Skip to content
Snippets Groups Projects
Commit ccf843af authored by Danny Auble's avatar Danny Auble
Browse files

removed whitespace

parent 3cd48428
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
doc/html/configurator.html.in
+ 135
− 136
View file @ ccf843af
...@@ -45,11 +45,11 @@ function get_accounting_storage_type_field(gather, form_storage) ...@@ -45,11 +45,11 @@ function get_accounting_storage_type_field(gather, form_storage)
{ {
if (form_storage[i].checked) if (form_storage[i].checked)
{ {
if(form_storage[i].value == "none" && !(gather == "none")) if(form_storage[i].value == "none" && !(gather == "none"))
{ {
return "filetxt" return "filetxt"
} }
else if(!(form_storage[i].value == "none") && gather == "none") else if(!(form_storage[i].value == "none") && gather == "none")
{ {
return "none" return "none"
} }
...@@ -117,7 +117,7 @@ function get_task_plugin_param() ...@@ -117,7 +117,7 @@ function get_task_plugin_param()
for (var i=0; i<document.config.task_plugin.length; i++) { for (var i=0; i<document.config.task_plugin.length; i++) {
if (document.config.task_plugin[i].checked) { if (document.config.task_plugin[i].checked) {
if (document.config.task_plugin[i].value == "affinity") { if (document.config.task_plugin[i].value == "affinity") {
return "TaskPluginParam=" + return "TaskPluginParam=" +
get_radio_value(document.config.task_plugin_param) get_radio_value(document.config.task_plugin_param)
} }
} }
...@@ -129,7 +129,7 @@ function hide_box() ...@@ -129,7 +129,7 @@ function hide_box()
{ {
var popup = document.getElementById('out_box'); var popup = document.getElementById('out_box');
popup.style.visibility = 'hidden'; popup.style.visibility = 'hidden';
} }
function displayfile() function displayfile()
...@@ -295,21 +295,21 @@ function displayfile() ...@@ -295,21 +295,21 @@ function displayfile()
get_field2(" CoresPerSocket",document.config.cores_per_socket) + get_field2(" CoresPerSocket",document.config.cores_per_socket) +
get_field2(" ThreadsPerCore",document.config.threads_per_core) + get_field2(" ThreadsPerCore",document.config.threads_per_core) +
" State=UNKNOWN <br>" + " State=UNKNOWN <br>" +
"PartitionName=" + document.config.partition_name.value + "PartitionName=" + document.config.partition_name.value +
" Nodes=" + document.config.node_name.value + " Nodes=" + document.config.node_name.value +
" Default=YES"+ " Default=YES"+
" MaxTime=" + document.config.max_time.value + " MaxTime=" + document.config.max_time.value +
" State=UP" " State=UP"
//scroll(0,0); //scroll(0,0);
//var popup = document.getElementById('out_box'); //var popup = document.getElementById('out_box');
//popup.innerHTML = "<a href='javascript:hide_box();'>close</a><br>"; //popup.innerHTML = "<a href='javascript:hide_box();'>close</a><br>";
//popup.innerHTML += "#BEGIN SLURM.CONF FILE<br><br>"; //popup.innerHTML += "#BEGIN SLURM.CONF FILE<br><br>";
//popup.innerHTML += printme; //popup.innerHTML += printme;
//popup.innerHTML += "<br><br>#END SLURM.CONF FILE<br>"; //popup.innerHTML += "<br><br>#END SLURM.CONF FILE<br>";
//popup.innerHTML += "<a href='javascript:hide_box();'>close</a>"; //popup.innerHTML += "<a href='javascript:hide_box();'>close</a>";
//popup.style.visibility = 'visible'; //popup.style.visibility = 'visible';
// OLD CODE // OLD CODE
...@@ -325,29 +325,29 @@ function displayfile() ...@@ -325,29 +325,29 @@ function displayfile()
<BODY> <BODY>
<FORM name=config> <FORM name=config>
<H1>SLURM Version @SLURM_MAJOR@.@SLURM_MINOR@ Configuration Tool</H1> <H1>SLURM Version @SLURM_MAJOR@.@SLURM_MINOR@ Configuration Tool</H1>
<P>This form can be used to create a SLURM configuration file with <P>This form can be used to create a SLURM configuration file with
you controlling many of the important configuration parameters.</P> you controlling many of the important configuration parameters.</P>
<P><B>This tool supports SLURM version @SLURM_MAJOR@.@SLURM_MINOR@ only.</B> <P><B>This tool supports SLURM version @SLURM_MAJOR@.@SLURM_MINOR@ only.</B>
Configuration files for other versions of SLURM should be built Configuration files for other versions of SLURM should be built
using the tool distributed with it in <i>doc/html/configurator.html</i>. using the tool distributed with it in <i>doc/html/configurator.html</i>.
Some parameters will be set to default values, but you can Some parameters will be set to default values, but you can
manually edit the resulting <I>slurm.conf</I> as desired manually edit the resulting <I>slurm.conf</I> as desired
for greater flexibility. See <I>man slurm.conf</I> for more for greater flexibility. See <I>man slurm.conf</I> for more
details about the configuration parameters.</P> details about the configuration parameters.</P>
<P>Note the while SLURM daemons create log files and other files as needed, <P>Note the while SLURM daemons create log files and other files as needed,
it treats the lack of parent directories as a fatal error. it treats the lack of parent directories as a fatal error.
This prevents the daemons from running if critical file systems are This prevents the daemons from running if critical file systems are
not mounted and will minimize the risk of cold-starting (starting not mounted and will minimize the risk of cold-starting (starting
without preserving jobs).</P> without preserving jobs).</P>
<P>Note that this configuration file must be installed on all nodes <P>Note that this configuration file must be installed on all nodes
in your cluster.</P> in your cluster.</P>
<P>After you have filled in the fields of interest, use the <P>After you have filled in the fields of interest, use the
"Submit" button on the bottom of the page to build the <I>slurm.conf</I> "Submit" button on the bottom of the page to build the <I>slurm.conf</I>
file. It will appear on your web browser. Save the file in text format file. It will appear on your web browser. Save the file in text format
as <I>slurm.conf</I> for use by SLURM. as <I>slurm.conf</I> for use by SLURM.
