1pbs_server_attributes(7B) PBS pbs_server_attributes(7B)
2
6 pbs_server_attributes - pbs server attributes
7
9 Server attributes can be read by any client; privilege is not required.
10 Most server attributes are alterable by a privileged client, run by a
11 user with administrator or operator privilege. Certain attributes
12 require the user to have full administrator privilege. The following
13 is a list of the server attributes.
14 accounting_keep_days
16 This defines the number of days that accounting files will be
17 kept. Default value: unset - pbs_server will never delete
18 accounting files Format: integer
19 acl_group_sloppy
21 This is a default value for the queue attribute of the same
22 name. Format: boolean, "TRUE", "True", "true", "Y", "y", "1",
23 "FALSE", "False", "false", "N", "n", "0"; default value: false
24 = disabled.
25 acl_host_enable
27 Attribute which when true directs the server to use the
28 acl_hosts access control lists. Requires full manager privi‐
29 lege to set or alter. Format: boolean, "TRUE", "True",
30 "true", "Y", "y", "1", "FALSE", "False", "false", "N", "n",
31 "0"; default value: false = disabled.
32 acl_hosts
34 List of hosts which may request services from this server.
35 This list contains the network name of the hosts. Local
36 requests, i.e. from the server's host itself, are aways
37 accepted even if the host is not included in the list. See
38 section 10.1, Authorization, in the PBS External Reference
39 Specification. Requires full manager privilege to set or
40 alter. Format: "[+|-]hostname.domain[,...]"; default value:
41 all hosts.
42 acl_logic_or
44 This is a default value for the queue attribute of the same
45 name. Format: boolean, "TRUE", "True", "true", "Y", "y", "1",
46 "FALSE", "False", "false", "N", "n", "0"; default value: false
47 = disabled.
48 acl_user_enable
50 Attribute which when true directs the server to use the server
51 level acl_users access list. Requires full manager privilege
52 to set or alter. Format: boolean (see acl_group_enable);
53 default value: disabled.
54 acl_users
56 List of users allowed or denied the ability to make any
57 requests of this server. See section 10.1, Authorization, in
58 the PBS External Reference Specification. If acl_user_enable
59 is set to true, only users listed in acl_users may submit to
60 or execute jobs in the queue. Requires full manager privilege
61 to set or alter. Format: "[+|-]user[@host][,...]"; default
62 value: all users allowed.
63 acl_roots
65 List of super users who may submit to and execute jobs at this
66 server. If the job execution id would be zero (0), then the
67 job owner, root@host, must be listed in this access control
68 list or the job is rejected. Format:
69 "[+|-]user[@host][,...]"; default value: no root jobs allowed.
70 allow_node_submit
72 Allow job submissions from compute nodes regardless of
73 ruserok(). Requires full manager privilege to set or alter.
74 Format: boolean; default value: disabled.
75 allow_proxy_user
77 Specifies users can proxy from one user to another. Proxy
78 requests will be either validated by ruserok() or by the
79 scheduler. Format: boolean; default value: false.
80 auto_node_np
82 Automatically configure a node's np value based on the ncpus
83 value from the status update. Requires full manager privilege
84 to set or alter. Format: boolean; default value: disabled.
85 clone_batch_delay
87 Number of seconds to delay between cloning a batch for a job
88 array. Format: integer; default value: 1.
89 clone_batch_size
91 Number of jobs to clone in a batch for a job array. Format:
92 integer; default value: 256.
93 credential_lifetime
95 The number of a seconds that a client connection may stay con‐
96 nected without re-authenticating. Default is 3600 seconds.
97 comment
99 A text string which may be set by the scheduler or other priv‐
100 ileged client to provide information to the batch system
101 users. Format: any string; default value: none.
102 default_node
104 A node specification to use if there is no other supplied
105 specification. This attribute is only used by servers where a
106 nodes file exist in the server_priv directory providing a list
107 of nodes to the server. If the nodes file does not exist,
108 this attribute is not set by default and is ignored if set.
109 The default value allows for jobs to share a single node.
110 Format: a node specification string; default value: 1#shared.
111 default_queue
113 The queue which is the target queue when a request does not
114 specify a queue name. Format: a queue name; default value:
115 none, must be set to an existing queue.
116 down_on_error
118 Set a node's state to "down" if MOM reports a message begin‐
119 ning with the string "ERROR". This might interfere with
120 moab's node error handling. See the HEALTH CHECK section in
121 pbs_mom(8B). This is an EXPERIMENTAL feature and may be
122 removed in the future. Format: boolean; default value: false.
123 disable_server_id_check
125 Makes it so the user for the job doesn't have to exist on the
126 server. The user must still exist on all the compute nodes or
127 the job will fail when it tries to execute. Format: boolean;
128 default value: false
129 extra_resc
131 Add additional string-type job resources. They have no effect
132 within TORQUE and are only advisible to the scheduler. They
133 can not be used for resources_default/min/max. Format: list;
134 default value: none.
135 job_force_cancel_time
137 If configured, number of seconds after a delete where a job
138 will be purged by the server. If not configured, no such thing
139 happens. Format: integer; default value: not used.
140 job_nanny
142 Enables the "job deletion nanny" feature. All job cancels
143 will create a repeating task that will resend KILL signals if
144 the initial job cancel failed. Further job cancels will be
145 rejected with the message "job cancel in progress." This is
146 useful for temporary failures with a job's execution node dur‐
147 ing a job delete request. It is possible that the job nanny
148 might interfere with job restarts, migrations, and checkpoint‐
149 ing. Format: boolean; default value: false.
150 job_start_timeout
152 Specifies the pbs_server to pbs_mom TCP socket timeout in sec‐
153 onds that is used when the pbs_server sends a job start to the
154 pbs_mom. It is useful when the mom has extra overhead
155 involved in starting jobs. If not specified then the
156 tcp_timeout value is used.
157 job_stat_rate
159 Moderates how often job stat requests will be issued from
160 pbs_server to the MOM daemons. If poll_jobs is unset or
161 false, then all jobs that haven't been updated in
162 job_stat_rate seconds will trigger a stat request. If
163 poll_jobs is true, then all jobs will be updated every
164 job_stat_rate.
165 (see poll_jobs) On active clusters, 60 or 120 might be rea‐
166 sonable. Default value: 45 seconds (PBS_RESTAT_JOB in
167 server_limits.h) Minimum value: 4 seconds (PBS_JOBSTAT_MIN in
168 server_limits.h)
169 keep_completed
171 Number of seconds to retain completed jobs in the C state.
172 This is overridden by the execution queue attribute of the
173 same name. Format: integer; default value: 0.
174 kill_delay
176 The amount of the time delay between the sending of SIGTERM
177 and SIGKILL when a qdel command is issued against a running
178 job. This is overridden by the execution queue attribute of
179 the same name. Format: integer seconds; default value: 2 sec‐
180 onds.
181 lock_file
183 Specifies the name and location of the lock file used to
184 determine which high availability server should be active.
185 If a full path is specified, it is used verbatim by TORQUE. If
186 a relative path is specified, TORQUE will prefix it with
187 $TORQUE_HOME/server_priv. Format: string; default value:
188 $TORQUE_HOME/server_priv/server.lock
189 lock_file_update_time
191 Specifies how often (in seconds) the thread will update the
192 lockfile. (for threaded high availability) Format: integer;
193 default value: 3
194 lock_file_check_time
196 Specifies how often (in seconds) a high availability server
197 will check to see if it should become active. (for threaded
198 high availability) Must be greater than lock_file_update_time.
199 Format: integer; default value: 9
200 log_events
202 A bit string which specifies the type of events which are
203 logged, see the section on Event Logging in chapter 3 of the
204 ERS. Format: integer; default value: 511, all events.
205 log_file_max_size
207 If this is set to a value > 0 then pbs_server will roll the
208 current log file to logfile.1 when its size is greater than or
209 equal to the value of log_file_max_size. This value is inter‐
210 preted as kilobytes.
211 log_file_roll_depth
213 If this is set to a value >=1 and log_file_max_size is set
214 then pbs_server will continue rolling the log files to log‐
215 file.log_file_roll_depth.
216 log_keep_days
218 If this is set then logs older than X days will be removed by
219 the server. Format: integer; default value: not enforced;
220 log_level
222 Controls the verbosity of server logs. This value ranges from
223 0 to 7 with 7 representing maximum verbosity. Format: inte‐
224 ger; default value: 0, minimum verbosity.
225 mail_body_fmt
227 Override the default format for the body of outgoing mail mes‐
228 sages. A number of printf-like format specifiers and escape
229 sequences can be used:
230 \n new line
232 \t horizontal tab
234 \\ backslash
236 \' single quote
238 \" double quote
240 %d details concerning the message
242 %h PBS host name
244 %i PBS job identifier
246 %j PBS job name
248 %m long reason for message
250 %r short reason for message
252 %% a single %
254 Format: a printf-like format string; Default value: "PBS Job Id:
256 %i\nJob Name: %j\nExec host: %h\n%m\n%d\n".
257 mail_domain
259 Override the default domain for outgoing mail messages. If
260 set, emails will be addressed to "euser@mail_domain". If
261 unset, the job's Job_Owner attribute will be used. If set to
262 "never", TORQUE will never send emails. Format: a domain
263 name; Default value: none.
264 mail_from
266 Set the From: header for all outgoing emails from pbs_server.
267 Format: an email address; Default value: "adm".
268 mail_subject_fmt
270 Override the default format for the subject of outgoing mail
271 messages. A number of printf-like format specifiers and escape
272 sequences can be used:
273 \n new line
275 \t horizontal tab
277 \\ backslash
279 \' single quote
281 \" double quote
283 %d details concerning the message
285 %h PBS host name
287 %i PBS job identifier
289 %j PBS job name
291 %m long reason for message
293 %r short reason for message
295 %% a single %
297 Format: a printf-like format string; Default value: "PBS JOB %i".
299 mail_uid
301 The uid from which server generated mail is sent to users.
302 Format: integer uid; default value: 0 for root.
303 managers
305 List of users granted batch administrator privileges. Format:
306 user@host.sub.domain[,user@host.sub.domain...]. The host,
307 sub-domain, or domain name may be wild carded by the use of an
308 * character, see the description of user access control lists
309 in chapter 10.1.1 of the ERS. Requires full manager privilege
310 to set or alter. Default value: root on the local host.
311 max_job_array_size
313 Specifies the maximum number of jobs that can be in any
314 requested job array. Arrays requesting more jobs than config‐
315 ured will be rejected. Format: integer; default value: Unlim‐
316 ited.
317 max_slot_limit
319 No array can request a slot limit greater than 10. Any array
320 that does not request a slot limit receives a slot limit of
321 10. Using the example above, slot requests greater than 10 are
322 rejected. A slot limit is the maximum number of jobs from an
323 array that can run concurrently. Format: integer; default
324 value: Unlimited.
325 max_running
327 The maximum number of jobs allowed to be selected for execu‐
328 tion at any given time. Advisory to the Scheduler, not
329 enforced by the server. Format: integer.
330 max_user_run
332 The maximum number of jobs owned by a single user that are
333 allowed to be running from this queue at one time. This
334 attribute is advisory to the Scheduler, it is not enforced by
335 the server. Format: integer; default value: none.
336 max_group_run
338 The maximum number of jobs owned by any users in a single
339 group that are allowed to be running from this queue at one
340 time. This attribute is advisory to the Scheduler, it is not
341 enforced by the server. Format: integer; default value: none.
342 mom_job_sync
344 Enables the "job sync on MOM" feature. When MOMs send a sta‐
345 tus update, and it includes a list of jobs, server will issue
346 job deletes for any jobs that don't actually exist. Format:
347 boolean; default value: true.
348 next_job_number
350 This hidden attribute is used to allow a manager to set the
351 value of the next job ID via qmgr. This attribute should
352 rarely be modified. Some sites may find it useful if they need
353 to recreate their pbs_server database (perhaps due to a format
354 change between major TORQUE versions) and they keep a database
355 of job information indexed by the job ID. The manager should
356 be careful to avoid setting the value to something that would
357 allow the next job number to conflict with a job already
358 queued, however Torque will handle this in a sane manner: the
359 job submission will be rejected and the next job number will
360 be incremented.
361 net_counter
363 Lists the 3 numbers representing the number of connections in
364 the last 5 seconds, 30 seconds, and 60 seconds. This is a
365 read-only attribute. Format: string; default value: none.
366 no_mail_force
368 This attribute can be set by a manager via qmgr. When set
369 true, Torque will not force mail to be sent to the user when
370 the jobs mail_points is set to "n". User will not receive
371 notice of job failures or deletions. Format: boolean; default
372 value: false.
373 node_check_rate
375 In OpenPBS, this was the rate at which pbs_server would poll
376 each node. In TORQUE, nodes periodically send updates without
377 solicitation from pbs_server; this attribute is now used as
378 the maximum number of seconds allowed without an update before
379 pbs_server will consider the node down. Format: integer;
380 default value: 150
381 node_ping_rate
383 Specifies the maximum interval (in seconds) between successive
384 pings sent from the pbs_server daemon to the pbs_mom daemon to
385 determine node/daemon health. Format: integer; default value:
386 300
387 node_pack
389 Controls how multiple processor nodes are allocated to jobs.
390 If this attribute is set to true, jobs will be assigned to the
391 multiple processor nodes with the fewest free processors.
392 This packs jobs into the fewest possible nodes leaving multi‐
393 ple processor nodes free for jobs which need many processors
394 on a node. If set to false, jobs will be scattered across
395 nodes reducing conflicts over memory between jobs. If unset,
396 the jobs are packed on nodes in the order that the nodes are
397 declared to the server (in the nodes file). Default value:
398 unset - assigned to nodes as nodes in order that were
399 declared.
400 node_suffix
402 Adds a domainname to node names before IP lookups. Format:
403 string; default value: none.
404 np_default
406 np_default allows the administrator to unify the number of
407 processors (np) on all nodes. The value can be dynamically
408 changed. A value of 0 tells pbs_server to use the value of np
409 found in the nodes file. The maximum value is 32767. Format:
410 integer; default value: not used.
411 operators
413 List of users granted batch operator privileges. Format of
414 the list is identical with managers above. Requires full man‐
415 ager privilege to set or alter. Default value: root on the
416 local host.
417 owner_purge
419 Allows job owners to forcibly purge their own jobs from the
420 server. This short-circuits the normal flow of events and is
421 hightly discouraged. Default value: unset
422 poll_jobs
424 Controls how pbs_server will send job status requests to MOMs.
425 When unset or false, statjob requests from clients (ie:
426 qstat(1B) or the scheduler) may trigger job status requests to
427 MOMs and and must wait until the MOMs have replied; this is
428 suitable for small to medium sized clusters. When set to
429 true, pbs_server will send periodic job status requests; this
430 is suitable for busy clusters with lots of jobs, lots of
431 clients, qstat(1B) is too slow, or your scheduler times out.
432 (see job_stat_rate) Default value: TRUE
433 query_other_jobs
435 The setting of this attribute controls if general users, other
436 than the job owner, are allowed to query the status of or
437 select the job. Format: boolean (see acl_host_enable);
438 Requires full manager privilege to set or alter. default
439 value: false - users may not query or select jobs owned by
440 other users.
441 resources_available
443 The list of resource and amounts available to jobs run by this
444 server. The sum of the resource of each type used by all jobs
445 running by this server cannot exceed the total amount listed
446 here. Advisory to the Scheduler, not enforced by the server.
