Hi team,
Does anyone know when a query fails or is retried or simply continues to run if a request for resources fails.
As far as I know, RESERVE acquisition can wait in queue for resources. During executing, additional acquisitions can be needed, if there is not enough memory available, query will fail (maybe will be retried?) .
We have a customer that noticed that, in dc_resource_acquisitions AcquireAdditional failed on a node, but query continued to execute, and there were other AcquireAdditional requests later.
Hence the question, why didn't the query fail? what are the conditions for the query to continue to execute, to fail or to be retried, if a resource acquisition request fails?
I attached the example from the customer, please note that the query is an INSERT with an constraint validation that's why there are 2 acquire requests.
Thanks,
Chaima