<P>For more information about SLURM, see <P>For more information about SLURM, see
...@@ -360,34 +360,34 @@ Define the hostname of the computer on which the SLURM controller and ...@@ -360,34 +360,34 @@ Define the hostname of the computer on which the SLURM controller and
optional backup controller will execute. You can also specify addresses optional backup controller will execute. You can also specify addresses
of these computers if desired (defaults to their hostnames). of these computers if desired (defaults to their hostnames).
The IP addresses can be either numeric IP addresses or names. The IP addresses can be either numeric IP addresses or names.
Hostname values should should not be the fully qualified domain Hostname values should should not be the fully qualified domain
name (e.g. use <I>linux</I> rather than <I>linux.llnl.gov</I>). name (e.g. use <I>linux</I> rather than <I>linux.llnl.gov</I>).
<P> <P>
<input type="text" name="control_machine" value="linux0"> <B>ControlMachine</B>: <input type="text" name="control_machine" value="linux0"> <B>ControlMachine</B>:
Master Controller Hostname Master Controller Hostname
<P> <P>
<input type="text" name="control_addr"> <B>ControlAddr</B>: Master Controller <input type="text" name="control_addr"> <B>ControlAddr</B>: Master Controller
Address (optional) Address (optional)
<P> <P>
<input type="text" name="backup_controller"> <B>BackupController</B>: Backup <input type="text" name="backup_controller"> <B>BackupController</B>: Backup
Controller Hostname (optional) Controller Hostname (optional)
<P> <P>
<input type="text" name="backup_addr"> <B>BackupAddr</B>: Backup Controller <input type="text" name="backup_addr"> <B>BackupAddr</B>: Backup Controller
Address (optional) Address (optional)
<P> <P>
<H2>Compute Machines</H2> <H2>Compute Machines</H2>
Define the machines on which user applications can run. Define the machines on which user applications can run.
You can also specify addresses of these computers if desired You can also specify addresses of these computers if desired
(defaults to their hostnames). (defaults to their hostnames).
Only a few of the possible parameters associated with the nodes will Only a few of the possible parameters associated with the nodes will
be set by this tool, but many others are available. be set by this tool, but many others are available.
Executing the command <i>slurmd -C</i> on each compute node will print its Executing the command <i>slurmd -C</i> on each compute node will print its
physical configuration (sockets, cores, real memory size, etc.), which physical configuration (sockets, cores, real memory size, etc.), which
can be used in constructing the <i>slurm.conf</i> file. can be used in constructing the <i>slurm.conf</i> file.
All of the nodes will be placed into a single partition (or queue) All of the nodes will be placed into a single partition (or queue)
with global access. Many options are available to group nodes into with global access. Many options are available to group nodes into
partitions with a wide variety of configuration parameters. partitions with a wide variety of configuration parameters.
Manually edit the <i>slurm.conf</i> produced to exercise these options. Manually edit the <i>slurm.conf</i> produced to exercise these options.
Node names and addresses may be specified using a numeric range specification. Node names and addresses may be specified using a numeric range specification.
...@@ -398,17 +398,17 @@ Compute nodes ...@@ -398,17 +398,17 @@ Compute nodes
<input type="text" name="node_addr"> <B>NodeAddr</B>: Compute node addresses <input type="text" name="node_addr"> <B>NodeAddr</B>: Compute node addresses
(optional) (optional)
<P> <P>
<input type="text" name="partition_name" value="debug"> <B>PartitionName</B>: <input type="text" name="partition_name" value="debug"> <B>PartitionName</B>:
Name of the one partition to be created Name of the one partition to be created
<P> <P>
<input type="text" name="max_time" value="INFINITE"> <B>MaxTime</B>: <input type="text" name="max_time" value="INFINITE"> <B>MaxTime</B>:
Maximum time limit of jobs in minutes or INFINITE Maximum time limit of jobs in minutes or INFINITE
<P> <P>
The following parameters describe a node's configuration. The following parameters describe a node's configuration.
Set a value for <B>Procs</B>. Set a value for <B>Procs</B>.
The other parameters are optional, but provide more control over scheduled resources: The other parameters are optional, but provide more control over scheduled resources:
<P> <P>
<input type="text" name="procs" value="1"> <B>Procs</B>: Count of processors <input type="text" name="procs" value="1"> <B>Procs</B>: Count of processors
on each compute node. on each compute node.
If Procs is omitted, it will be inferred from: If Procs is omitted, it will be inferred from:
Sockets, CoresPerSocket, and ThreadsPerCore. Sockets, CoresPerSocket, and ThreadsPerCore.
...@@ -429,15 +429,15 @@ the logical number of processors per socket. ...@@ -429,15 +429,15 @@ the logical number of processors per socket.
<B>ThreadsPerCore</B>: <B>ThreadsPerCore</B>:
Number of logical threads in a single physical core. Number of logical threads in a single physical core.
<P> <P>
<input type="text" name="memory" value=""> <B>RealMemory</B>: Amount <input type="text" name="memory" value=""> <B>RealMemory</B>: Amount
of real memory. This parameter is required when specifying Memory as a of real memory. This parameter is required when specifying Memory as a
consumable resource with the select/cons_res plug-in. See below consumable resource with the select/cons_res plug-in. See below
under Resource Selection. under Resource Selection.
<P> <P>
<H2>SLURM User</H2> <H2>SLURM User</H2>
The SLURM controller (slurmctld) can run without elevated privileges, The SLURM controller (slurmctld) can run without elevated privileges,
so it is recommended that a user "slurm" be created for it. For testing so it is recommended that a user "slurm" be created for it. For testing
purposes any user name can be used. purposes any user name can be used.
