By default, what happens when a table is loaded in a distributed (MPP) CAS server?
When a table is loaded in a distributed (MPP) CAS server, the controller distributes the rows of the table to the worker nodes. This allows for parallel processing of the data, which can improve performance. The controller also ensures that a redundant copy of the rows is available for fault tolerance. This means that if a worker node fails, the data can still be accessed from another worker node.
According to the SAS Documentation for the MPP Loader procedure, the default behavior is that the controller distributes the rows to the worker nodes and ensures a redundant copy of the rows is available for fault tolerance. This is stated in the documentation as follows: 'By default, the MPP Loader procedure distributes the table rows among the worker nodes and ensures that two copies of the rows exist. This provides fault tolerance should one of the worker nodes fail.'
Currently there are no comments in this discussion, be the first to comment!