Rerunning a process flow

Adeptia Connect has a rerun feature to execute aborted process flows. After identifying the reason for the failure, you can rectify and rerun the process flow for successful execution. A process flow may be aborted due to any reason (for example, incorrect data mapping or schema definition).

However, if it is aborted due to Kernel shutdown, the Recovery feature handles it automatically. All incomplete process flows that are recoverable are recovered through Recovery, not Rerun feature. 

This page contains the following information:

How Rerun works

During execution of a process flow, at every checkpoint, the Process Flow stores its current state in a rerun file. There is one rerun file for each execution of a process flow. These rerun files are stored in a rerun folder defined in the abpm.reprocessing.repository.root property. 

If the process flow aborts, the rerun file will have the state of the process flow till the last checkpoint. If no checkpoint is reached, no rerun file is created and the process flow can never be reprocessed after failure. In the rerun file, only the state of the process flow is stored. Intermediate data of the process flows are stored in the repository folder. If your process flow has checkpoints, and the rerun is enabled, then when you rerun, the Kernel looks for the rerun file for that particular process flow instance. If it finds such file, it picks up the intermediate data from there and resumes the process flow from last checkpoint state.

The rerun file remains in the rerun folder unless the process flow execution is completed successfully. The Intermediate data gets deleted automatically by Data Cleanup or a user can delete it manually to free disk space. Click here for more details.

Enabling Rerun

Perform the following steps to enable rerun:

1.  In your process flow, select each of your checkpoints one by one.
   
2. Change their respective Mode property to Rerun. This enables the rerun feature.

Using Checkpoints

Using Checkpoints, you can reprocess your aborted process flows. A checkpoint is an action that can be used while creating a process flow. However, you:

• Cannot reprocess the process flows which do not have checkpoints.
• Can add any number of checkpoints in a process flow.
• Must not use it after an asynchronous activity, which is generating a source stream. Otherwise, after rerun, the source stream (generated by an asynchronous activity) will not be available for the target activity (the activity consuming this stream).