Skip to content

Commit 3de48c7

Browse files
committed
DOC: More useful docs
1 parent 9ba3f88 commit 3de48c7

File tree

2 files changed

+16
-4
lines changed

2 files changed

+16
-4
lines changed

doc/users/plugins.rst

Lines changed: 15 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -174,7 +174,9 @@ Workflow execution with HTCondor DAGMan is done by calling::
174174
workflow.run(plugin='CondorDAGMan')
175175

176176
Job execution behavior can be tweaked with the following optional plug-in
177-
arguments::
177+
arguments. The value of most arguments can be a literal string or a filename,
178+
where in the latter case the content of the file will be used as the argument
179+
value::
178180

179181
submit_template : submit spec template for individual jobs in a DAG (see
180182
CondorDAGManPlugin.default_submit_template for the default.
@@ -190,14 +192,24 @@ arguments::
190192
wrapper_args : optional additional arguments to a wrapper command
191193
dagman_args : arguments to be prepended to the job execution script in the
192194
dagman call
195+
block : if True the plugin call will block until Condor has finished
196+
prcoessing the entire workflow (default: False)
193197

194198
Please see the `HTCondor documentation`_ for details on possible configuration
195199
options and command line arguments.
196200

197201
Using the ``wrapper_cmd`` argument it is possible to combine Nipype workflow
198202
execution with checkpoint/migration functionality offered by, for example,
199-
DMTCP_. On a Debian system, executing a workflow with support for
200-
checkpoint/migration for all nodes could look like this::
203+
DMTCP_. This is especially useful in the case of workflows with long running
204+
nodes, such as Freesurfer's recon-all pipeline, where Condor's job
205+
prioritization algorithm could lead to jobs being evicted from compute
206+
nodes in order to maximize overall troughput. With checkpoint/migration enabled
207+
such a job would be checkpointed prior eviction and resume work from the
208+
checkpointed state after being rescheduled -- instead of restarting from
209+
scratch.
210+
211+
On a Debian system, executing a workflow with support for checkpoint/migration
212+
for all nodes could look like this::
201213

202214
# define common parameters
203215
dmtcp_hdr = """

nipype/pipeline/plugins/dagman.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -36,7 +36,7 @@ class CondorDAGManPlugin(GraphPluginBase):
3636
- dagman_args : arguments to be prepended to the job execution script in the
3737
dagman call
3838
- block : if True the plugin call will block until Condor has finished
39-
prcoessing the entire workflow
39+
prcoessing the entire workflow (default: False)
4040
"""
4141

4242
default_submit_template = """

0 commit comments

Comments
 (0)