma separated (e.g. "alpha,beta,gamma") and/or a simple node range
expression may optionally be used to specify numeric ranges of nodes to avoid building a configuration
file with large numbers of entries. The node range expression can contain one pair of square brackets
with a sequence of comma separated numbers and/or ranges of numbers separated by a "-" (e.g.
"linux[0-64,128]", or "lx[15,18,32-33]"). Note that the numeric ranges can include one or more leading
zeros to indicate the numeric portion has a fixed number of digits (e.g. "linux[0000-1023]"). Up to two
numeric ranges can be included in the expression (e.g. "rack[0-63]_blade[0-41]"). If one or more numeric
expressions are included, one of them must be at the end of the name (e.g. "unit[0-31]rack" is invalid),
but arbitrary names can always be used in a comma separated list.
On BlueGene systems only, the square brackets should contain pairs of three digit numbers separated by a
"x". These numbers indicate the boundaries of a rectangular prism (e.g. "bgl[000x144,400x544]"). See
BlueGene documentation for more details. The node configuration specified the following information:
NodeName
Name that SLURM uses to refer to a node (or base partition for BlueGene systems). Typically this
would be the string that "/bin/hostname -s" returns. It may also be the fully qualified domain
name as returned by "/bin/hostname -f" (e.g. "foo1.bar.com"), or any valid domain name associated
with the host through the host database (/etc/hosts) or DNS, depending on the resolver settings.
Note that if the short form of the hostname is not used, it may prevent use of hostlist
expressions (the numeric portion in brackets must be at the end of the string). Only short
hostname forms are compatible with the switch/elan, switch/federation and switch/nrt plugins at
this time. It may also be an arbitrary string if NodeHostname is specified. If the NodeName is
"DEFAULT", the values specified with that record will apply to subsequent node specifications
unless explicitly set to other values in that node record or replaced with a different set of
default values. Each line where NodeName is "DEFAULT" will replace or add to previous default
values and not a reinitialize the default values. For architectures in which the node order is
significant, nodes will be considered consecutive in the order defined. For example, if the
configuration for "NodeName=charlie" immediately follows the configuration for "NodeName=baker"
they will be considered adjacent in the computer.
NodeHostname
Typically this would be the string that "/bin/hostname -s" returns. It may also be the fully
qualified domain name as returned by "/bin/hostname -f" (e.g. "foo1.bar.com"), or any valid domain
name associated with the host through the host database (/etc/hosts) or DNS, depending on the
resolver settings. Note that if the short form of the hostname is not used, it may prevent use of
hostlist expressions (the numeric portion in brackets must be at the end of the string). Only
short hostname forms are compatible with the switch/elan, switch/federation and switch/nrt plugins
at this time. A node range expression can be used to specify a set of nodes. If an expression is
used, the number of nodes identified by NodeHostname on a line in the configuration file must be
identical to the number of nodes identified by NodeName. By default, the NodeHostname will be
identical in value to NodeName.
NodeAddr
Name that a node should be referred to in establishing a communications path. This name will be
used as an argument to the gethostbyname() function for identification. If a node range
expression is used to designate multiple nodes, they must exactly match the entries in the
NodeName (e.g. "NodeName=lx[0-7] NodeAddr="elx[0-7]"). NodeAddr may also contain IP addresses.
By default, the NodeAddr will be identical in value to NodeHostname.
Boards Number of Baseboards in nodes with a baseboard controller. Note that when Boards is specified,
SocketsPerBoard, CoresPerSocket, and ThreadsPerCore should be specified. Boards and CPUs are
mutually exclusive. The default value is 1.
CoresPerSocket
Number of cores in a single physical processor socket (e.g. "2"). The CoresPerSocket value
describes physical cores, not the logical number of processors per socket. NOTE: If you have
multi-core processors, you will likely need to specify this parameter in order to optimize
scheduling. The default value is 1.
CPUs Number of logical processors on the node (e.g. "2"). If CPUs is omitted, it will set equal to the
product of Sockets, CoresPerSocket, and ThreadsPerCore. CPUs and Boards are mutually exclusive.
The default value is 1.
Feature
A comma delimited list of arbitrary strings indicative of some characteristic associated with the
node. There is no value associated with a feature at this time, a node either has a feature or it
does not. If desired a feature may contain a numeric component indicating, for example, processor
speed. By default a node has no features. Also see Gres.
Gres A comma delimited list of generic resources specifications for a node. Each resource
specification consists of a name followed by an optional colon with a numeric value (default value
is one) (e.g. "Gres=bandwidth:10000,gpu:2"). A suffix of "K", "M" or "G" may be used to multiply
the number by 1024, 1048576 or 1073741824 respectively (e.g. "Gres=bandwidth:4G,gpu:4").. By
default a node has no generic resources and its maximum count is 4,294,967,295. Also see Feature.
