Session Property Managers

    Add an file with the following contents to enable the built-in manager that reads a JSON config file:

    Change the value of session-property-manager.config-file to point to a JSON config file, which can be an absolute path, or a path relative to the Presto data directory.

    This configuration file consists of a list of match rules, each of which specify a list of conditions that the query must meet, and a list of session properties that should be applied by default. All matching rules contribute to constructing a list of session properties. Rules are applied in the order they are specified. Rules specified later in the file override values for properties that have been previously encountered.

    • user (optional): regex to match against user name.

    • source (optional): regex to match against source string.

    • queryType (optional): string to match against the type of the query submitted:

      • DATA_DEFINITION: Queries that alter/create/drop the metadata of schemas/tables/views, and that manage prepared statements, privileges, sessions, and transactions.

      • DESCRIBE: DESCRIBE, DESCRIBE INPUT, DESCRIBE OUTPUT, and SHOW queries.

      • EXPLAIN: EXPLAIN queries.

      • INSERT: INSERT and CREATE TABLE AS queries.

      • SELECT: SELECT queries.

    • (optional): list of tags. To match, every tag in this list must be in the list of client-provided tags associated with the query.

    • group (optional): regex to match against the fully qualified name of the resource group the query is routed to.

    • clientInfo (optional): regex to match against the client info text supplied by the client

    • sessionProperties: map with string keys and values. Each entry is a system or catalog property name and corresponding value. Values must be specified as strings, no matter the actual data type.

    Example

    Consider the following set of requirements:

    • All interactive queries are routed to subgroups under the global.interactive group, and have an execution time limit of 1 hour (tighter than the constraint on global).

    • All ETL queries (tagged with ‘etl’) are routed to subgroups under the global.pipeline group, and must be configured with certain properties to control writer behavior.

    • All high memory ETL queries (tagged with ‘high_mem_etl’) are routed to subgroups under the global.pipeline group, and must be configured to enable .

    These requirements can be expressed with the following rules:

    1. [
    2.   {
    3.     "group": "global.*",
    4.     "sessionProperties": {
    5.       "query_max_execution_time": "8h",
    6.     }
    7.   },
    8.   {
    9.     "group": "global.interactive.*",
    10.     "sessionProperties": {
    12.     }
    13.   },
    14.     "group": "global.pipeline.*",
    15.     "clientTags": ["etl"],
    16.     "sessionProperties": {
    17.       "scale_writers": "true",
    18.       "writer_min_size": "1GB"
    19.     }
    20.   },
    21.   {
    22.     "group": "global.pipeline.*",
    23.     "clientTags": ["high_mem_etl"],
    24.     "sessionProperties": {
    25.       "exchange_materialization_strategy": "ALL",
    26.       "partitioning_provider_catalog": "hive",
    27.       "hash_partition_count": 4096
    28.     }
    29.   }