In the previous section we have implemented our first task (the t1.ecf file).
The t1.ecf script needs to be preprocessed to generate the job file.
This pre-processing is done automatically by ecflow_server the when the task is about to run.
However it is possible to check the job creation before the suite definition is loaded into the ecflow_server.
Text
Automated job creation checking is only available with Python.
If the ecflow_server can’t locate the ecf script, please see ecf file location algorithm
Python
The process of job creation can be checked before the suite definition
is loaded into the ecflow_server. The following checks are done:
- Locating ecf script files, corresponding to the task in the suite definition.
- Performing pre-processing
When the suite definition is large and has many ecf script this
checking can save a lot of time.
The following point’s should be noted about about job creation checking:
- It is independent of the ecflow_server.Hence ECF_PORT and ECF_NODE in the job file will have default values.
- Job files have a .job0 extension, whereas the server will always generate jobs with a extension .job<1-n>, i.e t1.job1, t1.job2.The numbers correspond to ECF_TRYNO which is never zero.
Checking is done using ecflow.Defs.check_job_creation
#!/usr/bin/env python2.7
import os
import ecflow
print "Creating suite definition"
defs = ecflow.Defs()
suite = defs.add_suite("test")
suite.add_variable("ECF_HOME", os.path.join(os.getenv("HOME") + "course"))
suite.add_task("t1")
print defs
print "Checking job creation: .ecf -> .job0"
print defs.check_job_creation()
# We can assert, so that we only progress, once all job creation works
# assert len(defs.check_job_creation()) == 0, "Job generation failed"
What to do
- Add job creation checking.
- Examine the job file
$HOME/course/test/t1.job0