Port The port number that the SLURM compute node daemon, slurmd, listens to for work on this particular
node. By default there is a single port number for all slurmd daemons on all compute nodes as
defined by the SlurmdPort configuration parameter. Use of this option is not generally recommended
except for development or testing purposes. If multiple slurmd daemons execute on a node this can
specify a range of ports
Procs See CPUs.
RealMemory
Size of real memory on the node in MegaBytes (e.g. "2048"). The default value is 1.
Reason Identifies the reason for a node being in state "DOWN", "DRAINED" "DRAINING", "FAIL" or "FAILING".
Use quotes to enclose a reason having more than one word.
Sockets
Number of physical processor sockets/chips on the node (e.g. "2"). If Sockets is omitted, it will
be inferred from CPUs, CoresPerSocket, and ThreadsPerCore. NOTE: If you have multi-core
processors, you will likely need to specify these parameters. Sockets and SocketsPerBoard are
mutually exclusive. If Sockets is specified when Boards is also used, Sockets is interpreted as
SocketsPerBoard rather than total sockets. The default value is 1.
SocketsPerBoard
Number of physical processor sockets/chips on a baseboard. Sockets and SocketsPerBoard are
mutually exclusive. The default value is 1.
State State of the node with respect to the initiation of user jobs. Acceptable values are "DOWN",
"DRAIN", "FAIL", "FAILING" and "UNKNOWN". "DOWN" indicates the node failed and is unavailable to
be allocated work. "DRAIN" indicates the node is unavailable to be allocated work. "FAIL"
indicates the node is expected to fail soon, has no jobs allocated to it, and will not be
allocated to any new jobs. "FAILING" indicates the node is expected to fail soon, has one or more
jobs allocated to it, but will not be allocated to any new jobs. "UNKNOWN" indicates the node's
state is undefined (BUSY or IDLE), but will be established when the slurmd daemon on that node
registers. The default value is "UNKNOWN". Also see the DownNodes parameter below.
ThreadsPerCore
Number of logical threads in a single physical core (e.g. "2"). Note that the SLURM can allocate
resources to jobs down to the resolution of a core. If your system is configured with more than
one thread per core, execution of a different job on each thread is not supported unless you
configure SelectTypeParameters=CR_CPU plus CPUs; do not configure Sockets, CoresPerSocket or
ThreadsPerCore. A job can execute a one task per thread from within one job step or execute a
distinct job step on each of the threads. Note also if you are running with more than 1 thread
per core and running the select/cons_res plugin you will want to set the SelectTypeParameters
variable to something other than CR_CPU to avoid unexpected results. The default value is 1.
TmpDisk
Total size of temporary disk storage in TmpFS in MegaBytes (e.g. "16384"). TmpFS (for "Temporary
File System") identifies the location which jobs should use for temporary storage. Note this does
not indicate the amount of free space available to the user on the node, only the total file
system size. The system administration should insure this file system is purged as needed so that
user jobs have access to most of this space. The Prolog and/or Epilog programs (specified in the
configuration file) might be used to insure the file system is kept clean. The default value is
0.
Weight The priority of the node for scheduling purposes. All things being equal, jobs will be allocated
the nodes with the lowest weight which satisfies their requirements. For example, a heterogeneous
collection of nodes might be placed into a single partition for greater system utilization,
responsiveness and capability. It would be preferable to allocate smaller memory nodes rather than
larger memory nodes if either will satisfy a job's requirements. The units of weight are
arbitrary, but larger weights should be assigned to nodes with more processors, memory, disk
space, higher processor speed, etc. Note that if a job allocation request can not be satisfied
using the nodes with the lowest weight, the set of nodes with the next lowest weight is added to
the set of nodes under consideration for use (repeat as needed for higher weight values). If you
absolutely want to minimize the number of higher weight nodes allocated to a job (at a cost of
higher scheduling overhead), give each node a distinct Weight value and they will be added to the
pool of nodes being considered for scheduling individually. The default value is 1.
The "DownNodes=" configuration permits you to mark certain nodes as in a DOWN, DRAIN, FAIL, or FAILING
state without altering the permanent configuration information listed under a "NodeName=" specification.
DownNodes
Any node name, or list of node names, from the "NodeName=" specifications.
Reason Identifies the reason for a node being in state "DOWN", "DRAIN", "FAIL" or "FAILING. Use quotes
to enclose a reason having more than one word.
State State of the node with respect to the initiation of user jobs. Acceptable values are "BUSY",
"DOWN", "DRAIN", "FAIL", "FAILING, "IDLE", and "UNKNOWN".
CLOUD Indicates the node exists in the cloud. It's initial state will be treated as powered
down. The node will be available for use after it's state is recovered from SLURM's
state save file or the slurmd daemon starts on the compute node.
DOWN Indicates the node failed and is unavailable to be allocated work.
