1RSL(5) Grid Community Toolkit Manual RSL(5)
2
6 rsl - GRAM5 RSL Attributes
7
9 arguments
10 The command line arguments for the executable. Use quotes, if a
11 space is required in a single argument.
12 count
14 The number of executions of the executable. [Default:
15 <literal>1</literal>]
16 directory
18 Specifies the path of the directory the jobmanager will use as the
19 default directory for the requested job. [Default:
20 <literal>$(HOME)</literal>]
21 dry_run
23 If dryrun = yes then the jobmanager will not submit the job for
24 execution and will return success. [Default: <literal>no</literal>]
25 environment
27 The environment variables that will be defined for the executable
28 in addition to default set that is given to the job by the
29 jobmanager.
30 executable
32 The name of the executable file to run on the remote machine. If
33 the value is a GASS URL, the file is transferred to the remote gass
34 cache before executing the job and removed after the job has
35 terminated.
36 expiration
38 Time (in seconds) after a a job fails to receive a two-phase commit
39 end signal before it is cleaned up. [Default:
40 <literal>14400</literal>]
41 file_clean_up
43 Specifies a list of files which will be removed after the job is
44 completed.
45 file_stage_in
47 Specifies a list of ("remote URL" "local file") pairs which
48 indicate files to be staged to the nodes which will run the job.
49 file_stage_in_shared
51 Specifies a list of ("remote URL" "local file") pairs which
52 indicate files to be staged into the cache. A symlink from the
53 cache to the "local file" path will be made.
54 file_stage_out
56 Specifies a list of ("local file" "remote URL") pairs which
57 indicate files to be staged from the job to a GASS-compatible file
58 server.
59 gass_cache
61 Specifies location to override the GASS cache location.
62 gram_my_job
64 Obsolete and ignored. [Default: <literal>collective</literal>]
65 host_count
67 Only applies to clusters of SMP computers, such as newer IBM SP
68 systems. Defines the number of nodes ("pizza boxes") to distribute
69 the "count" processes across.
70 job_type
72 This specifies how the jobmanager should start the job. Possible
73 values are single (even if the count > 1, only start 1 process or
74 thread), multiple (start count processes or threads), mpi (use the
75 appropriate method (e.g. mpirun) to start a program compiled with a
76 vendor-provided MPI library. Program is started with count nodes),
77 and condor (starts condor jobs in the "condor" universe.) [Default:
78 <literal>multiple</literal>]
79 library_path
81 Specifies a list of paths to be appended to the system-specific
82 library path environment variables. [Default:
83 <literal>$(GLOBUS_LOCATION)/lib</literal>]
84 loglevel
86 Override the default log level for this job. The value of this
87 attribute consists of a combination of the strings FATAL, ERROR,
88 WARN, INFO, DEBUG, TRACE joined by the | character
89 logpattern
91 Override the default log path pattern for this job. The value of
92 this attribute is a string (potentially containing RSL
93 substitutions) that is evaluated to the path to write the log to.
94 If the resulting string contains the string $(DATE) (or any other
95 RSL substitution), it will be reevaluated at log time.
96 max_cpu_time
98 Explicitly set the maximum cputime for a single execution of the
99 executable. The units is in minutes. The value will go through an
100 atoi() conversion in order to get an integer. If the GRAM scheduler
101 cannot set cputime, then an error will be returned.
102 max_memory
104 Explicitly set the maximum amount of memory for a single execution
105 of the executable. The units is in Megabytes. The value will go
106 through an atoi() conversion in order to get an integer. If the
107 GRAM scheduler cannot set maxMemory, then an error will be
108 returned.
109 max_time
111 The maximum walltime or cputime for a single execution of the
112 executable. Walltime or cputime is selected by the GRAM scheduler
113 being interfaced. The units is in minutes. The value will go
114 through an atoi() conversion in order to get an integer.
115 max_wall_time
117 Explicitly set the maximum walltime for a single execution of the
118 executable. The units is in minutes. The value will go through an
119 atoi() conversion in order to get an integer. If the GRAM scheduler
120 cannot set walltime, then an error will be returned.
121 min_memory
123 Explicitly set the minimum amount of memory for a single execution
124 of the executable. The units is in Megabytes. The value will go
125 through an atoi() conversion in order to get an integer. If the
126 GRAM scheduler cannot set minMemory, then an error will be
127 returned.
128 project
130 Target the job to be allocated to a project account as defined by
131 the scheduler at the defined (remote) resource.
132 proxy_timeout
134 Obsolete and ignored. Now a job-manager-wide setting.
135 queue
137 Target the job to a queue (class) name as defined by the scheduler
138 at the defined (remote) resource.