447 Format: "resources_available.resource_name=value[,...]".
448 resources_cost
450 The cost factors of various types of resources. These values
451 are used in determining the order of releasing members of syn‐
452 chronous job sets, see the section on Synchronize Job Starts.
453 For the most part, these value are purely arbitrary and have
454 meaning only in the relative values between systems. The cost
455 of the resources requested by a job is the sum of the products
456 of the various resources_cost s and the amount of each
457 resource requested by the job. It is not necessary to assign
458 a cost for each possible resource, only those which the site
459 wishes to be considered in synchronous job scheduling. For‐
460 mat: "resources_cost.resource_name=value[,...]"; default
461 value: none, cost of resource is not computed.
462 resources_default
464 The list of default resource values that are set as limits for
465 a job executing on this server when the job does not specify a
466 limit, and there is no queue default. Format:
467 "resources_default.resource_name=value[,...]"; default value:
468 no limit.
469 resources_max
471 The maximum amount of each resource which can be requested by
472 a single job executing on this server if there is not a
473 resources_max valued defined for the queue in which the job
474 resides. Format: "resources_max.resource_name=value[,...]";
475 default value: infinite usage.
476 sched_version
478 A string specifying the scheduler version. Schedulers should
479 check this string when starting and not become active if the
480 wrong string is found. This is ignored by pbs_server.
481 scheduler_iteration
483 The time, in seconds, between iterations of attempts by the
484 batch server to schedule jobs. On each iteration, the server
485 examines the available resources and runnable jobs to see if a
486 job can be initiated. This examination also occurs whenever a
487 running batch job terminates or a new job is placed in the
488 queued state in an execution queue. Format: integer seconds;
489 default value: 10 minutes, set by {PBS_SCHEDULE_CYCLE} in
490 server_limits.h.
491 scheduling
493 Controls if the server will request job scheduling by the PBS
494 job scheduler. If true, the scheduler will be called as
495 required; if false, the scheduler will not be called and no
496 job will be placed into execution unless the server is
497 directed to do so by an operator or administrator. Setting or
498 resetting this attribute to true results in an immediate call
499 to the scheduler. For more information, see the section
500 Scheduler - Server Interaction in the PBS Administrator Guide.
501 Format: boolean (see acl_host_enable); default value: value of
502 -a option when server is invoked, if -a is not specified, the
503 value is recovered from the prior server run. If it has never
504 been set, the value is "false".
505 server_name
507 The name of the server which is the same as the host name. If
508 the hostname resolves to an external IP address, then set this
509 to a name that resolves to the internal IP.
510 submit_hosts
512 A list of hostnames allowed to submit jobs to this batch
513 server regardless of ruserok().
514 system_cost
516 An arbitrary value factored into the resource cost of any job
517 managed by this server for the purpose of selecting which mem‐
518 ber of synchronous set is released first, see resources_cost
519 and section 3.2.2, Synchronize Job Starts. [default value:
520 none, cost of resource is not computed]
521 tcp_timeout
523 Specifies the pbs_server to pbs_mom TCP socket timeout in sec‐
524 onds. Format: integer; default value: 6.
525 The following attributes are read-only, they are maintained by the
527 server and cannot be changed by a client.
528 pbs_version
530 The release version number of the server.
531 resources_assigned
533 The total amount of certain types of resources allocated to
534 running jobs.
535 server_state
537 The current state of the server:
538 Active The server is running and will invoke the job scheduler
540 as required to schedule jobs for execution.
541 Idle The server is running but will not invoke the job
543 scheduler.
544 Scheduling
546 The server is running and there is an outstanding
547 request to the job scheduler.
548 Terminating
550 The server is terminating. No additional jobs will be
551 scheduled.
552 Terminating, Delayed
554 The server is terminating in delayed mode. The server
555 will not run any new jobs and will shutdown when the
556 last currently executing job completes.
557 state_count
559 The total number of jobs managed by the server currently in
560 each state.
561 total_jobs
563 The total number of jobs currently managed by the server.
564
566 the PBS ERS, qmgr(1B), pbs_resources(7B)
567Local pbs_server_attributes(7B)