DRAIN Indicates the node is unavailable to be allocated work.on.
FAIL Indicates the node is expected to fail soon, has no jobs allocated to it, and will not
be allocated to any new jobs.
FAILING Indicates the node is expected to fail soon, has one or more jobs allocated to it, but
will not be allocated to any new jobs.
FUTURE Indicates the node is defined for future use and need not exist when the SLURM daemons
are started. These nodes can be made available for use simply by updating the node state
using the scontrol command rather than restarting the slurmctld daemon. After these
nodes are made available, change their State in the slurm.conf file. Until these nodes
are made available, they will not be seen using any SLURM commands or nor will any
attempt be made to contact them.
UNKNOWN Indicates the node's state is undefined (BUSY or IDLE), but will be established when the
slurmd daemon on that node registers. The default value is "UNKNOWN".
On computers where frontend nodes are used to execute batch scripts rather than compute nodes (BlueGene
or Cray systems), one may configure one or more frontend nodes using the configuration parameters defined
below. These options are very similar to those used in configuring compute nodes. These options may only
be used on systems configured and built with the appropriate parameters (--have-front-end,
--enable-bluegene-emulation) or a system determined to have the appropriate architecture by the configure
script (BlueGene or Cray systems). The front end configuration specifies the following information:
AllowGroups
Comma separated list of group names which may execute jobs on this front end node. By default, all
groups may use this front end node. May not be used with the DenyGroups option.
AllowUsers
Comma separated list of user names which may execute jobs on this front end node. By default, all
users may use this front end node. May not be used with the DenyUsers option.
DenyGroups
Comma separated list of group names which are prevented from executing jobs on this front end
node. May not be used with the AllowGroups option.
DenyUsers
Comma separated list of user names which are prevented from executing jobs on this front end node.
May not be used with the AllowUsers option.
FrontendName
Name that SLURM uses to refer to a frontend node. Typically this would be the string that
"/bin/hostname -s" returns. It may also be the fully qualified domain name as returned by
"/bin/hostname -f" (e.g. "foo1.bar.com"), or any valid domain name associated with the host
through the host database (/etc/hosts) or DNS, depending on the resolver settings. Note that if
the short form of the hostname is not used, it may prevent use of hostlist expressions (the
numeric portion in brackets must be at the end of the string). If the FrontendName is "DEFAULT",
the values specified with that record will apply to subsequent node specifications unless
explicitly set to other values in that frontend node record or replaced with a different set of
default values. Each line where FrontendName is "DEFAULT" will replace or add to previous default
values and not a reinitialize the default values. Note that since the naming of front end nodes
would typically not follow that of the compute nodes (e.g. lacking X, Y and Z coordinates found in
the compute node naming scheme), each front end node name should be listed separately and without
a hostlist expression (i.e. frontend00,frontend01" rather than "frontend[00-01]").</p>
FrontendAddr
Name that a frontend node should be referred to in establishing a communications path. This name
will be used as an argument to the gethostbyname() function for identification. As with
FrontendName, list the individual node addresses rather than using a hostlist expression. The
number of FrontendAddr records per line must equal the number of FrontendName records per line
(i.e. you can't map to node names to one address). FrontendAddr may also contain IP addresses.
By default, the FrontendAddr will be identical in value to FrontendName.
Port The port number that the SLURM compute node daemon, slurmd, listens to for work on this particular
frontend node. By default there is a single port number for all slurmd daemons on all frontend
nodes as defined by the SlurmdPort configuration parameter. Use of this option is not generally
recommended except for development or testing purposes.
Reason Identifies the reason for a frontend node being in state "DOWN", "DRAINED" "DRAINING", "FAIL" or
"FAILING". Use quotes to enclose a reason having more than one word.
State State of the frontend node with respect to the initiation of user jobs. Acceptable values are
"DOWN", "DRAIN", "FAIL", "FAILING" and "UNKNOWN". "DOWN" indicates the frontend node has failed
and is unavailable to be allocated work. "DRAIN" indicates the frontend node is unavailable to be
allocated work. "FAIL" indicates the frontend node is expected to fail soon, has no jobs
allocated to it, and will not be allocated to any new jobs. "FAILING" indicates the frontend node
is expected to fail soon, has one or more jobs allocated to it, but will not be allocated to any
new jobs. "UNKNOWN" indicates the frontend node's state is undefined (BUSY or IDLE), but will be
established when the slurmd daemon on that node registers. The default value is "UNKNOWN". Also
see the DownNodes parameter below.
For example: "FrontendName=frontend[00-03] FrontendAddr=efrontend[00-03] State=UNKNOWN" is used to
define four front end nodes for running slurmd daemons.
