Replies: 3 comments
-
Had the exact same thought! |
Beta Was this translation helpful? Give feedback.
-
@djouallah - did my vote! Btw , can you please elaborate on read-only iceberg or delta metadata? |
Beta Was this translation helpful? Give feedback.
-
ducklake use the same approach adopted by BigQuery/Snowflake/Fabric DWH, you write iceberg using a sql based catalog and for other engines to read your data, you just export the metadata (or it is automatically exported), in this model, there is only 1 writer, and other engines are just readers. see example : https://cloud.google.com/bigquery/docs/exporting-data#export_table_metadata |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
it will be useful to have a function that can be called manually to export iceberg metadata so other client can read the table directly if for whatever reason, they can't access the SQL DB hosting the metadata, I understand the source of the truth is always the SQL DB.
Another reason is if the Metadata DB got corrupted for any reason, a user will have some form of a backup, if all you have is only parquet files, it will be impossible to rebuild the latest state of the table.
Beta Was this translation helpful? Give feedback.
All reactions