[HASH | ORDER] GROUP | Whether distinct values in group by and distinct clauses should be determined calculating a hash or by sorting them |
[CONCAT | HASH | MERGE] UNION | How equal records of union statements are deteremined |
[LOOP | MERGE | HASH] JOIN | Specifies that all join operations of a query are performed with the same strategy. Compare with with join hints which allow to specify different strategies when joining multiple tables. |
EXPAND VIEWS | Expand indexed views. |
FAST n | Optimize for fast retrieval of the first n records. Compare with Oracle's first_rows hint. |
FORCE ORDER | Preserve the join order specified in the from clause of the SQL statement. Compare with Oracle's ordered hint. |
[FORCE | DISABLE] EXTERNALPUSHDOWN | Force/disable the pushdown of the computation of qualifying expressions in Hadoop. Only applies to PolyBase, has no effect on Azure Storage. |
[FORCE | DISABLE] SCALEOUTEXECUTION | |
IGNORE_NONCLUSTERED_COLUMNSTORE_INDEX | |
KEEP PLAN | Relax estimated recompile threshold for a query. (The threshold determines when an automatic query-recompilation takes place because of a change of the number of indexed columns in DML queries) |
KEEPFIXED PLAN | Do not recompile a plan even if statistics have changed. With keepfixed plan , a query is only recompiled if the schema of the underlying tables changes or if sp_recompile is executed against these tables. |
MAX_GRANT_PERCENT|MIN_GRANT_PERCENT = n | Maximum/minimum granted memory as percentage of configured memory limit (n = 0 … 100 ) |
MAXDOP <integer_value> | Override the maximum degree of parallelism (see sp_configure ). Setting this value to 0 lets SQL Server choose the max degree of the parallelism. |
MAXRECURSION n | Maximum number of recursion that is allowed for the query. (Probably useful in conjunction with the with common table expressions). |
NO_PERFORMANCE_SPOOL | |
OPTIMIZE FOR (@var { UNKNOWN | = <literal_constant> } [ , …n ] ) | Optimize the query for a specified value for the variable @var . UNKNOWN lets SQL server use statistical data instead of the initial data. |
OPTIMIZE FOR UNKNOWN | Use average selectivity for predicates. |
PARAMETERIZATION { SIMPLE | FORCED } | |
QUERYTRACEON n | |
RECOMPILE | Create a new and temporary plan just for the execution of the given SQL statement. |
ROBUST PLAN | |
USE HINT ( hint-name [ , …n ] ) | Provide additional hints which are specified by hint-name . |
USE PLAN N'<xml_plan>' | |
TABLE HINT ( <exposed_object_name> [ , <table_hint> [ [, ]...n ] ] ) | |