139 remote_io_url
141 Writes the given value (a URL base string) to a file, and adds the
142 path to that file to the environment through the
143 GLOBUS_REMOTE_IO_URL environment variable. If this is specified as
144 part of a job restart RSL, the job manager will update the file’s
145 contents. This is intended for jobs that want to access files via
146 GASS, but the URL of the GASS server has changed due to a GASS
147 server restart.
148 restart
150 Start a new job manager, but instead of submitting a new job, start
151 managing an existing job. The job manager will search for the job
152 state file created by the original job manager. If it finds the
153 file and successfully reads it, it will become the new manager of
154 the job, sending callbacks on status and streaming stdout/err if
155 appropriate. It will fail if it detects that the old jobmanager is
156 still alive (via a timestamp in the state file). If stdout or
157 stderr was being streamed over the network, new stdout and stderr
158 attributes can be specified in the restart RSL and the jobmanager
159 will stream to the new locations (useful when output is going to a
160 GASS server started by the client that’s listening on a dynamic
161 port, and the client was restarted). The new job manager will
162 return a new contact string that should be used to communicate with
163 it. If a jobmanager is restarted multiple times, any of the
164 previous contact strings can be given for the restart attribute.
165 rsl_substitution
167 Specifies a list of values which can be substituted into other rsl
168 attributes' values through the $(SUBSTITUTION) mechanism.
169 save_state
171 Causes the jobmanager to save it’s job state information to a
172 persistent file on disk. If the job manager exits or is suspended,
173 the client can later start up a new job manager which can continue
174 monitoring the job.
175 savejobdescription
177 Save a copy of the job description to $HOME [Default:
178 <literal>no</literal>]
179 scratch_dir
181 Specifies the location to create a scratch subdirectory in. A
182 SCRATCH_DIRECTORY RSL substitution will be filled with the name of
183 the directory which is created.
184 stderr
186 The name of the remote file to store the standard error from the
187 job. If the value is a GASS URL, the standard error from the job is
188 transferred dynamically during the execution of the job. There are
189 two accepted forms of this value. It can consist of a single
190 destination: stderr = URL, or a sequence of destinations: stderr =
191 (DESTINATION) (DESTINATION). In the latter case, the DESTINATION
192 may itself be a URL or a sequence of an x-gass-cache URL followed
193 by a cache tag. [Default: <literal>/dev/null</literal>]
194 stderr_position
196 Specifies where in the file remote standard error streaming should
197 be restarted from. Must be 0.
198 stdin
200 The name of the file to be used as standard input for the
201 executable on the remote machine. If the value is a GASS URL, the
202 file is transferred to the remote gass cache before executing the
203 job and removed after the job has terminated. [Default:
204 <literal>/dev/null</literal>]
205 stdout
207 The name of the remote file to store the standard output from the
208 job. If the value is a GASS URL, the standard output from the job
209 is transferred dynamically during the execution of the job. There
210 are two accepted forms of this value. It can consist of a single
211 destination: stdout = URL, or a sequence of destinations: stdout =
212 (DESTINATION) (DESTINATION). In the latter case, the DESTINATION
213 may itself be a URL or a sequence of an x-gass-cache URL followed
214 by a cache tag. [Default: <literal>/dev/null</literal>]
215 stdout_position
217 Specifies where in the file remote output streaming should be
218 restarted from. Must be 0.
219 two_phase
221 Use a two-phase commit for job submission and completion. The job
222 manager will respond to the initial job request with a
223 WAITING_FOR_COMMIT error. It will then wait for a signal from the
224 client before doing the actual job submission. The integer supplied
225 is the number of seconds the job manager should wait before timing
226 out. If the job manager times out before receiving the commit
227 signal, or if a client issues a cancel signal, the job manager will
228 clean up the job’s files and exit, sending a callback with the job
229 status as GLOBUS_GRAM_PROTOCOL_JOB_STATE_FAILED. After the job
230 manager sends a DONE or FAILED callback, it will wait for a commit
231 signal from the client. If it receives one, it cleans up and exits
232 as usual. If it times out and save_state was enabled, it will leave
233 all of the job’s files in place and exit (assuming the client is
234 down and will attempt a job restart later). The timeoutvalue can be
235 extended via a signal. When one of the following errors occurs, the
236 job manager does not delete the job state file when it exits:
237 GLOBUS_GRAM_PROTOCOL_ERROR_COMMIT_TIMED_OUT,
238 GLOBUS_GRAM_PROTOCOL_ERROR_TTL_EXPIRED,
239 GLOBUS_GRAM_PROTOCOL_ERROR_JM_STOPPED,
240 GLOBUS_GRAM_PROTOCOL_ERROR_USER_PROXY_EXPIRED. In these cases, it
241 can not be restarted, so the job manager will not wait for the
242 commit signal after sending the FAILED callback
243 username
245 Verify that the job is running as this user.
246
248 Copyright © 1999-2016 University of Chicago
249Grid Community Toolkit 6 03/31/2018 RSL(5)