<P> <P>
<input type="text" name="slurm_user" value="slurm"> <B>SlurmUser</B> <input type="text" name="slurm_user" value="slurm"> <B>SlurmUser</B>
...@@ -453,7 +453,7 @@ entries to prevent this from happening by setting ...@@ -453,7 +453,7 @@ entries to prevent this from happening by setting
<B>CacheGroups</B>=1. Reconfiguring ("scontrol reconfig") with <B>CacheGroups</B>=1. Reconfiguring ("scontrol reconfig") with
<B>CacheGroups</B>=0 will cause slurmd to purge the cache. Select one <B>CacheGroups</B>=0 will cause slurmd to purge the cache. Select one
value for <B>CacheGroups</B>:<BR> value for <B>CacheGroups</B>:<BR>
<input type="radio" name="cache_groups" value="0" checked> <input type="radio" name="cache_groups" value="0" checked>
<B>0</B>: for normal environment.<BR> <B>0</B>: for normal environment.<BR>
<input type="radio" name="cache_groups" value="1"> <input type="radio" name="cache_groups" value="1">
<B>1</B>: for slow NIS environment. <B>1</B>: for slow NIS environment.
...@@ -464,10 +464,10 @@ after making any changes to system password or group databases. ...@@ -464,10 +464,10 @@ after making any changes to system password or group databases.
<P> <P>
<H2>SLURM Port Numbers</H2> <H2>SLURM Port Numbers</H2>
The SLURM controller (slurmctld) requires a unique port for communications The SLURM controller (slurmctld) requires a unique port for communications
as do the SLURM compute node daemons (slurmd). If not set, slurm ports as do the SLURM compute node daemons (slurmd). If not set, slurm ports
are set by checking for an entry in <I>/etc/services</I> and if that are set by checking for an entry in <I>/etc/services</I> and if that
fails by using an interval default set at SLURM build time. fails by using an interval default set at SLURM build time.
<P> <P>
<input type="text" name="slurmctld_port" value="6817"> <B>SlurmctldPort</B> <input type="text" name="slurmctld_port" value="6817"> <B>SlurmctldPort</B>
<P> <P>
...@@ -477,21 +477,21 @@ fails by using an interval default set at SLURM build time. ...@@ -477,21 +477,21 @@ fails by using an interval default set at SLURM build time.
<H2>Authentication and Security</H2> <H2>Authentication and Security</H2>
Define the method used for authenticating communicating between SLURM components.<BR> Define the method used for authenticating communicating between SLURM components.<BR>
Select one value for <B>AuthType</B>:<BR> Select one value for <B>AuthType</B>:<BR>
<input type="radio" name="auth_type" value="none"> <B>None</B>: No authentication, <input type="radio" name="auth_type" value="none"> <B>None</B>: No authentication,
not recommended production use<br> not recommended production use<br>
<input type="radio" name="auth_type" value="authd"> <B>Authd</B>: Brent Chun's <input type="radio" name="auth_type" value="authd"> <B>Authd</B>: Brent Chun's
<A href="http://www.theether.org/authd/">authd</A><BR> <A href="http://www.theether.org/authd/">authd</A><BR>
<input type="radio" name="auth_type" value="munge" checked> <B>Munge</B>: LLNL's <input type="radio" name="auth_type" value="munge" checked> <B>Munge</B>: LLNL's
<A href="http://home.gna.org/munge/">Munge</A><BR> <A href="http://home.gna.org/munge/">Munge</A><BR>
<P> <P>
Library used for job step cryptographic signature generation.<BR> Library used for job step cryptographic signature generation.<BR>
Select one value for <B>CryptoType</B>:<BR> Select one value for <B>CryptoType</B>:<BR>
<input type="radio" name="crypto_type" value="munge" checked><B>Munge</B>: LLNL's <input type="radio" name="crypto_type" value="munge" checked><B>Munge</B>: LLNL's
<A href="http://home.gna.org/munge/">Munge</A> (has Gnu Public License)<BR> <A href="http://home.gna.org/munge/">Munge</A> (has Gnu Public License)<BR>
<input type="radio" name="crypto_type" value="openssl"> <B>OpenSSL</B>: <input type="radio" name="crypto_type" value="openssl"> <B>OpenSSL</B>:
<A href="http://www.openssl.org/">OpenSSL</A> <A href="http://www.openssl.org/">OpenSSL</A>
<P> <P>
Define the location of public and private keys used by SLURM's Define the location of public and private keys used by SLURM's
cryptographic signature generation plugin (CryptoType).<br> cryptographic signature generation plugin (CryptoType).<br>
<b>These values are only used if CryptoType=OpenSSL.</b><br> <b>These values are only used if CryptoType=OpenSSL.</b><br>
These files need to be generated by the SLURM administrator. These files need to be generated by the SLURM administrator.
...@@ -503,8 +503,8 @@ Specify fully qualified pathnames. ...@@ -503,8 +503,8 @@ Specify fully qualified pathnames.
<P> <P>
<H2>State Preservation</H2> <H2>State Preservation</H2>
Define the location of a directory where the slurmctld daemon saves its state. Define the location of a directory where the slurmctld daemon saves its state.
This should be a fully qualified pathname which can be read and written to This should be a fully qualified pathname which can be read and written to
by the SLURM user on both the control machine and backup controller (if configured). by the SLURM user on both the control machine and backup controller (if configured).
The location of a directory where slurmd saves state should also be defined. The location of a directory where slurmd saves state should also be defined.
This must be a unique directory on each compute server (local disk). This must be a unique directory on each compute server (local disk).
