Impala workload management table format
Learn about the available columns in the query history and live query system tables.
Table Format
The following columns are available as part of the query history and live query system tables:
| Column Name | Description | Data Type | Sample Value |
|---|---|---|---|
| cluster_id | String specified through the Impala startup flag to uniquely identify an instance. | string | cluster-123 |
| query_id | Impala assigned query identifier. | string | 214d08bef0831e7a:3c65392400000000 |
| session_id | Impala assigned session identifier. | string | ea4f661af43993d8:587839553a41adb8 |
| session_type | Client session type. | string | HIVESERVER2 |
| hiveserver2_protocol_version | Version of the HiveServer (HS2) protocol that was used by the client when connecting. | string | HIVE_CLI_SERVICE_PROTOCOL_V6 |
| db_user | Effective user on the cluster. | string | csso_name |
| db_user_connection | Username from an authenticated client. | string | csso_name |
| db_name | Name of the database being queried. | string | default |
| impala_coordinator | Name of the coordinator for the query. | string | coord-22899:27000 |
| query_status | Status of the query when it completes. | string | OK |
| query_state | Final state of the query. | string | FINISHED |
| impala_query_end_state | Final Impala state of the query. | string | FINISHED |
| query_type | Type of the query. | string | QUERY |
| network_address | Client IP address and port. | string | 127.0.0.1:40120 |
| start_time_utc | Time when the query started. Time zone is in UTC. | timestamp | 2024-07-17 17:13:46.414316000 |
| total_time_ms | Difference between the query end time and start time, in milliseconds (digits after the decimal point represent milliseconds). | decimal(18,3) | 136.121 |
| query_opts_config | List of query options stored as a single string containing comma-separated values of key-value pairs. | string | TIMEZONE=America/Los_Angeles,CLIENT_IDENTIFIER=Impala Shell v4.4.0a1 (04bdb4d) built on Mon Nov 20 10:49:35 PST 2023 |
| resource_pool | Name of the resource pool for the query. | string | default-pool |
| per_host_mem_estimate | Size, in bytes of the per-host memory estimate. | bigint | 5 |
| dedicated_coord_mem_estimate | Size, in bytes of the dedicated coordinator memory estimate. | bigint | 4 |
| per_host_fragment_instances | Comma-separated string listing each host and its fragment instances. | string | myhost-1:27000=1,myhost-2:27001=2 |
| backends_count | Count of the number of backends used by this query. | integer | 2 |
| admission_result | Result of the admission (not applicable to DDLs). | string | Admitted immediately |
| cluster_memory_admitted | Cluster memory, in bytes that was admitted. | integer | 4 |
| executor_group | Name of the executor group. | string | executor_group |
| executor_groups | List of all executor groups including the groups that were considered and rejected as part of Workload Aware Auto Scaling. | string | executor_group1, executor_group2… |
| exec_summary | Full text of the executor summary. | string | |
| num_rows_fetched | Number of rows fetched by the query. | bigint | 6001215 |
| row_materialization_rows_per_sec | Count of the number of rows materialized per second. | bigint | 3780 |
| row_materialization_time_ms | Time spent materializing rows converted to milliseconds. | decimal(18,3) | 1.58 |
| compressed_bytes_spilled | Count of bytes that were written (or spilled) to scratch disk space. | bigint | 241515 |
| event_planning_finished | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 27.253 |
| event_submit_for_admission | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 30.204 |
| event_completed_admission | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 30.986 |
| event_all_backends_started | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 31.969 |
| event_rows_available | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 31.969 |
| event_first_row_fetched | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 135.175 |
| event_last_row_fetched | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 135.181 |
| event_unregister_query | Event from the timeline.The value represents the number of milliseconds since the query was received. | decimal(18,3) | 141.435 |
| read_io_wait_total_ms | Total read I/O wait time converted to milliseconds. | bigint | 15.091 |
| read_io_wait_mean_ms | Average read I/O wait time across executors converted to milliseconds | bigint | 35.515 |
| bytes_read_cache_total | Total bytes read from the data cache | bigint | 45823 |
| bytes_read_total | Total bytes read | bigint | 745227 |
| pernode_peak_mem_min | Minimum value of all the per-node peak memory usages | bigint | 5552846 |
| pernode_peak_mem_max | Maximum value of all the per-node peak memory usages | bigint | 5552846 |
| pernode_peak_mem_mean | Mean value of all the per-node peak memory usages | bigint | 5552846 |
| sql | SQL statement as provided by the user | string | SELECT db_user, total_time_ms from impala_query_log where query_state = 'EXCEPTION'; |
| plan | Full text of the query plan | string | |
| tables_queried | Comma-separated string containing all the tables queried in the SQL statement. Aliased tables are resolved to their actual table names. | string | db.tbl,db.tbl |
| select_columns | Comma-separated string containing all columns from the select list of the sql. Aliased columns are resolved to their actual column names. Each column is in the format database.table.column_name. | string | db.tbl.col1,db.tbl.col2 |
| where_columns | Comma-separated string containing all columns from the where list of the sql. Aliased columns are resolved to their actual column names. Each column is in the format database.table.column_name. | string | db.tbl.col1,db.tbl.col2 |
| join_columns | Comma-separated string containing all columns from the sql used in a join. Aliased columns are resolved to their actual column names. Each column is in the format database.table.column_name. | string | db.tbl.col1,db.tbl.col2 |
| aggregate_columns | Comma-separated string containing all columns from the group by and having lists of the sql. Aliased columns are resolved to their actual column names. Each column is in the format database.table.column_name. | string | db.tbl.col1,db.tbl.col2 |
| orderby_columns | Comma-separated string containing all columns from the order by list of the sql. Aliased columns are resolved to their actual column names. Each column is in the format database.table.column_name. | string | db.tbl.col1,db.tbl.col2 |
| coordinator_slots | Number of query slots used by the query on the coordinator. | bigint | 1 |
| executor_slots | Numberof query slots used by the query on the executors. The value in this column represents the slots used by a single executor, not the total number of slots across all executors. | bigint | 1 |
