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 acl_group_sloppy
16 This is a default value for the queue attribute of the same
17 name. Format: boolean, "TRUE", "True", "true", "Y", "y", "1",
18 "FALSE", "False", "false", "N", "n", "0"; default value: false
19 = disabled.
20 acl_host_enable
22 Attribute which when true directs the server to use the
23 acl_hosts access control lists. Requires full manager privi‐
24 lege to set or alter. Format: boolean, "TRUE", "True",
25 "true", "Y", "y", "1", "FALSE", "False", "false", "N", "n",
26 "0"; default value: false = disabled.
27 acl_hosts
29 List of hosts which may request services from this server.
30 This list contains the network name of the hosts. Local
31 requests, i.e. from the server's host itself, are aways
32 accepted even if the host is not included in the list. See
33 section 10.1, Authorization, in the PBS External Reference
34 Specification. Requires full manager privilege to set or
35 alter. Format: "[+|-]hostname.domain[,...]"; default value:
36 all hosts.
37 acl_logic_or
39 This is a default value for the queue attribute of the same
40 name. Format: boolean, "TRUE", "True", "true", "Y", "y", "1",
41 "FALSE", "False", "false", "N", "n", "0"; default value: false
42 = disabled.
43 acl_user_enable
45 Attribute which when true directs the server to use the server
46 level acl_users access list. Requires full manager privilege
47 to set or alter. Format: boolean (see acl_group_enable);
48 default value: disabled.
49 acl_users
51 List of users allowed or denied the ability to make any
52 requests of this server. See section 10.1, Authorization, in
53 the PBS External Reference Specification. If acl_user_enable
54 is set to true, only users listed in acl_users may submit to
55 or execute jobs in the queue. Requires full manager privilege
56 to set or alter. Format: "[+|-]user[@host][,...]"; default
57 value: all users allowed.
58 acl_roots
60 List of super users who may submit to and execute jobs at this
61 server. If the job execution id would be zero (0), then the
62 job owner, root@host, must be listed in this access control
63 list or the job is rejected. Format:
64 "[+|-]user[@host][,...]"; default value: no root jobs allowed.
65 allow_node_submit
67 Allow job submissions from compute nodes regardless of
68 ruserok(). Requires full manager privilege to set or alter.
69 Format: boolean; default value: disabled.
70 auto_node_np
72 Automatically configure a node's np value based on the ncpus
73 value from the status update. Requires full manager privilege
74 to set or alter. Format: boolean; default value: disabled.
75 comment
77 A text string which may be set by the scheduler or other priv‐
78 ileged client to provide information to the batch system
79 users. Format: any string; default value: none.
80 default_node
82 A node specification to use if there is no other supplied
83 specification. This attribute is only used by servers where a
84 nodes file exist in the server_priv directory providing a list
85 of nodes to the server. If the nodes file does not exist,
86 this attribute is not set by default and is ignored if set.
87 The default value allows for jobs to share a single node.
88 Format: a node specification string; default value: 1#shared.
89 default_queue
91 The queue which is the target queue when a request does not
92 specify a queue name. Format: a queue name; default value:
93 none, must be set to an existing queue.
94 down_on_error
96 Set a node's state to "down" if MOM reports a message begin‐
97 ning with the string "ERROR". This might interfere with
98 moab's node error handling. See the HEALTH CHECK section in
99 pbs_mom(8B). This is an EXPERIMENTAL feature and may be
100 removed in the future. Format: boolean; default value: false.
101 job_nanny
103 Enables the experimental "job deletion nanny" feature. All
104 job cancels will create a repeating task that will resend KILL
105 signals if the initial job cancel failed. Further job cancels
106 will be rejected with the message "job cancel in progress."
107 This is useful for temporary failures with a job's execution
108 node during a job delete request. It is possible that the job
109 nanny might interfere with job restarts, migrations, and
110 checkpointing. This is an EXPERIMENTAL feature and may be
111 removed in the future. Format: boolean; default value: false.
112 job_stat_rate
114 Moderates how often job stat requests will be issued from
115 pbs_server to the MOM daemons. If poll_jobs is unset or
116 false, then all jobs that haven't been updated in
117 job_stat_rate seconds will trigger a stat request. If
118 poll_jobs is true, then all jobs will be updated every
119 job_stat_rate.
120 (see poll_jobs) On active clusters, 60 or 120 might be rea‐
121 sonable. Default value: 45 seconds (PBS_RESTAT_JOB in
122 server_limits.h) Minimum value: 4 seconds (PBS_JOBSTAT_MIN in
123 server_limits.h)
124 keep_completed
126 Number of seconds to retain completed jobs in the C state.
127 This is overriden by the execution queue attribute of the same
128 name. EXPERIMENTAL Format: integer; default value: 0.
129 kill_delay
131 The amount of the time delay between the sending of SIGTERM
132 and SIGKILL when a qdel command is issued against a running
133 job. This is overriden by the execution queue attribute of
134 the same name. Format: integer seconds; default value: 2 sec‐
135 onds.
136 log_events
138 A bit string which specifies the type of events which are
139 logged, see the section on Event Logging in chapter 3 of the
140 ERS. Format: integer; default value: 511, all events.
141 log_file_max_size
143 If this is set to a value > 0 then pbs_server will roll the
144 current log file to logfile.1 when its size is greater than or
145 equal to the value of log_file_max_size. This value is inter‐
146 preted as kilobytes.
147 log_file_roll_depth
149 If this is set to a value >=1 and log_file_max_size is set
150 then pbs_server will continue rolling the log files to log‐
151 file.log_file_roll_depth.
152 log_level
154 Controls the verbosity of server logs. This value ranges from
155 0 to 7 with 7 representing maximum verbosity. Format: inte‐
156 ger; default value: 0, minimum verbosity.
157 mail_domain
159 Override the default domain for outgoing mail messages. If
160 set, emails will be addressed to "euser@mail_domain". If
161 unset, the job's Job_Owner attribute will be used. Format: a
162 domain name; Default value: none.
163 mail_uid
165 The uid from which server generated mail is sent to users.
166 Format: integer uid; default value: 0 for root.
167 managers
169 List of users granted batch administrator privileges. Format:
170 user@host.sub.domain[,user@host.sub.domain...]. The host,
171 sub-domain, or domain name may be wild carded by the use of an
172 * character, see the description of user access control lists
173 in chapter 10.1.1 of the ERS. Requires full manager privilege
174 to set or alter. Default value: root on the local host.
175 max_running
177 The maximum number of jobs allowed to be selected for execu‐
178 tion at any given time. Advisory to the Scheduler, not
179 enforced by the server. Format: integer.
180 max_user_run
182 The maximum number of jobs owned by a single user that are
183 allowed to be running from this queue at one time. This
184 attribute is advisory to the Scheduler, it is not enforced by
185 the server. Format: integer; default value: none.
186 max_group_run
188 The maximum number of jobs owned by any users in a single
189 group that are allowed to be running from this queue at one
190 time. This attribute is advisory to the Scheduler, it is not
191 enforced by the server. Format: integer; default value: none.
192 mom_job_sync
194 Enables the experimental "job sync on MOM" feature. When
195 MOMs send a status update, and it includes a list of jobs,
196 server will issue job deletes for any jobs that don't actually
197 exist. This is an EXPERIMENTAL feature and may be removed in
198 the future. Format: boolean; default value: false.
199 next_job_number
201 This hidden attribute is used to allow a manager to set the
202 value of the next job ID via qmgr. This attribute should
203 rarely be modified. Some sites may find it useful if they need
204 to recreate their pbs_server database (perhaps due to a format
205 change between major TORQUE versions) and they keep a database
206 of job information indexed by the job ID. The manager should
207 be careful to avoid setting the value to something that would
208 allow the next job number to conflict with a job already
209 queued, however Torque will handle this in a sane manner: the
210 job submission will be rejected and the next job number will
211 be incremented.
212 node_check_rate
214 In OpenPBS, this was the rate at which pbs_server would poll
215 each node. In TORQUE, nodes periodically send updates without
216 solicitation from pbs_server; this attribute is now used as
217 the maximum number of seconds allowed without an update before
218 pbs_server will consider the node down. Format: integer;
219 default value: 150
220 node_pack
222 Controls how multiple processor nodes are allocated to jobs.
223 If this attribute is set to true, jobs will be assigned to the
224 multiple processor nodes with the fewest free processors.
225 This packs jobs into the fewest possible nodes leaving multi‐
226 ple processor nodes free for jobs which need many processors
227 on a node. If set to false, jobs will be scattered across
228 nodes reducing conflicts over memory between jobs. If unset,
229 the jobs are packed on nodes in the order that the nodes are
230 declared to the server (in the nodes file). Default value:
231 unset - assigned to nodes as nodes in order that were
232 declared.
233 operators
235 List of users granted batch operator privileges. Format of
236 the list is identical with managers above. Requires full man‐
237 ager privilege to set or alter. Default value: root on the
238 local host.
