Connectors and Job Servers

Many Connectors require a Job Server with a particular Service associated with it. For example, to use the Azure Machine Learning Connector, you need a Job Server with the ServiceForRedwood_AzureML Service.

When you create a Connection for a particular Connector, RunMyJobs will prompt you to create a Queue and Job Server for that Connector, and will automatically assign the appropriate Service to the Job Server. For example, if you install the Azure Synapse Connector, RunMyJobs will create an Azure Job Server with the required ServiceForRedwood_AzureSynapse Service.

However, some Connectors share a single Connection Type. For example, Azure Synapse and Azure Machine Learning both use the Azure Connection type. If you install the Azure Synapse Connector first, and then install the Azure Machine Learning Connector, RunMyJobs will not prompt you to create a Job Server for Azure Machine Learning, because a Job Server for the Azure Connection Type already exists. If you then try to submit an Azure Machine Learning process, you may see an error similar to the following.

In this situation, there are two solutions:

• Stop the existing Job Server, then edit its definition and add the required Service.

• Create a new Job Server for the new Connector and add the required Service to it.

The option you choose depends on your needs. For example, if you have two Connectors with the same Connection Type, but you don't use them often, having both of them on a single Job Server might be fine. But if the Connectors have a high volume of traffic, you may want to have a separate Job Server for each Connector for efficiency reasons.