The partition configuration permits you to establish different job limits or access controls for various
groups (or partitions) of nodes. Nodes may be in more than one partition, making partitions serve as
general purpose queues. For example one may put the same set of nodes into two different partitions,
each with different constraints (time limit, job sizes, groups allowed to use the partition, etc.). Jobs
are allocated resources within a single partition. Default values can be specified with a record in
which PartitionName is "DEFAULT". The default entry values will apply only to lines following it in the
configuration file and the default values can be reset multiple times in the configuration file with
multiple entries where "PartitionName=DEFAULT". The "PartitionName=" specification must be placed on
every line describing the configuration of partitions. Each line where PartitionName is "DEFAULT" will
replace or add to previous default values and not a reinitialize the default values. A single partition
name can not appear as a PartitionName value in more than one line (duplicate partition name records will
be ignored). If a partition that is in use is deleted from the configuration and slurm is restarted or
reconfigured (scontrol reconfigure), jobs using the partition are canceled. NOTE: Put all parameters for
each partition on a single line. Each line of partition configuration information should represent a
different partition. The partition configuration file contains the following information:
AllocNodes
Comma separated list of nodes from which users can submit jobs in the partition. Node names may
be specified using the node range expression syntax described above. The default value is "ALL".
AllowGroups
Comma separated list of group names which may execute jobs in the partition. If at least one
group associated with the user attempting to execute the job is in AllowGroups, he will be
permitted to use this partition. Jobs executed as user root can use any partition without regard
to the value of AllowGroups. If user root attempts to execute a job as another user (e.g. using
srun's --uid option), this other user must be in one of groups identified by AllowGroups for the
job to successfully execute. The default value is "ALL". NOTE: For performance reasons, SLURM
maintains a list of user IDs allowed to use each partition and this is checked at job submission
time. This list of user IDs is updated when the slurmctld daemon is restarted, reconfigured (e.g.
"scontrol reconfig") or the partition's AllowGroups value is reset, even if is value is unchanged
(e.g. "scontrol update PartitionName=name AllowGroups=group"). For a user's access to a partition
to change, both his group membership must change and SLURM's internal user ID list must change
using one of the methods described above.
Alternate
Partition name of alternate partition to be used if the state of this partition is "DRAIN" or
"INACTIVE."
Default
If this keyword is set, jobs submitted without a partition specification will utilize this
partition. Possible values are "YES" and "NO". The default value is "NO".
DefMemPerCPU
Default real memory size available per allocated CPU in MegaBytes. Used to avoid over-subscribing
memory and causing paging. DefMemPerCPU would generally be used if individual processors are
allocated to jobs (SelectType=select/cons_res). If not set, the DefMemPerCPU value for the entire
cluster will be used. Also see DefMemPerNode and MaxMemPerCPU. DefMemPerCPU and DefMemPerNode
are mutually exclusive. NOTE: Enforcement of memory limits currently requires enabling of
accounting, which samples memory use on a periodic basis (data need not be stored, just
collected).
DefMemPerNode
Default real memory size available per allocated node in MegaBytes. Used to avoid
over-subscribing memory and causing paging. DefMemPerNode would generally be used if whole nodes
are allocated to jobs (SelectType=select/linear) and resources are shared (Shared=yes or
Shared=force). If not set, the DefMemPerNode value for the entire cluster will be used. Also see
DefMemPerCPU and MaxMemPerNode. DefMemPerCPU and DefMemPerNode are mutually exclusive. NOTE:
Enforcement of memory limits currently requires enabling of accounting, which samples memory use
on a periodic basis (data need not be stored, just collected).
DefaultTime
Run time limit used for jobs that don't specify a value. If not set then MaxTime will be used.
Format is the same as for MaxTime.
DisableRootJobs
If set to "YES" then user root will be prevented from running any jobs on this partition. The
default value will be the value of DisableRootJobs set outside of a partition specification (which
is "NO", allowing user root to execute jobs).
GraceTime
Specifies, in units of seconds, the preemption grace time to be extended to a job which has been
selected for preemption. The default value is zero, no preemption grace time is allowed on this
partition. (Meaningful only for PreemptMode=CANCEL)
Hidden Specifies if the partition and its jobs are to be hidden by default. Hidden partitions will by
default not be reported by the SLURM APIs or commands. Possible values are "YES" and "NO". The
default value is "NO". Note that partitions that a user lacks access to by virtue of the
AllowGroups parameter will also be hidden by default.
MaxCPUsPerNode
Maximum number of CPUs on any node available to all jobs from this partition. This can be
especially useful to schedule GPUs. For example a node can be associated with two Slurm partitions
(e.g. "cpu" and "gpu") and the partition/queue "cpu" could be limited to only a subset of the
node's CPUs, insuring that one or more CPUs would be available to jobs in the "gpu"
partition/queue.
