JobScheduling¶
The Job Scheduling Executor takes the information gained from all previous optimizers and makes a scheduling decision for the jobs.
Subsequent to this jobs are added into a Task Queue and pilot agents can be submitted.
All issues preventing the successful resolution of a site candidate are discovered here where all information is available.
This Executor will fail affected jobs meaningfully.
- class DIRAC.WorkloadManagementSystem.Executor.JobScheduling.JobScheduling¶
Bases:
OptimizerExecutor
The specific Optimizer must provide the following methods: - optimizeJob() - the main method called for each job and it can provide: - initializeOptimizer() before each execution cycle
- deserializeTask(taskStub)¶
- disableFastTrackForTask()¶
- classmethod ex_enabled()¶
- classmethod ex_getExtraArguments()¶
- classmethod ex_getMind()¶
- classmethod ex_getOption(optName, defaultValue=None)¶
- classmethod ex_getProperty(optName)¶
- classmethod ex_optimizerName()¶
- classmethod ex_setMind(mindName, **extraArgs)¶
- classmethod ex_setOption(optName, value)¶
- classmethod ex_setProperty(optName, value)¶
- fastTrackDispatch(jid, jobState)¶
- freezeTask(freezeTime)¶
- classmethod initialize()¶
- classmethod initializeOptimizer()¶
Initialization of the optimizer.
- isTaskFrozen()¶
- property jobLog¶
- optimizeJob(jid, jobState)¶
Banned sites are removed from the destination list.
Get input files
Production jobs are sent directly to TQ
Check if staging is necessary
- processTask(jid, jobState)¶
- retrieveOptimizerParam(name)¶
- serializeTask(cjs)¶
- setNextOptimizer(jobState=None)¶
- storeOptimizerParam(name, value)¶