When you have a Pega application running longer time in production you might already have a big case volume. In this time if you expose a new column to the case table and deploy to Pega the column population job triggered by the deployment could take longer time to run. This also could cause some […]