MaxMemPerCPU
Maximum real memory size available per allocated CPU in MegaBytes. Used to avoid over-subscribing
memory and causing paging. MaxMemPerCPU would generally be used if individual processors are
allocated to jobs (SelectType=select/cons_res). If not set, the MaxMemPerCPU value for the entire
cluster will be used. Also see DefMemPerCPU and MaxMemPerNode. MaxMemPerCPU and MaxMemPerNode
are mutually exclusive. NOTE: Enforcement of memory limits currently requires enabling of
accounting, which samples memory use on a periodic basis (data need not be stored, just
collected).
MaxMemPerNode
Maximum real memory size available per allocated node in MegaBytes. Used to avoid
over-subscribing memory and causing paging. MaxMemPerNode would generally be used if whole nodes
are allocated to jobs (SelectType=select/linear) and resources are shared (Shared=yes or
Shared=force). If not set, the MaxMemPerNode value for the entire cluster will be used. Also see
DefMemPerNode and MaxMemPerCPU. MaxMemPerCPU and MaxMemPerNode are mutually exclusive. NOTE:
Enforcement of memory limits currently requires enabling of accounting, which samples memory use
on a periodic basis (data need not be stored, just collected).
MaxNodes
Maximum count of nodes which may be allocated to any single job. For BlueGene systems this will
be a c-nodes count and will be converted to a midplane count with a reduction in resolution. The
default value is "UNLIMITED", which is represented internally as -1. This limit does not apply to
jobs executed by SlurmUser or user root.
MaxTime
Maximum run time limit for jobs. Format is minutes, minutes:seconds, hours:minutes:seconds,
days-hours, days-hours:minutes, days-hours:minutes:seconds or "UNLIMITED". Time resolution is one
minute and second values are rounded up to the next minute. This limit does not apply to jobs
executed by SlurmUser or user root.
MinNodes
Minimum count of nodes which may be allocated to any single job. For BlueGene systems this will
be a c-nodes count and will be converted to a midplane count with a reduction in resolution. The
default value is 1. This limit does not apply to jobs executed by SlurmUser or user root.
Nodes Comma separated list of nodes (or base partitions for BlueGene systems) which are associated with
this partition. Node names may be specified using the node range expression syntax described
above. A blank list of nodes (i.e. "Nodes= ") can be used if one wants a partition to exist, but
have no resources (possibly on a temporary basis).
PartitionName
Name by which the partition may be referenced (e.g. "Interactive"). This name can be specified by
users when submitting jobs. If the PartitionName is "DEFAULT", the values specified with that
record will apply to subsequent partition specifications unless explicitly set to other values in
that partition record or replaced with a different set of default values. Each line where
PartitionName is "DEFAULT" will replace or add to previous default values and not a reinitialize
the default values.
PreemptMode
Mechanism used to preempt jobs from this partition when PreemptType=preempt/partition_prio is
configured. This partition specific PreemptMode configuration parameter will override the
PreemptMode configuration parameter set for the cluster as a whole. The cluster-level PreemptMode
must include the GANG option if PreemptMode is configured to SUSPEND for any partition. The
cluster-level PreemptMode must not be OFF if PreemptMode is enabled for any partition. See the
description of the cluster-level PreemptMode configuration parameter above for further
information.
Priority
Jobs submitted to a higher priority partition will be dispatched before pending jobs in lower
priority partitions and if possible they will preempt running jobs from lower priority partitions.
Note that a partition's priority takes precedence over a job's priority. The value may not exceed
65533.
ReqResv
Specifies users of this partition are required to designate a reservation when submitting a job.
This option can be useful in restricting usage of a partition that may have higher priority or
additional resources to be allowed only within a reservation. Possible values are "YES" and "NO".
The default value is "NO".
RootOnly
Specifies if only user ID zero (i.e. user root) may allocate resources in this partition. User
root may allocate resources for any other user, but the request must be initiated by user root.
This option can be useful for a partition to be managed by some external entity (e.g. a
higher-level job manager) and prevents users from directly using those resources. Possible values
are "YES" and "NO". The default value is "NO".
SelectTypeParameters
Partition-specific resource allocation type. Supported values are CR_Core and CR_Socket. Use
requires the system-wide SelectTypeParameters value be set plus CR_ALLOCATE_FULL_SOCKET.
Shared Controls the ability of the partition to execute more than one job at a time on each resource
(node, socket or core depending upon the value of SelectTypeParameters). If resources are to be
shared, avoiding memory over-subscription is very important. SelectTypeParameters should be
configured to treat memory as a consumable resource and the --mem option should be used for job
allocations. Sharing of resources is typically useful only when using gang scheduling
(PreemptMode=suspend or PreemptMode=kill). Possible values for Shared are "EXCLUSIVE", "FORCE",
"YES", and "NO". Note that a value of "YES" or "FORCE" can negatively impact performance for
systems with many thousands of running jobs. The default value is "NO". For more information see
the following web pages:
http://slurm.schedmd.com/cons_res.html,
http://slurm.schedmd.com/cons_res_share.html,
http://slurm.schedmd.com/gang_scheduling.html, and
http://slurm.schedmd.com/preempt.html.
