In the Iceberg Working Session, we briefly discussed that we could potentially benefit from moving away from SNAPPY and into a more modern analytics compression algorithm.

In this task, we should:

• Pick one of our big tables (>= 1 TB) that we think would benefit
• Backfill Iceberg tables with different compression algorithms, such as SNAPPY, GZIP, and ZSTD.
• Run a couple of WMF real world queries on each.
• Compare and contrast SELECT benefits against compression time and compression ratio.
• Decide if we should move out