...@@ -520,37 +520,37 @@ Define when a non-responding (DOWN) node is returned to service.<BR> ...@@ -520,37 +520,37 @@ Define when a non-responding (DOWN) node is returned to service.<BR>
Select one value for <B>ReturnToService</B>:<BR> Select one value for <B>ReturnToService</B>:<BR>
<input type="radio" name="return_to_service" value="0"> <input type="radio" name="return_to_service" value="0">
<B>0</B>: When explicitly restored to service by an administrator.<BR> <B>0</B>: When explicitly restored to service by an administrator.<BR>
<input type="radio" name="return_to_service" value="1" checked> <input type="radio" name="return_to_service" value="1" checked>
<B>1</B>: Automatically, when slurmd daemon registers with valid configuration<BR> <B>1</B>: Automatically, when slurmd daemon registers with valid configuration<BR>
<P> <P>
<H2>Scheduling</H2> <H2>Scheduling</H2>
Define the mechanism to be used for controlling job ordering.<BR> Define the mechanism to be used for controlling job ordering.<BR>
Select one value for <B>SchedulerType</B>:<BR> Select one value for <B>SchedulerType</B>:<BR>
<input type="radio" name="sched_type" value="builtin"> <B>Builtin</B>: First-In <input type="radio" name="sched_type" value="builtin"> <B>Builtin</B>: First-In
First-Out (FIFO)<BR> First-Out (FIFO)<BR>
<input type="radio" name="sched_type" value="backfill" checked> <B>Backfill</B>: <input type="radio" name="sched_type" value="backfill" checked> <B>Backfill</B>:
FIFO with backfill<BR> FIFO with backfill<BR>
<input type="radio" name="sched_type" value="gang"> <B>Gang</B>: Gang scheduling <input type="radio" name="sched_type" value="gang"> <B>Gang</B>: Gang scheduling
(time-slicing for parallel jobs)<BR> (time-slicing for parallel jobs)<BR>
<input type="radio" name="sched_type" value="wiki"> <B>Wiki</B>: Wiki interface <input type="radio" name="sched_type" value="wiki"> <B>Wiki</B>: Wiki interface
to Maui (configuration parameter <B>SchedulerPort</B> must specified)<BR> to Maui (configuration parameter <B>SchedulerPort</B> must specified)<BR>
<input type="radio" name="sched_type" value="wiki2"> <B>Wiki2</B>: Wiki interface <input type="radio" name="sched_type" value="wiki2"> <B>Wiki2</B>: Wiki interface
to Moab (configuration parameter <B>SchedulerPort</B> must specified)<BR> to Moab (configuration parameter <B>SchedulerPort</B> must specified)<BR>
<P> <P>
<input type="text" name="scheduler_port" value="7321"> <B>SchedulerPort</B>: scheduler <input type="text" name="scheduler_port" value="7321"> <B>SchedulerPort</B>: scheduler
communications port (used by Wiki and Wiki2 only) communications port (used by Wiki and Wiki2 only)
<P> <P>
Define what node configuration (sockets, cores, memory, etc.) should be used. Define what node configuration (sockets, cores, memory, etc.) should be used.
Using values defined in the configuration file will provide faster scheduling.<BR> Using values defined in the configuration file will provide faster scheduling.<BR>
Select one value for <B>FastSchedule</B>:<BR> Select one value for <B>FastSchedule</B>:<BR>
<input type="radio" name="fast_schedule" value="1" checked> <input type="radio" name="fast_schedule" value="1" checked>
<B>1</B>: Use node configuration values defined in configuration file<BR> <B>1</B>: Use node configuration values defined in configuration file<BR>
<input type="radio" name="fast_schedule" value="0"> <input type="radio" name="fast_schedule" value="0">
<B>0</B>: Use node configuration values actually found on each node <B>0</B>: Use node configuration values actually found on each node
(if configured with with gang scheduling or allocation of individual (if configured with with gang scheduling or allocation of individual
processors to jobs rather than only whole node allocations, the processor processors to jobs rather than only whole node allocations, the processor
count on the node should match the configured value to avoid having extra count on the node should match the configured value to avoid having extra
processors left idle) processors left idle)
<P> <P>
...@@ -558,46 +558,46 @@ processors left idle) ...@@ -558,46 +558,46 @@ processors left idle)
Define the node interconnect used.<BR> Define the node interconnect used.<BR>
Select one value for <B>SwitchType</B>:<BR> Select one value for <B>SwitchType</B>:<BR>
<input type="radio" name="switch_type" value="elan"> <B>Elan</B>: Quadrics Elan3 or Elan4<BR> <input type="radio" name="switch_type" value="elan"> <B>Elan</B>: Quadrics Elan3 or Elan4<BR>
<input type="radio" name="switch_type" value="federation"> <B>Federation</B>: IBM <input type="radio" name="switch_type" value="federation"> <B>Federation</B>: IBM
Federation Switch<BR> Federation Switch<BR>
<input type="radio" name="switch_type" value="none" checked> <B>None</B>: No special <input type="radio" name="switch_type" value="none" checked> <B>None</B>: No special
handling required (InfiniBand, Myrinet, Ethernet, etc.)<BR> handling required (InfiniBand, Myrinet, Ethernet, etc.)<BR>
<P> <P>
<H2>Default MPI Type</H2> <H2>Default MPI Type</H2>
Specify the type of MPI to be used by default. SLURM will configure environment Specify the type of MPI to be used by default. SLURM will configure environment
variables accordingly. Users can over-ride this specification with an srun option.<BR> variables accordingly. Users can over-ride this specification with an srun option.<BR>
Select one value for <B>MpiDefault</B>:<BR> Select one value for <B>MpiDefault</B>:<BR>
<input type="radio" name="mpi_default" value="mpichgm"> <B>MPICH-GM</B><BR> <input type="radio" name="mpi_default" value="mpichgm"> <B>MPICH-GM</B><BR>
<input type="radio" name="mpi_default" value="mpichmx"> <B>MPICH-MX</B><BR> <input type="radio" name="mpi_default" value="mpichmx"> <B>MPICH-MX</B><BR>
<input type="radio" name="mpi_default" value="mpich1_p4"> <B>MPICH1-P4</B><BR> <input type="radio" name="mpi_default" value="mpich1_p4"> <B>MPICH1-P4</B><BR>
<input type="radio" name="mpi_default" value="mpich1_shmem"> <B>MPICH1-SHMEM</B>: <input type="radio" name="mpi_default" value="mpich1_shmem"> <B>MPICH1-SHMEM</B>:
This also works for MVAPICH-SHMEM.<BR> This also works for MVAPICH-SHMEM.<BR>
<input type="radio" name="mpi_default" value="mvapich"> <B>MVAPICH</B><BR> <input type="radio" name="mpi_default" value="mvapich"> <B>MVAPICH</B><BR>
<input type="radio" name="mpi_default" value="none" checked> <B>None</B>: <input type="radio" name="mpi_default" value="none" checked> <B>None</B>:
+This works for most other MPI types including MPICH2, LAM MPI and Open MPI.<BR> +This works for most other MPI types including MPICH2, LAM MPI and Open MPI.<BR>
<P> <P>
<H2>Process Tracking</H2> <H2>Process Tracking</H2>
Define the algorithm used to identify which processes are associated with a Define the algorithm used to identify which processes are associated with a
given job. This is used signal, kill, and account for the processes associated given job. This is used signal, kill, and account for the processes associated
with a job step.<BR> with a job step.<BR>
Select one value for <B>ProctrackType</B>:<BR> Select one value for <B>ProctrackType</B>:<BR>
<input type="radio" name="proctrack_type" value="aix"> <B>AIX</B>: Use AIX kernel <input type="radio" name="proctrack_type" value="aix"> <B>AIX</B>: Use AIX kernel
extension, recommended for AIX systems<BR> extension, recommended for AIX systems<BR>
<input type="radio" name="proctrack_type" value="cgroup"> <B>Cgroup</B>: Use <input type="radio" name="proctrack_type" value="cgroup"> <B>Cgroup</B>: Use
Linux <i>cgroups</i> to create a job container and track processes. Linux <i>cgroups</i> to create a job container and track processes.
