Vertica Analytics Platform Version 9.2.x Documentation
Specifies partitioning of table data, through a
PARTITION BY clause in the table definition:
PARTITION BY partition‑expression [ GROUP BY group‑expression ] [ ActivePartitionCountExpr ]
For each table row, resolves to a partition key that is derived from one or more table columns.
Avoid partitioning tables on LONG VARBINARY and LONG VARCHAR columns. Doing so can adversely impact performance.
For each table row, resolves to a partition group key that is derived from the partition key. Vertica uses group keys to merge partitions into separate partition groups.
...PARTITION BY (i+j) GROUP BY (
Specifies how many partitions are active for this table, specified as follows:
Partitioning Requirements and Restrictions
PARTITION BY expressions can specify leaf expressions, functions, and operators. The following requirements and restrictions apply:
- All table projections must include all columns referenced in the expression; otherwise, Vertica cannot resolve the expression.
- The expression can reference multiple columns, but it must resolve to a single non-null value for each row.
- All leaf expressions must be constants or table columns.
- All other expressions must be functions and operators. The following restrictions apply to functions:
- The expression cannot include queries.
GROUP BY expressions do not support modulo (%) operations.
Partitioning Tables in the Administrator's Guide
Was this topic helpful?