EXCLUSIVE Allocates entire nodes to jobs even with select/cons_res configured. Jobs that run in
partitions with "Shared=EXCLUSIVE" will have exclusive access to all allocated nodes.
FORCE Makes all resources in the partition available for sharing without any means for users
to disable it. May be followed with a colon and maximum number of jobs in running or
suspended state. For example "Shared=FORCE:4" enables each node, socket or core to
execute up to four jobs at once. Recommended only for BlueGene systems configured
with small blocks or for systems running with gang scheduling
(SchedulerType=sched/gang).
YES Makes all resources in the partition available for sharing upon request by the job.
Resources will only be over-subscribed when explicitly requested by the user using the
"--share" option on job submission. May be followed with a colon and maximum number
of jobs in running or suspended state. For example "Shared=YES:4" enables each node,
socket or core to execute up to four jobs at once. Recommended only for systems
running with gang scheduling (SchedulerType=sched/gang).
NO Selected resources are allocated to a single job. No resource will be allocated to
more than one job.
State State of partition or availability for use. Possible values are "UP", "DOWN", "DRAIN" and
"INACTIVE". The default value is "UP". See also the related "Alternate" keyword.
UP Designates that new jobs may queued on the partition, and that jobs may be allocated
nodes and run from the partition.
DOWN Designates that new jobs may be queued on the partition, but queued jobs may not be
allocated nodes and run from the partition. Jobs already running on the partition
continue to run. The jobs must be explicitly canceled to force their termination.
DRAIN Designates that no new jobs may be queued on the partition (job submission requests will
be denied with an error message), but jobs already queued on the partition may be
allocated nodes and run. See also the "Alternate" partition specification.
INACTIVE Designates that no new jobs may be queued on the partition, and jobs already queued may
not be allocated nodes and run. See also the "Alternate" partition specification.
Prolog and Epilog Scripts
There are a variety of prolog and epilog program options that execute with various permissions and at
various times. The four options most likely to be used are: Prolog and Epilog (executed once on each
compute node for each job) plus PrologSlurmctld and EpilogSlurmctld (executed once on the ControlMachine
for each job).
NOTE: Standard output and error messages are normally not preserved. Explicitly write output and error
messages to an appropriate location if you wish to preserve that information.
NOTE: The Prolog script is ONLY run on any individual node when it first sees a job step from a new
allocation; it does not run the Prolog immediately when an allocation is granted. If no job steps from
an allocation are run on a node, it will never run the Prolog for that allocation. The Epilog, on the
other hand, always runs on every node of an allocation when the allocation is released.
If the Epilog fails (returns a non-zero exit code), this will result in the node being set to a DOWN
state. If the EpilogSlurmctld fails (returns a non-zero exit code), this will only be logged. If the
Prolog fails (returns a non-zero exit code), this will result in the node being set to a DOWN state and
the job requeued to executed on another node. If the PrologSlurmctld fails (returns a non-zero exit
code), this will result in the job requeued to executed on another node if possible. Only batch jobs can
be requeued. Interactive jobs (salloc and srun) will be cancelled if the PrologSlurmctld fails.
Information about the job is passed to the script using environment variables. Unless otherwise
specified, these environment variables are available to all of the programs.
BASIL_RESERVATION_ID
Basil reservation ID. Available on Cray XT/XE systems only.
MPIRUN_PARTITION
BlueGene partition name. Available on BlueGene systems only.
SLURM_JOB_ACCOUNT
Account name used for the job. Available in PrologSlurmctld and EpilogSlurmctld only.
SLURM_JOB_CONSTRAINTS
Features required to run the job. Available in PrologSlurmctld and EpilogSlurmctld only.
SLURM_JOB_DERIVED_EC
The highest exit code of all of the job steps. Available in EpilogSlurmctld only.
SLURM_JOB_EXIT_CODE
The exit code of the job script (or salloc). Available in EpilogSlurmctld only.
SLURM_JOB_GID
Group ID of the job's owner. Available in PrologSlurmctld and EpilogSlurmctld only.
SLURM_JOB_GROUP
Group name of the job's owner. Available in PrologSlurmctld and EpilogSlurmctld only.
SLURM_JOB_ID
Job ID.
SLURM_JOB_NAME
Name of the job. Available in PrologSlurmctld and EpilogSlurmctld only.
SLURM_JOB_NODELIST
Nodes assigned to job. A SLURM hostlist expression. "scontrol show hostnames" can be used to
convert this to a list of individual host names. Available in PrologSlurmctld and EpilogSlurmctld
only.
SLURM_JOB_PARTITION
Partition that job runs in. Available in PrologSlurmctld and EpilogSlurmctld only.