Build a <i>cgroup.conf</i> file as well<BR> Build a <i>cgroup.conf</i> file as well<BR>
<input type="radio" name="proctrack_type" value="pgid" checked> <B>Pgid</B>: Use Unix <input type="radio" name="proctrack_type" value="pgid" checked> <B>Pgid</B>: Use Unix
Process Group ID, processes changing their process group ID can escape from SLURM Process Group ID, processes changing their process group ID can escape from SLURM
control<BR> control<BR>
<input type="radio" name="proctrack_type" value="linuxproc"> <B>LinuxProc</B>: Use <input type="radio" name="proctrack_type" value="linuxproc"> <B>LinuxProc</B>: Use
parent process ID records, required for MPICH-GM use, processes can escape parent process ID records, required for MPICH-GM use, processes can escape
from SLURM control<BR> from SLURM control<BR>
<input type="radio" name="proctrack_type" value="rms"> <B>RMS</B>: Use Quadrics <input type="radio" name="proctrack_type" value="rms"> <B>RMS</B>: Use Quadrics
kernel infrastructure, recommended for systems where this is available<BR> kernel infrastructure, recommended for systems where this is available<BR>
<input type="radio" name="proctrack_type" value="sgi_job"> <B>SGI's PAGG <input type="radio" name="proctrack_type" value="sgi_job"> <B>SGI's PAGG
module</B>: Use <A HREF="http://oss.sgi.com/projects/pagg/">SGI's Process module</B>: Use <A HREF="http://oss.sgi.com/projects/pagg/">SGI's Process
Aggregates (PAGG) kernel module</A>, recommended where available<BR> Aggregates (PAGG) kernel module</A>, recommended where available<BR>
<P> <P>
...@@ -609,17 +609,17 @@ Select one value for <B>SelectType</B>:<BR> ...@@ -609,17 +609,17 @@ Select one value for <B>SelectType</B>:<BR>
<DL> <DL>
<DL> <DL>
<DT><B>SelectTypeParameters</B> (As used by <I>SelectType=Cons_res</I> only): <DT><B>SelectTypeParameters</B> (As used by <I>SelectType=Cons_res</I> only):
<DD> Note: The -E extension for sockets, cores, and threads <DD> Note: The -E extension for sockets, cores, and threads
are ignored within the node allocation mechanism are ignored within the node allocation mechanism
when CR_CPU or CR_CPU_MEMORY is selected. when CR_CPU or CR_CPU_MEMORY is selected.
They are considered to compute the total number of They are considered to compute the total number of
tasks when -n is not specified tasks when -n is not specified
<DD> Note: CR_MEMORY assumes MaxShare value of one of higher <DD> Note: CR_MEMORY assumes MaxShare value of one of higher
<DT> <input type="radio" name="cons_res_params" value="CR_CPU" checked <DT> <input type="radio" name="cons_res_params" value="CR_CPU" checked
onClick="javascript:set_select_type(this, 'cons_res')"> onClick="javascript:set_select_type(this, 'cons_res')">
<B>CR_CPU</B>: (default) <B>CR_CPU</B>: (default)
CPUs as consumable resources. CPUs as consumable resources.
<DD> No notion of sockets, cores, or threads. <DD> No notion of sockets, cores, or threads.
On a multi-core system, cores will be considered CPUs. On a multi-core system, cores will be considered CPUs.
On a multi-core/hyperthread system, threads will be considered CPUs. On a multi-core/hyperthread system, threads will be considered CPUs.
On a single-core systems CPUs are CPUs. ;-) On a single-core systems CPUs are CPUs. ;-)
...@@ -628,10 +628,10 @@ Select one value for <B>SelectType</B>:<BR> ...@@ -628,10 +628,10 @@ Select one value for <B>SelectType</B>:<BR>
<B>CR_Socket</B>: Sockets as a consumable resource. <B>CR_Socket</B>: Sockets as a consumable resource.
<DT> <input type="radio" name="cons_res_params" value="CR_Core" <DT> <input type="radio" name="cons_res_params" value="CR_Core"
onClick="javascript:set_select_type(this)"> onClick="javascript:set_select_type(this)">
<B>CR_Core</B>: Cores as a consumable resource. <B>CR_Core</B>: Cores as a consumable resource.
<DT> <input type="radio" name="cons_res_params" value="CR_Memory" <DT> <input type="radio" name="cons_res_params" value="CR_Memory"
onClick="javascript:set_select_type(this)"> onClick="javascript:set_select_type(this)">
<B>CR_Memory</B>: Memory as a consumable resource. <B>CR_Memory</B>: Memory as a consumable resource.
<DD> Note: CR_Memory assumes MaxShare value of one of higher <DD> Note: CR_Memory assumes MaxShare value of one of higher
<DT> <input type="radio" name="cons_res_params" value="CR_CPU_Memory" <DT> <input type="radio" name="cons_res_params" value="CR_CPU_Memory"
onClick="javascript:set_select_type(this)"> onClick="javascript:set_select_type(this)">
...@@ -648,19 +648,19 @@ Select one value for <B>SelectType</B>:<BR> ...@@ -648,19 +648,19 @@ Select one value for <B>SelectType</B>:<BR>
</DL> </DL>
</DL> </DL>
<input type="radio" name="select_type" value="linear" checked> <input type="radio" name="select_type" value="linear" checked>
<B>Linear</B>: Node-base <B>Linear</B>: Node-base
resource allocation, does not manage individual processor allocation<BR> resource allocation, does not manage individual processor allocation<BR>
<input type="radio" name="select_type" value="bluegene"> <input type="radio" name="select_type" value="bluegene">
<B>BlueGene</B>: For IBM Blue Gene systems only<BR> <B>BlueGene</B>: For IBM Blue Gene systems only<BR>
<P> <P>
<H2>Task Launch</H2> <H2>Task Launch</H2>
Define a task launch plugin. This may be used to Define a task launch plugin. This may be used to
provide resource management within a node (e.g. pinning provide resource management within a node (e.g. pinning
tasks to specific processors). tasks to specific processors).
