Lyftron stored procedures

Apart from emulating SQL Server's stored procedures, Lyftron introduces some of its own. All of them start with lyft_ prefix to be easily distinguished from regular SQL Server stored procedures.

Administrative

Administrative stored procedures should be used by power users and administrators. Note that most of the tasks can also be achieved conveniently using Lyftron Admin Portal user interface.

Stored procedure name Description
lyft_flush_plan_cache Flushes the whole execution plan cache.
lyft_run_garbage_collection Attempts garbage collection to reclaim unused memory.
lyft_set_default_federator Configures federator for an agent. Given connection will be the new default federator for the designated agent.

Cache management

Stored procedure name Description
lyft_create_view_copy Creates a primary/secondary view copy for the materialized view.
lyft_evict_spark_inmem_cached_table Removes data from Spark in-memory cache for given table.
lyft_full_load_cached_view_copy Loads cache for a given view.
lyft_full_load_view_copy Fills a primary/secondary view copy.
lyft_incremental_insert Inserts new rows into an incremantally cached view.
lyft_incremental_load_cached_view_copy Conducts the whole process of an incremental load of new rows.
lyft_invalidate_spark_cached_table Invalidates and then loads Spark in-memory cache for given table.
lyft_populate_cached_table Materializes a cached view.
lyft_refresh_view
lyft_set_reload_mode Sets reload mode for given view.
lyft_set_view_cache_mode Sets the cache mode for given view.
lyft_set_view_copy_status
lyft_set_incremental_marker Sets new incremental marker value for given table or view.
lyft_set_incremental_marker_column Sets incremental marker column for given table or view. The column will be later used for incremental load process.
lyft_switch_view_copy Switches view copy between the primary and secondary table for cached views with two rotated tables.
lyft_trim_view_copy Drops persistent cache and recreates it.

Connection management

Stored procedure name Description
sp_addlinkedserver Creates a connection to a data source. A connection allows for access to a single data source (endpoint).
sp_addlinkedsrvlogin Creates or updates a mapping between a connection on the local instance of Lyftron and a security account on a remote data source.
sp_dropserver Removes a connection from the list of known connections on the instance of Lyftron.
sp_testlinkedserver Tests the connection to a data source.
sp_updatelinkedserver Updates a connection to a data source.

Diagnostics

Stored procedure name Description
lyft_set_logging_level Dynamically changes Lyftron logging level.
lyft_help_jobs Fetches a data set with details of all jobs that the schedule is aware of.

Job management

Stored procedure name Description
sp_add_job Adds a job executed by the job scheduler.
sp_add_jobstep Adds a step (operation) to a job.
sp_add_schedule Creates a schedule that can be used by any number of jobs.
sp_attach_schedule Sets a schedule for a job.
sp_detach_schedule Removes an association between a schedule and a job.
sp_delete_job Deletes a job.
sp_start_job Instructs the scheduler to execute a job immediately.
sp_update_job Changes the attributes of a job.
sp_update_jobstep Changes the setting for a step in a job that is used to perform automated activities.
sp_update_schedule Changes the settings for a schedule.
sp_delete_jobstep Deleted a step from the job.
sp_delete_schedule Deletes a schedule.

Utility

Stored procedure name Description
lyft_execute_provider_sql_on_connection Executes provider-specific SQL query or command on a given connection.
lyft_update_object_statistics Updates statistics (rowcount and distinct count) for a given table.
lyft_update_row_count_object_statistics Updates rowcount statistics for a given table.

Metadata management

Stored procedure name Description
lyft_get_partitioning_schema Retrieves partitioning schema for a given view.
lyft_set_partitioning_schema Sets partitioning mode and schema for a given view.
sp_generate_views Generates views on target database based on objects (tables or views) from a source database.
sp_import_metadata Generates views on target database based on objects (tables or views) from a source database.

Spark management

Stored procedure name Description
lyft_spark_force_become_active Transitions the node to Active, regardless of Zookeeper state.
lyft_spark_force_become_standby Transitions the node to StandBy, regardless of Zookeeper state.
lyft_spark_instance_kill Forces an OS-level kill on the given Spark process.
lyft_spark_instance_start Tries to find a running Spark process in the OS. If not found, a new process is started.
lyft_spark_instance_stop Sends a stop-server signal to the remote driver.
lyft_spark_release Releases Spark context and transitions driver to non-initialized state.
lyft_spark_stop_driver Sends a stop-server signal to the remote driver.