SVCS_S3QUERY_SUMMARY
Use the SVCS_S3QUERY_SUMMARY view to get a summary of all Redshift Spectrum queries (S3 queries) that have been run on the system. One segment can perform one external table scan.
Note
System views with the prefix SVCS provide details about queries on both the main and concurrency scaling clusters. The views are similar to the views with the prefix SVL except that the SVL views provide information only for queries run on the main cluster.
SVCS_S3QUERY_SUMMARY is visible to all users. Superusers can see all rows; regular users can see only their own data. For more information, see Visibility of data in system tables and views.
For information about SVL_S3QUERY, see SVL_S3QUERY.
Table columns
| Column name | Data type | Description |
|---|---|---|
| userid | integer | The ID of the user that generated the given entry. |
| query | integer | The query ID. You can use this value to join various other system tables and views. |
| xid | bigint | The transaction ID. |
| pid | integer | The process ID. |
| segment | integer | The segment number. A query consists of multiple segments, and each segment consists of one or more steps. |
| step | integer | The query step that ran. |
| starttime | timestamp | The time in UTC that the Redshift Spectrum query in this segment started running. One segment can have one external table scan. |
| endtime | timestamp | The time in UTC that the Redshift Spectrum query in this segment completed. One segment can have one external table scan. |
| elapsed | integer | The length of time that it took the Redshift Spectrum query in this segment to run (in microseconds). |
| aborted | integer | If a query was stopped by the system or canceled
by the user, this column contains 1. If the
query ran to completion, this column contains
0. |
| external_table_name | char(136) | The internal format of name of the external name of the table for the external table scan. |
| file_format | character(16) | The file format of the external table data. |
| is_partitioned | char(1) | If true (t), this column
value indicates that the external table is partitioned. |
| is_rrscan | char(1) | If true (t), this column
value indicates that a range-restricted scan was applied. |
| is_nested | varchar(1) | If true (t), this column
value indicates that the nested column data type is accessed. |
| s3_scanned_rows | bigint | The number of rows scanned from Amazon S3 and sent to the Redshift Spectrum layer. |
| s3_scanned_bytes | bigint | The number of bytes scanned from Amazon S3 and sent to the Redshift Spectrum layer, based on compressed data. |
| s3query_returned_rows | bigint | The number of rows returned from the Redshift Spectrum layer to the cluster. |
| s3query_returned_bytes | bigint | The number of bytes returned from the Redshift Spectrum layer to the cluster. A large amount of data returned to Amazon Redshift might affect system performance. |
| files | integer | The number of files that were processed for this Redshift Spectrum query. A small number of files limits the benefits of parallel processing. |
| files_max | integer | The maximum number of file processed on one slice. |
| files_avg | integer | The average number of file processed on one slice. |
| splits | bigint | The number of splits processed for this segment. The number of splits processed on this slice. With large splitable data files, for example, data files larger than about 512 MB, Redshift Spectrum tries to split the files into multiple S3 requests for parallel processing. |
| splits_max | integer | The maximum number of splits processed on this slice. |
| splits_avg | bigint | The average number of splits processed on this slice. |
| total_split_size | bigint | The total size of all splits processed. |
| max_split_size | bigint | The maximum split size processed, in bytes. |
| avg_split_size | bigint | The average split size processed, in bytes. |
| total_retries | bigint | The total number of retries for the Redshift Spectrum query in this segment. |
| max_retries | integer | The maximum number of retries for one individual processed file. |
| max_request_duration | bigint | The maximum duration of an individual file request (in microseconds). Long running queries might indicate a bottleneck. |
| avg_request_duration | bigint | The average duration of the file requests (in microseconds). |
| max_request_parallelism | integer | The maximum number of parallel requests at one slice for this Redshift Spectrum query. |
| avg_request_parallelism | double precision | The average number of parallel requests at one slice for this Redshift Spectrum query. |
| total_slowdown_count | bigint | The total number of Amazon S3 requests with a slow down error that occurred during the external table scan. |
| max_slowdown_count | integer | The maximum number of Amazon S3 requests with a slow down error that occurred during the external table scan on one slice. |
Sample query
The following example gets the scan step details for the last query run.
select query, segment, elapsed, s3_scanned_rows, s3_scanned_bytes, s3query_returned_rows, s3query_returned_bytes, files from svcs_s3query_summary where query = pg_last_query_id() order by query,segment;
query | segment | elapsed | s3_scanned_rows | s3_scanned_bytes | s3query_returned_rows | s3query_returned_bytes | files ------+---------+---------+-----------------+------------------+-----------------------+------------------------+------ 4587 | 2 | 67811 | 0 | 0 | 0 | 0 | 0 4587 | 2 | 591568 | 172462 | 11260097 | 8513 | 170260 | 1 4587 | 2 | 216849 | 0 | 0 | 0 | 0 | 0 4587 | 2 | 216671 | 0 | 0 | 0 | 0 | 0
