1slurmd(8) Slurm Daemon slurmd(8)
2
6 slurmd - The compute node daemon for Slurm.
7
10 slurmd [OPTIONS...]
11
14 slurmd is the compute node daemon of Slurm. It monitors all tasks run‐
15 ning on the compute node , accepts work (tasks), launches tasks, and
16 kills running tasks upon request.
17 OPTIONS
19 -b Report node rebooted when daemon restarted. Used for testing
21 purposes.
22 -c Clear system locks as needed. This may be required if slurmd
25 terminated abnormally.
26 -C Print actual hardware configuration and exit. The format of out‐
29 put is the same as used in slurm.conf to describe a node's con‐
30 figuration plus its uptime.
31 --conf-server <host>[:<port>]
34 Comma-separated list of controllers, the first being the primary
35 slurmctld. A port can (optionally) be specified for each con‐
36 troller. These hosts are where the slurmd will fetch the config‐
37 uration from when running in "configless" mode.
38 -d <file>
41 Specify the fully qualified pathname to the slurmstepd program
42 to be used for shepherding user job steps. This can be useful
43 for testing purposes.
44 -D Run slurmd in the foreground. Error and debug messages will be
46 copied to stderr.
47 -f <file>
49 Read configuration from the specified file. See NOTES below.
50 -F[feature]
52 Start this node as a Dynamic Future node. It will try to match a
53 node definition with a state of FUTURE, optionally using the
54 specified feature to match the node definition.
55 -G Print Generic RESource (GRES) configuration (based upon
57 slurm.conf GRES merged with gres.conf contents for this node)
58 and exit.
59 -h Help; print a brief summary of command options.
61 -L <file>
63 Write log messages to the specified file.
64 -M Lock slurmd pages into system memory using mlockall (2) to dis‐
66 able paging of the slurmd process. This may help in cases where
67 nodes are marked DOWN during periods of heavy swap activity. If
68 the mlockall (2) system call is not available, an error will be
69 printed to the log and slurmd will continue as normal.
70 It is suggested to set LaunchParameters=slurmstepd_memlock in
72 slurm.conf(5) when setting -M.
73 -n <value>
76 Set the daemon's nice value to the specified value, typically a
77 negative number. Also note the PropagatePrioProcess configura‐
78 tion parameter.
79 -N <nodename>
82 Run the daemon with the given nodename. Used to emulate a larger
83 system with more than one slurmd daemon per node. Requires that
84 Slurm be built using the --enable-multiple-slurmd configure op‐
85 tion.
86 -s Change working directory of slurmd to SlurmdLogFile path if pos‐
89 sible, or to SlurmdSpoolDir otherwise. If both of them fail it
90 will fallback to /var/tmp.
91 -v Verbose operation. Multiple -v's increase verbosity.
94 -V, --version
96 Print version information and exit.
97
100 The following environment variables can be used to override settings
101 compiled into slurmd.
102 SLURM_CONF The location of the Slurm configuration file. This
104 is overridden by explicitly naming a configuration
105 file on the command line.
106
109 SIGTERM SIGINT
110 slurmd will shutdown cleanly, waiting for in-progress rollups to
111 finish.
112 SIGHUP Reloads the slurm configuration files, similar to 'scontrol re‐
114 configure'.
115 SIGUSR2
117 Reread the log level from the configs, and then reopen the log
118 file. This should be used when setting up logrotate(8).
119 SIGPIPE
121 This signal is explicitly ignored.
122
125 If slurmd is started with the -D option then the core file will be
126 written to the current working directory. Otherwise if SlurmdLogFile
127 is a fully qualified path name (starting with a slash), the core file
128 will be written to the same directory as the log file. Otherwise the
129 core file will be written to the SlurmSpoolDir directory, or
130 "/var/tmp/" as a last resort. If none of the above directories can be
131 written, no core file will be produced.
132
135 It may be useful to experiment with different slurmd specific configu‐
136 ration parameters using a distinct configuration file (e.g. timeouts).
137 However, this special configuration file will not be used by the slurm‐
138 ctld daemon or the Slurm programs, unless you specifically tell each of
139 them to use it. If you desire changing communication ports, the loca‐
140 tion of the temporary file system, or other parameters used by other
141 Slurm components, change the common configuration file, slurm.conf.
142 If you are using configless mode with a login node that runs a lot of
144 client commands, you may consider running slurmd on that machine so it
145 can manage a cached version of the configuration files. Otherwise, each
146 client command will use the DNS record to contact the controller and
147 get the configuration information, which could place additional load on
148 the controller.
149
152 Copyright (C) 2002-2007 The Regents of the University of California.
153 Copyright (C) 2008-2010 Lawrence Livermore National Security. Copy‐
154 right (C) 2010-2021 SchedMD LLC. Produced at Lawrence Livermore Na‐
155 tional Laboratory (cf, DISCLAIMER).
156 This file is part of Slurm, a resource management program. For de‐
158 tails, see <https://slurm.schedmd.com/>.
159 Slurm is free software; you can redistribute it and/or modify it under
161 the terms of the GNU General Public License as published by the Free
162 Software Foundation; either version 2 of the License, or (at your op‐
163 tion) any later version.
164 Slurm is distributed in the hope that it will be useful, but WITHOUT
166 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
167 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
168 for more details.
169
172 /etc/slurm.conf
173
176 slurm.conf(5), slurmctld(8)
177June 2021 Slurm Daemon slurmd(8)