Select one value for <B>TaskPlugin</B>:<BR> Select one value for <B>TaskPlugin</B>:<BR>
<input type="radio" name="task_plugin" value="none" checked> <B>None</B>: No task launch actions<BR> <input type="radio" name="task_plugin" value="none" checked> <B>None</B>: No task launch actions<BR>
<input type="radio" name="task_plugin" value="affinity"> <B>Affinity</B>: <input type="radio" name="task_plugin" value="affinity"> <B>Affinity</B>:
CPU affinity support CPU affinity support
(see srun man pages for the --cpu_bind, --mem_bind, and -E options) (see srun man pages for the --cpu_bind, --mem_bind, and -E options)
<DL><DL> <DL><DL>
...@@ -676,7 +676,7 @@ CPU affinity support ...@@ -676,7 +676,7 @@ CPU affinity support
<H2>Prolog and Epilog</H2> <H2>Prolog and Epilog</H2>
<P> <P>
<B>Prolog/Epilog</B>: Fully qualified path that will be executed as <B>Prolog/Epilog</B>: Fully qualified path that will be executed as
root on every node of a user's job before the job's tasks root on every node of a user's job before the job's tasks
will be initiated there and after that job has terminated. will be initiated there and after that job has terminated.
These parameters are optional. These parameters are optional.
...@@ -686,8 +686,8 @@ These parameters are optional. ...@@ -686,8 +686,8 @@ These parameters are optional.
</DL> </DL>
<P> <P>
<B>SrunProlog/Epilog</B>: Fully qualified path to be executed by srun at <B>SrunProlog/Epilog</B>: Fully qualified path to be executed by srun at
job step initiation and termination. These parameters may be overridden by job step initiation and termination. These parameters may be overridden by
srun's --prolog and --epilog options srun's --prolog and --epilog options
These parameters are optional. These parameters are optional.
<DL> <DL>
...@@ -696,7 +696,7 @@ These parameters are optional. ...@@ -696,7 +696,7 @@ These parameters are optional.
</DL> </DL>
<P> <P>
<B>TaskProlog/Epilog</B>: Fully qualified path to be executed as the user <B>TaskProlog/Epilog</B>: Fully qualified path to be executed as the user
before each task begins execution and after each task terminates. before each task begins execution and after each task terminates.
These parameters are optional. These parameters are optional.
<DL> <DL>
...@@ -705,9 +705,9 @@ These parameters are optional. ...@@ -705,9 +705,9 @@ These parameters are optional.
</DL> </DL>
<H2>Event Logging</H2> <H2>Event Logging</H2>
Slurmctld and slurmd daemons can each be configured with different Slurmctld and slurmd daemons can each be configured with different
levels of logging verbosity from 0 (quiet) to 7 (extremely verbose). levels of logging verbosity from 0 (quiet) to 7 (extremely verbose).
Each may also be configured to use debug files. Use fully qualified Each may also be configured to use debug files. Use fully qualified
pathnames for the files. pathnames for the files.
<P> <P>
<input type="text" name="slurmctld_debug" value="3"> <B>SlurmctldDebug</B> (0 to 7) <input type="text" name="slurmctld_debug" value="3"> <B>SlurmctldDebug</B> (0 to 7)
...@@ -716,47 +716,47 @@ pathnames for the files. ...@@ -716,47 +716,47 @@ pathnames for the files.
<P> <P>
<input type="text" name="slurmd_debug" value="3"> <B>SlurmdDebug</B> (0 to 7) <input type="text" name="slurmd_debug" value="3"> <B>SlurmdDebug</B> (0 to 7)
<P> <P>
<input type="text" name="slurmd_logfile" value=""> <B>SlurmdLogFile</B> (default is none, <input type="text" name="slurmd_logfile" value=""> <B>SlurmdLogFile</B> (default is none,
log goes to syslog, string "%h" in name gets replaced with hostname) log goes to syslog, string "%h" in name gets replaced with hostname)
<P> <P>
<H2>Job Completion Logging</H2> <H2>Job Completion Logging</H2>
Define the job completion logging mechanism to be used.<BR> Define the job completion logging mechanism to be used.<BR>
Select one value for <B>JobCompType</B>:<BR> Select one value for <B>JobCompType</B>:<BR>
<input type="radio" name="job_comp_type" value="none" checked> <B>None</B>: <input type="radio" name="job_comp_type" value="none" checked> <B>None</B>:
No job completion logging<BR> No job completion logging<BR>
<input type="radio" name="job_comp_type" value="filetxt"> <B>FileTxt</B>: <input type="radio" name="job_comp_type" value="filetxt"> <B>FileTxt</B>:
Write job completion status to a text file<BR> Write job completion status to a text file<BR>
<input type="radio" name="job_comp_type" value="script"> <B>Script</B>: <input type="radio" name="job_comp_type" value="script"> <B>Script</B>:
Use an arbitrary script to log job completion<BR> Use an arbitrary script to log job completion<BR>
<input type="radio" name="job_comp_type" value="mysql"> <B>MySQL</B>: <input type="radio" name="job_comp_type" value="mysql"> <B>MySQL</B>:
Write completion status to a MySQL database<BR> Write completion status to a MySQL database<BR>
<input type="radio" name="job_comp_type" value="pgsql"> <B>PGSQL</B>: <input type="radio" name="job_comp_type" value="pgsql"> <B>PGSQL</B>:
Write completion status to a PostreSQL database<BR> Write completion status to a PostreSQL database<BR>
<input type="radio" name="job_comp_type" value="slurmdbd"> <B>SlurmDBD</B>: <input type="radio" name="job_comp_type" value="slurmdbd"> <B>SlurmDBD</B>:
Write completion status to Slurm a database daemon (serving multiple Slurm clusters) Write completion status to Slurm a database daemon (serving multiple Slurm clusters)
which will write to some database<BR> which will write to some database<BR>
<P> <P>
<input type="text" name="job_comp_loc" value=""> <B>JobCompLoc</B>: <input type="text" name="job_comp_loc" value=""> <B>JobCompLoc</B>:
This is the location of the text file to be written to (if JobCompType=filetst) This is the location of the text file to be written to (if JobCompType=filetst)
or the script to be run (if JobCompType=script) or database name (for other values or the script to be run (if JobCompType=script) or database name (for other values
of JobCompType). of JobCompType).