239 poll_jobs
241 Controls how pbs_server will send job status requests to MOMs.
242 When unset or false, statjob requests from clients (ie:
243 qstat(1B) or the scheduler) may trigger job status requests to
244 MOMs and and must wait until the MOMs have replied; this is
245 suitable for small to medium sized clusters. When set to
246 true, pbs_server will send periodic job status requests; this
247 is suitable for busy clusters with lots of jobs, lots of
248 clients, qstat(1B) is too slow, or your scheduler times out.
249 (see job_stat_rate) Default value: TRUE
250 query_other_jobs
252 The setting of this attribute controls if general users, other
253 than the job owner, are allowed to query the status of or
254 select the job. Format: boolean (see acl_host_enable);
255 Requires full manager privilege to set or alter. default
256 value: false - users may not query or select jobs owned by
257 other users.
258 resources_available
260 The list of resource and amounts available to jobs run by this
261 server. The sum of the resource of each type used by all jobs
262 running by this server cannot exceed the total amount listed
263 here. Advisory to the Scheduler, not enforced by the server.
264 Format: "resources_available.resource_name=value[,...]".
265 resources_cost
267 The cost factors of various types of resources. These values
268 are used in determining the order of releasing members of syn‐
269 chronous job sets, see the section on Synchronize Job Starts.
270 For the most part, these value are purely arbitrary and have
271 meaning only in the relative values between systems. The cost
272 of the resources requested by a job is the sum of the products
273 of the various resources_cost s and the amount of each
274 resource requested by the job. It is not necessary to assign
275 a cost for each possible resource, only those which the site
276 wishes to be considered in synchronous job scheduling. For‐
277 mat: "resources_cost.resource_name=value[,...]"; default
278 value: none, cost of resource is not computed.
279 resources_default
281 The list of default resource values that are set as limits for
282 a job executing on this server when the job does not specify a
283 limit, and there is no queue default. Format:
284 "resources_default.resource_name=value[,...]"; default value:
285 no limit.
286 resources_max
288 The maximum amount of each resource which can be requested by
289 a single job executing on this server if there is not a
290 resources_max valued defined for the queue in which the job
291 resides. Format: "resources_max.resource_name=value[,...]";
292 default value: infinite usage.
293 scheduler_iteration
295 The time, in seconds, between iterations of attempts by the
296 batch server to schedule jobs. On each iteration, the server
297 examines the available resources and runnable jobs to see if a
298 job can be initiated. This examination also occurs whenever a
299 running batch job terminates or a new job is placed in the
300 queued state in an execution queue. Format: integer seconds;
301 default value: 10 minutes, set by {PBS_SCHEDULE_CYCLE} in
302 server_limits.h.
303 scheduling
305 Controls if the server will request job scheduling by the PBS
306 job scheduler. If true, the scheduler will be called as
307 required; if false, the scheduler will not be called and no
308 job will be placed into execution unless the server is
309 directed to do so by an operator or administrator. Setting or
310 resetting this attribute to true results in an immediate call
311 to the scheduler. For more information, see the section
312 Scheduler - Server Interaction in the PBS Administrator Guide.
313 Format: boolean (see acl_host_enable); default value: value of
314 -a option when server is invoked, if -a is not specified, the
315 value is recovered from the prior server run. If it has never
316 been set, the value is "false".
317 server_name
319 The name of the server which is the same as the host name. If
320 the hostname resolves to an external IP address, then set this
321 to a name that resolves to the internal IP.
322 submit_hosts
324 A list of hostnames allowed to submit jobs to this batch
325 server regardless of ruserok().
326 system_cost
328 An arbitrary value factored into the resource cost of any job
329 managed by this server for the purpose of selecting which mem‐
330 ber of synchronous set is released first, see resources_cost
331 and section 3.2.2, Synchronize Job Starts. [default value:
332 none, cost of resource is not computed]
333 The following attributes are read-only, they are maintained by the
335 server and cannot be changed by a client.
336 resources_assigned
338 The total amount of certain types of resources allocated to
339 running jobs.
340 server_state
342 The current state of the server:
343 Active The server is running and will invoke the job scheduler
345 as required to schedule jobs for execution.
346 Idle The server is running but will not invoke the job
348 scheduler.
349 Scheduling
351 The server is running and there is an outstanding
352 request to the job scheduler.
353 Terminating
355 The server is terminating. No additional jobs will be
356 scheduled.
357 Terminating, Delayed
359 The server is terminating in delayed mode. The server
360 will not run any new jobs and will shutdown when the
361 last currently executing job completes.
362 state_count
364 The total number of jobs managed by the server currently in
365 each state.
366 total_jobs
368 The total number of jobs currently managed by the server.
369 PBS_version
371 The release version number of the server.
372
374 the PBS ERS, qmgr(1B), pbs_resources(7B)
375Local pbs_server_attributes(7B)