Hello The community article Configure a Compute Context with Pre-started Compute Servers from the command-line has demonstrated that we can have multiple pre-started compute -servers. Wondering if there is a way to load balance these compute servers. The idea is that when a job is started using rest-api and using the new compute context, it should go to an idle compute server.
When you configure reusable compute servers with an available minimum, then when a new compute server request comes in for that context it would use one of those reusable servers already running unless they are all already in use.
-- Greg Wootton | Principal Systems Technical Support Engineer
Thanks @gwootton Say for example I have a new compute context to create two pre-launched servers. Now if both the pre-launched compute-server are in use and new request comes for the same context, will it create two more servers OR fail?
It would create one new server to handle the new request.
-- Greg Wootton | Principal Systems Technical Support Engineer
The SAS Users Group for Administrators (SUGA) is open to all SAS administrators and architects who install, update, manage or maintain a SAS deployment.