When I was exploring BigQuery's public dataset today and the details shown from BigQuery information schema I noticed some of the BigQuery tables use a storage tier called "COMPRESSED". This is quite interesting because the logical bytes of the table and the billable bytes of the table are significantly different which mean the cost on querying those tables will be significantly cheaper. Anyone seen this before or know how to make use of it?
But this doesn't seem to be documented anywhere and all tables I've created seem to either have Active or Long Term but I've not seen anything labelled as COMPRESSED.
The following query got me these information
SELECT
*
FROM
`bigquery-public-data.london_bicycles.INFORMATION_SCHEMA.PARTITIONS`
And some of the details I retrieved as as follows and the logical type is about 6 times bigger in size for table cycle_hire
Row table_catalog table_schema table_name partition_id total_rows total_logical_bytes total_billable_bytes last_modified_time storage_tier
Row table_catalog table_schema table_name partition_id total_rows total_logical_bytes total_billable_bytes last_modified_time storage_tier
Row | table_catalog | table_schema | table_name | partition_id | total_rows | total_logical_bytes | total_billable_bytes | last_modified_time | storage_tier | |
1 | bigquery-public-data | london_bicycles | cycle_stations | null | 785 | 78650 | 30982 | 2022-09-09 21:02:32.345000 UTC | COMPRESSED | |
2 | bigquery-public-data | london_bicycles | cycle_hire | null | 24369201 | 2784394803 | 469461272 | 2019-05-08 11:21:52.696000 UTC | COMPRESSED |
LinkedIn
Twitter