<p><b>Options below are for use with a database to specify where the database is running and how to connect to it</b><br> <p><b>Options below are for use with a database to specify where the database is running and how to connect to it</b><br>
<input type="text" name="job_comp_host" value=""> <B>JobCompHost</B>: <input type="text" name="job_comp_host" value=""> <B>JobCompHost</B>:
Host the database is running on for Job completion<br> Host the database is running on for Job completion<br>
<input type="text" name="job_comp_port" value=""> <B>JobCompPort</B>: <input type="text" name="job_comp_port" value=""> <B>JobCompPort</B>:
Port the database server is listening on for Job completion<br> Port the database server is listening on for Job completion<br>
<input type="text" name="job_comp_user" value=""> <B>JobCompUser</B>: <input type="text" name="job_comp_user" value=""> <B>JobCompUser</B>:
User we are to use to talk to the database for Job completion<br> User we are to use to talk to the database for Job completion<br>
<input type="text" name="job_comp_pass" value=""> <B>JobCompPass</B>: <input type="text" name="job_comp_pass" value=""> <B>JobCompPass</B>:
Password we are to use to talk to the database for Job completion<br> Password we are to use to talk to the database for Job completion<br>
<P> <P>
<H2>Job Accounting Gather</H2> <H2>Job Accounting Gather</H2>
SLURM accounts for resource use per job. System specifics can be polled SLURM accounts for resource use per job. System specifics can be polled
determined by system type<BR> determined by system type<BR>
Select one value for <B>JobAcctGatherType</B>:<BR> Select one value for <B>JobAcctGatherType</B>:<BR>
<input type="radio" name="job_acct_gather_type" value="none" checked> <B>None</B>: No <input type="radio" name="job_acct_gather_type" value="none" checked> <B>None</B>: No
job accounting<BR> job accounting<BR>
<input type="radio" name="job_acct_gather_type" value="aix"> <B>AIX</B>: Specifc <input type="radio" name="job_acct_gather_type" value="aix"> <B>AIX</B>: Specifc
AIX process table information gathered, use with AIX systems only<BR> AIX process table information gathered, use with AIX systems only<BR>
...@@ -769,39 +769,39 @@ polling interval in seconds. Zero disables periodic sampling.<BR> ...@@ -769,39 +769,39 @@ polling interval in seconds. Zero disables periodic sampling.<BR>
<H2>Job Accounting Storage</H2> <H2>Job Accounting Storage</H2>
Used with the Job Accounting Gather SLURM can store the accounting information in many different fashions. Fill in your systems choice here<BR> Used with the Job Accounting Gather SLURM can store the accounting information in many different fashions. Fill in your systems choice here<BR>
Select one value for <B>AccountingStorageType</B>:<BR> Select one value for <B>AccountingStorageType</B>:<BR>
<input type="radio" name="accounting_storage_type" value="none" checked> <B>None</B>: <input type="radio" name="accounting_storage_type" value="none" checked> <B>None</B>:
No job accounting storage<BR> No job accounting storage<BR>
<input type="radio" name="accounting_storage_type" value="filetxt"> <B>FileTxt</B>: <input type="radio" name="accounting_storage_type" value="filetxt"> <B>FileTxt</B>:
Write job accounting to a text file<BR> Write job accounting to a text file<BR>
<input type="radio" name="accounting_storage_type" value="gold"> <B>Gold</B>: <input type="radio" name="accounting_storage_type" value="gold"> <B>Gold</B>:
Write completion status to Gold database daemon which can securely Write completion status to Gold database daemon which can securely
save the data from many Slurm managed clusters into a common database<BR> save the data from many Slurm managed clusters into a common database<BR>
<input type="radio" name="accounting_storage_type" value="mysql"> <B>MySQL</B>: <input type="radio" name="accounting_storage_type" value="mysql"> <B>MySQL</B>:
Write job accounting to a MySQL database<BR> Write job accounting to a MySQL database<BR>
<input type="radio" name="accounting_storage_type" value="pgsql"> <B>PGSQL</B>: <input type="radio" name="accounting_storage_type" value="pgsql"> <B>PGSQL</B>:
Write job accounting to a PostreSQL database<BR> Write job accounting to a PostreSQL database<BR>
<input type="radio" name="accounting_storage_type" value="slurmdbd"> <B>SlurmDBD</B>: <input type="radio" name="accounting_storage_type" value="slurmdbd"> <B>SlurmDBD</B>:
Write job accounting to Slurm DBD (database daemon) which can securely Write job accounting to Slurm DBD (database daemon) which can securely
save the data from many Slurm managed clusters into a common database<BR> save the data from many Slurm managed clusters into a common database<BR>
<input type="text" name="accounting_storage_loc" value=""> <B>AccountingStorageLoc</B>: <input type="text" name="accounting_storage_loc" value=""> <B>AccountingStorageLoc</B>:
Location specification or database name. Location specification or database name.
This is the location of the text file to be written to (used by Log only). This is the location of the text file to be written to (used by Log only).
Use a fully qualified pathname. If using a database it is the name of the database you will use or create for the stored data.<br> Use a fully qualified pathname. If using a database it is the name of the database you will use or create for the stored data.<br>
<p><b>Options below are for use with a database to specify where the database is running and how to connect to it</b><br> <p><b>Options below are for use with a database to specify where the database is running and how to connect to it</b><br>
<input type="text" name="accounting_storage_host" value=""> <B>AccountingStorageHost</B>: <input type="text" name="accounting_storage_host" value=""> <B>AccountingStorageHost</B>:
Host the database is running on for Job Accounting<br> Host the database is running on for Job Accounting<br>
<input type="text" name="accounting_storage_port" value=""> <B>AccountingStoragePort</B>: <input type="text" name="accounting_storage_port" value=""> <B>AccountingStoragePort</B>:
Port the database server is listening on for Job Accounting<br> Port the database server is listening on for Job Accounting<br>
<input type="text" name="accounting_storage_user" value=""> <B>AccountingStorageUser</B>: <input type="text" name="accounting_storage_user" value=""> <B>AccountingStorageUser</B>:
User we are to use to talk to the database for Job Accounting<br> User we are to use to talk to the database for Job Accounting<br>
<input type="text" name="accounting_storage_pass" value=""> <B>AccountingStoragePass</B>: <input type="text" name="accounting_storage_pass" value=""> <B>AccountingStoragePass</B>:
Password we are to use to talk to the database for Job Accounting. Password we are to use to talk to the database for Job Accounting.
In the case of SlurmDBD, this will be an alternate socket name for use with a Munge In the case of SlurmDBD, this will be an alternate socket name for use with a Munge
daemon providing enterprise-wide authentication (while the default Munge socket daemon providing enterprise-wide authentication (while the default Munge socket
would provide cluster-wide authentication only).<br> would provide cluster-wide authentication only).<br>
<input type="text" name="cluster_name" value="cluster"> <B>ClusterName</B>: <input type="text" name="cluster_name" value="cluster"> <B>ClusterName</B>:
Name to be recorded in database for jobs from this cluster. Name to be recorded in database for jobs from this cluster.
This is important if a single database is used to record information This is important if a single database is used to record information
from multiple Slurm-managed clusters.<br> from multiple Slurm-managed clusters.<br>
<P> <P>
...@@ -811,47 +811,47 @@ Define the location into which we can record the daemon's process ID. ...@@ -811,47 +811,47 @@ Define the location into which we can record the daemon's process ID.
This is used for locate the appropriate daemon for signalling. This is used for locate the appropriate daemon for signalling.
Specify a specify the fully qualified pathname for the file. Specify a specify the fully qualified pathname for the file.
<P> <P>
<input type="text" name="slurmctld_pid_file" value="/var/run/slurmctld.pid"> <input type="text" name="slurmctld_pid_file" value="/var/run/slurmctld.pid">
<B>SlurmctldPidFile</B> <B>SlurmctldPidFile</B>
<P> <P>
<input type="text" name="slurmd_pid_file" value="/var/run/slurmd.pid"> <input type="text" name="slurmd_pid_file" value="/var/run/slurmd.pid">
<B>SlurmdPidFile</B> <B>SlurmdPidFile</B>
<P> <P>
<H2>Timers</H2> <H2>Timers</H2>
SLURM has a variety of timers to control when to consider a node DOWN, SLURM has a variety of timers to control when to consider a node DOWN,
when to purge job records, how long to give a job to gracefully terminate, etc. when to purge job records, how long to give a job to gracefully terminate, etc.
<P> <P>
<input type="text" name="slurmctld_timeout" value="120"> <input type="text" name="slurmctld_timeout" value="120">
<B>SlurmctldTimeout</B>: How many seconds the backup controller waits before <B>SlurmctldTimeout</B>: How many seconds the backup controller waits before
becoming the master controller becoming the master controller
<P> <P>
<input type="text" name="slurmd_timeout" value="300"> <input type="text" name="slurmd_timeout" value="300">
<B>SlurmdTimeout</B>: How many seconds the SLURM controller waits for the slurmd <B>SlurmdTimeout</B>: How many seconds the SLURM controller waits for the slurmd
to respond to a request before considering the node DOWN to respond to a request before considering the node DOWN
<P> <P>
<input type="text" name="inactive_limit" value="0"> <input type="text" name="inactive_limit" value="0">
<B>InactiveLimit</B>: How many seconds the SLURM controller waits for srun <B>InactiveLimit</B>: How many seconds the SLURM controller waits for srun
commands to respond before considering the job or job step inactive and commands to respond before considering the job or job step inactive and
terminating it. A value of zero indicates unlimited wait terminating it. A value of zero indicates unlimited wait
<P> <P>
<input type="text" name="min_job_age" value="300"> <input type="text" name="min_job_age" value="300">
<B>MinJobAge</B>: How many seconds the SLURM controller waits after a <B>MinJobAge</B>: How many seconds the SLURM controller waits after a
job terminates before purging its record. A record of the job will job terminates before purging its record. A record of the job will
persist in job completion and/or accounting records indefinitely, persist in job completion and/or accounting records indefinitely,
but will no longer be visible with the squeue command after puring but will no longer be visible with the squeue command after puring
<P> <P>
<input type="text" name="kill_wait" value="30"> <input type="text" name="kill_wait" value="30">
<B>KillWait</B>: How many seconds a job is given to gracefully terminate <B>KillWait</B>: How many seconds a job is given to gracefully terminate
after reaching its time limit and being sent SIGTERM before sending after reaching its time limit and being sent SIGTERM before sending
a SIGKILLL a SIGKILLL
<P> <P>
<input type="text" name="wait_time" value="0"> <input type="text" name="wait_time" value="0">
<B>WaitTime</B>: How many seconds after a job step's first task terminates <B>WaitTime</B>: How many seconds after a job step's first task terminates
before terminating all remaining tasks. A value of zero indicates unlimited wait before terminating all remaining tasks. A value of zero indicates unlimited wait
<P> <P>
<BR> <BR>
<BR> <BR>
<input type=button value="Submit" onClick="javascript:displayfile()"> <input type=button value="Submit" onClick="javascript:displayfile()">
<input type=reset value="Reset Form"> <input type=reset value="Reset Form">
...@@ -861,4 +861,3 @@ before terminating all remaining tasks. A value of zero indicates unlimited wait ...@@ -861,4 +861,3 @@ before terminating all remaining tasks. A value of zero indicates unlimited wait
<P class="footer">LLNL-WEB-402631<BR> <P class="footer">LLNL-WEB-402631<BR>
Last modified 25 August 2010</P> Last modified 25 August 2010</P>
</BODY> </BODY>
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
Nutzungsbedingungen | Datenschutzerklärung | Barrierefreiheitserklärung