SLURM_JOB_UID
User ID of the job's owner.
SLURM_JOB_USER
User name of the job's owner.
NETWORK TOPOLOGY
SLURM is able to optimize job allocations to minimize network contention. Special SLURM logic is used to
optimize allocations on systems with a three-dimensional interconnect (BlueGene, Sun Constellation, etc.)
and information about configuring those systems are available on web pages available here:
<http://slurm.schedmd.com/>. For a hierarchical network, SLURM needs to have detailed information about
how nodes are configured on the network switches.
Given network topology information, SLURM allocates all of a job's resources onto a single leaf of the
network (if possible) using a best-fit algorithm. Otherwise it will allocate a job's resources onto
multiple leaf switches so as to minimize the use of higher-level switches. The TopologyPlugin parameter
controls which plugin is used to collect network topology information. The only values presently
supported are "topology/3d_torus" (default for IBM BlueGene, Sun Constellation and Cray XT/XE systems,
performs best-fit logic over three-dimensional topology), "topology/none" (default for other systems,
best-fit logic over one-dimensional topology), "topology/tree" (determine the network topology based upon
information contained in a topology.conf file, see "man topology.conf" for more information). Future
plugins may gather topology information directly from the network. The topology information is optional.
If not provided, SLURM will perform a best-fit algorithm assuming the nodes are in a one-dimensional
array as configured and the communications cost is related to the node distance in this array.
RELOCATING CONTROLLERS
If the cluster's computers used for the primary or backup controller will be out of service for an
extended period of time, it may be desirable to relocate them. In order to do so, follow this procedure:
1. Stop the SLURM daemons
2. Modify the slurm.conf file appropriately
3. Distribute the updated slurm.conf file to all nodes
4. Restart the SLURM daemons
There should be no loss of any running or pending jobs. Insure that any nodes added to the cluster have
the current slurm.conf file installed.
CAUTION: If two nodes are simultaneously configured as the primary controller (two nodes on which
ControlMachine specify the local host and the slurmctld daemon is executing on each), system behavior
will be destructive. If a compute node has an incorrect ControlMachine or BackupController parameter,
that node may be rendered unusable, but no other harm will result.
EXAMPLE
#
# Sample /etc/slurm.conf for dev[0-25].llnl.gov
# Author: John Doe
# Date: 11/06/2001
#
ControlMachine=dev0
ControlAddr=edev0
BackupController=dev1
BackupAddr=edev1
#
AuthType=auth/munge
Epilog=/usr/local/slurm/epilog
Prolog=/usr/local/slurm/prolog
FastSchedule=1
FirstJobId=65536
InactiveLimit=120
JobCompType=jobcomp/filetxt
JobCompLoc=/var/log/slurm/jobcomp
KillWait=30
MaxJobCount=10000
MinJobAge=3600
PluginDir=/usr/local/lib:/usr/local/slurm/lib
ReturnToService=0
SchedulerType=sched/backfill
SlurmctldLogFile=/var/log/slurm/slurmctld.log
SlurmdLogFile=/var/log/slurm/slurmd.log
SlurmctldPort=7002
SlurmdPort=7003
SlurmdSpoolDir=/usr/local/slurm/slurmd.spool
StateSaveLocation=/usr/local/slurm/slurm.state
SwitchType=switch/elan
TmpFS=/tmp
WaitTime=30
JobCredentialPrivateKey=/usr/local/slurm/private.key
JobCredentialPublicCertificate=/usr/local/slurm/public.cert
#
# Node Configurations
#
NodeName=DEFAULT CPUs=2 RealMemory=2000 TmpDisk=64000
NodeName=DEFAULT State=UNKNOWN
NodeName=dev[0-25] NodeAddr=edev[0-25] Weight=16
# Update records for specific DOWN nodes
DownNodes=dev20 State=DOWN Reason="power,ETA=Dec25"
#
# Partition Configurations
#
PartitionName=DEFAULT MaxTime=30 MaxNodes=10 State=UP
PartitionName=debug Nodes=dev[0-8,18-25] Default=YES
PartitionName=batch Nodes=dev[9-17] MinNodes=4
PartitionName=long Nodes=dev[9-17] MaxTime=120 AllowGroups=admin
FILE AND DIRECTORY PERMISSIONS
There are three classes of files: Files used by slurmctld must be accessible by user SlurmUser and
accessible by the primary and backup control machines. Files used by slurmd must be accessible by user
root and accessible from every compute node. A few files need to be accessible by normal users on all
login and compute nodes. While many files and directories are listed below, most of them will not be
used with most configurations.
AccountingStorageLoc
If this specifies a file, it must be writable by user SlurmUser. The file must be accessible by
the primary and backup control machines. It is recommended that the file be readable by all users
from login and compute nodes.
