Starburst Galaxy

  •  Get started

  •  Working with data

  •  Data engineering

  •  Developer tools

  •  Cluster administration

  •  Troubleshooting

  • Galaxy status

  •  Reference
  • Galaxy Telemetry #

    The Galaxy Telemetry catalog is a built-in catalog that lets you access Starburst Galaxy managed datasets such as query history.

    The catalog is available by default. However, you must add it to an AWS cluster, in the us-east-1 region, in order to access it. The catalog is read-only and unmodifiable.

    Query history #

    Query history data is located in the query_history table within the public schema. New history data is accessible within two hours of query completion.

    Users can query up to the last 30 days of their query history by selecting from galaxy_telemetry.public.query_history.

    The query_history table contains the following columns:

    Column Type Description
    cluster_name varchar Name of the cluster running the query.
    email varchar Email address of the user running the query.
    role_name varchar Role of the user running the query.
    create_time timestamp Timestamp when the query was received.
    execution_start_time timestamp Timestamp when query execution started.
    end_time timestamp Timestamp when results were finished being consumed by the client.
    session_catalog varchar The session catalog.
    session_schema varchar The session schema.
    session_properties map(varchar, varchar) Temporary properties for the query session.
    remote_client_address varchar Address of the client that submitted the query.
    user_agent varchar User agent that submitted the query.
    query_type varchar Query type, such as SELECT, UPDATE, or INSERT.
    query_id varchar Unique identifier of the query.
    query varchar Full SQL statement of the submitted query.
    query_plan varchar Full explain plan of the submitted query, including costs for each stage. Identical to the output from the EXPLAIN ANALYZE SQL statement.
    query_state varchar The state of the query, such as FINISHED, FAILED, or RUNNING.
    update_type varchar The type of update the query performed, such as CREATE TABLE. This column is null for SELECT queries.
    tables array(ROW(catalog, varchar, schema varchar, table varchar)) Catalog, schema, and table names accessed by the query.
    internal_network_bytes bigint Number of bytes of data exchanged between nodes during query execution.
    internal_network_rows bigint Number of rows of data exchanged between nodes during query execution.
    output_bytes bigint Number of filtered bytes on query output.
    output_rows bigint Number of filtered rows on query output.
    peak_task_total_memory _bytes bigint Maximum amount of user and system memory reserved at one time by one task.
    peak_task_user_memory _bytes bigint Maximum amount of user memory reserved at one time by one task.
    peak_user_memory _bytes bigint Maximum amount of memory directly tied to query resources used by a query at one time.
    physical_input_bytes bigint Number of uncompressed bytes read from the source.
    physical_input_rows bigint Number of rows read from the source.
    read_bytes bigint Sum of all bytes used in the query input.
    read_rows bigint Sum of all rows used in the query input.
    written_bytes bigint Number of bytes of inserted rows.
    written_rows bigint Number of inserted rows.
    original_query_id varchar Unique identifier of the original query, in case the results were served from cache.
    client_info varchar Information about the client submitting the query.
    index_and_cache_usage_overall double Combined score for performance gains from filtering and projection.
    index_and_cache_usage_filtering double Indicates index usage for filtering operations.
    index_and_cache_usage_projection double Indicates data cache usage for projection operations.
    source varchar Client tool or driver used to execute query.
    date varchar Date (yyyy-MM-dd) of end_time.
    hour varchar Hour (00-23) of end_time.

    Set permissions #

    The galaxy_telemetry catalog is owned by the accountadmin role. Read-only access can be granted to other roles.

    SQL support #

    The catalog provides read access to data and metadata in the datasets. It supports the following features: