Mail Address : info@applicationsrus.com

Convert Workflow queue from SQL Server to Message Queue


Use Case

A major player in the Medical Treatment Reporting business has many Workflow programs that used SQL Server as source to get the next record to process in a workflow. This was done via a Polling process where SQL table was pooled every second to see if more work was needed. While this worked okay it put quite a load on the database, especially since there was not always work to be done. The client decided it wanted to use a more efficient method of getting next workflow record that would not be load on the database and would not be constantly running every second for each Workflow.

Solution

Applications R-Us designed a solution that would implement a “pushing” strategy where instead of constantly “polling” the DB every second each Workflow was assigned to a AWS RabbitMQ message broker. Using this design the Workflow would stay dormant until a new entry was sent from the RabbitMQ message broker. This greatly reduce the load on the SQL Server database as well as tying up other Server resources as it was only working when information was “pushed: from RabbitMQ.

Technologies Utilized

.NET, C#, AWS Parameter Store, AWS RabbitMQ and Windows Scheduled Tasks

Click here to see Project Site