Epilog Must be executable by user root. It is recommended that the file be readable by all users. The
file must exist on every compute node.
EpilogSlurmctld
Must be executable by user SlurmUser. It is recommended that the file be readable by all users.
The file must be accessible by the primary and backup control machines.
HealthCheckProgram
Must be executable by user root. It is recommended that the file be readable by all users. The
file must exist on every compute node.
JobCheckpointDir
Must be writable by user SlurmUser and no other users. The file must be accessible by the primary
and backup control machines.
JobCompLoc
If this specifies a file, it must be writable by user SlurmUser. The file must be accessible by
the primary and backup control machines.
JobCredentialPrivateKey
Must be readable only by user SlurmUser and writable by no other users. The file must be
accessible by the primary and backup control machines.
JobCredentialPublicCertificate
Readable to all users on all nodes. Must not be writable by regular users.
MailProg
Must be executable by user SlurmUser. Must not be writable by regular users. The file must be
accessible by the primary and backup control machines.
Prolog Must be executable by user root. It is recommended that the file be readable by all users. The
file must exist on every compute node.
PrologSlurmctld
Must be executable by user SlurmUser. It is recommended that the file be readable by all users.
The file must be accessible by the primary and backup control machines.
ResumeProgram
Must be executable by user SlurmUser. The file must be accessible by the primary and backup
control machines.
SallocDefaultCommand
Must be executable by all users. The file must exist on every login and compute node.
slurm.conf
Readable to all users on all nodes. Must not be writable by regular users.
SlurmctldLogFile
Must be writable by user SlurmUser. The file must be accessible by the primary and backup control
machines.
SlurmctldPidFile
Must be writable by user root. Preferably writable and removable by SlurmUser. The file must be
accessible by the primary and backup control machines.
SlurmdLogFile
Must be writable by user root. A distinct file must exist on each compute node.
SlurmdPidFile
Must be writable by user root. A distinct file must exist on each compute node.
SlurmdSpoolDir
Must be writable by user root. A distinct file must exist on each compute node.
SrunEpilog
Must be executable by all users. The file must exist on every login and compute node.
SrunProlog
Must be executable by all users. The file must exist on every login and compute node.
StateSaveLocation
Must be writable by user SlurmUser. The file must be accessible by the primary and backup control
machines.
SuspendProgram
Must be executable by user SlurmUser. The file must be accessible by the primary and backup
control machines.
TaskEpilog
Must be executable by all users. The file must exist on every compute node.
TaskProlog
Must be executable by all users. The file must exist on every compute node.
UnkillableStepProgram
Must be executable by user SlurmUser. The file must be accessible by the primary and backup
control machines.
LOGGING
Note that while SLURM daemons create log files and other files as needed, it treats the lack of parent
directories as a fatal error. This prevents the daemons from running if critical file systems are not
mounted and will minimize the risk of cold-starting (starting without preserving jobs).
Log files and job accounting files, may need to be created/owned by the "SlurmUser" uid to be
successfully accessed. Use the "chown" and "chmod" commands to set the ownership and permissions
appropriately. See the section FILE AND DIRECTORY PERMISSIONS for information about the various files
and directories used by SLURM.
It is recommended that the logrotate utility be used to insure that various log files do not become too
large. This also applies to text files used for accounting, process tracking, and the slurmdbd log if
they are used.
Here is a sample logrotate configuration. Make appropriate site modifications and save as
/etc/logrotate.d/slurm on all nodes. See the logrotate man page for more details.
##
# SLURM Logrotate Configuration
##
/var/log/slurm/*log {
compress
missingok
nocopytruncate
nocreate
nodelaycompress
nomail
notifempty
noolddir
rotate 5
sharedscripts
size=5M
create 640 slurm root
postrotate
/etc/init.d/slurm reconfig
endscript
}
COPYING
Copyright (C) 2002-2007 The Regents of the University of California. Produced at Lawrence Livermore
National Laboratory (cf, DISCLAIMER).
Copyright (C) 2008-2010 Lawrence Livermore National Security.
Copyright (C) 2010-2013 SchedMD LLC.
This file is part of SLURM, a resource management program. For details, see <http://slurm.schedmd.com/>.
SLURM is free software; you can redistribute it and/or modify it under the terms of the GNU General
Public License as published by the Free Software Foundation; either version 2 of the License, or (at your
option) any later version.
SLURM 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 General Public
License for more details.
FILES
/etc/slurm.conf
SEE ALSO
bluegene.conf(5), cgroup.conf(5), gethostbyname (3), getrlimit (2), gres.conf(5), group (5), hostname
(1), scontrol(1), slurmctld(8), slurmd(8), slurmdbd(8), slurmdbd.conf(5), srun(1), spank(8), syslog (2),
topology.conf(5), wiki.conf(5)
slurm.conf 2.6 March 2013 slurm.conf(5)