CaGrid Best Practices

From CaGrid

This page documents best practices that community members have found useful when using caGrid. Please contribute new best practices and add detail to existing best practices.

Synchronize against the trust fabric

Sync against your Grid GTS service regularly to ensure that any changes in the trust fabric are reflected on your Grid node. This can be done by using the SyncGTS service.

Enable CQL and DCQL Validation

Configure data services to validate CQL and DCQL queries.

Enable logging on critical services

Enable logging on Grid data services to monitor access to the data. A best practice is to write the users Grid identity to the log and save the log on backups.

Use caGrid Transfer service when moving large amounts of data

Use caGrid transfer for moving large sets of data around the Grid.

Use a stateful Grid service for long-running Grid executions

Using a “job invocation” (service contexts) pattern in a stateful Grid service for executing analytical service operations (jobs) that take more than a few seconds to complete.

Write analytical services for scalability

Using thread pools in analytical services for scalability.

Use the Globus best practices

Please refer to Globus best practices.

Best Practices Contributors

Ron Price at the University of Utah