Method: Webhookdb::Replicator::Base.chunked_row_update_bounds

Defined in:
lib/webhookdb/replicator/base.rb

.chunked_row_update_bounds(max_pk, chunk_size: 1_000_000) ⇒ Object

Return an array of tuples used for splitting UPDATE queries so locks are not held on the entire table when backfilling values when adding new columns. See ensure_all_columns_modification.

The returned chunks are like: [[0, 100], [100, 200], [200]], and meant to be used like ‘0 < pk <= 100`, `100 < pk <= 200`, `p, > 200`.

Note that final value in the array is a single item, used like ‘pk > chunks[0]`.



593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
 
# File 'lib/webhookdb/replicator/base.rb', line 593

def self.chunked_row_update_bounds(max_pk, chunk_size: 1_000_000)
  result = []
  chunk_lower_pk = 0
  chunk_upper_pk = chunk_size
  while chunk_upper_pk <= max_pk
    # Get chunks like 0 < pk <= 100, 100 < pk <= 200, etc
    # Each loop we increment one row chunk size, until we find the chunk containing our max PK.
    # Ie if row chunk size is 100, and max_pk is 450, the final chunk here is 400-500.
    result << [chunk_lower_pk, chunk_upper_pk]
    chunk_lower_pk += chunk_size
    chunk_upper_pk += chunk_size
  end
  # Finally, one final chunk for all rows greater than our biggest chunk.
  # For example, with a row chunk size of 100, and max_pk of 450, we got a final chunk of 400-500.
  # But we could have gotten 100 writes (with a new max pk of 550), so this 'pk > 500' catches those.
  result << [chunk_lower_pk]
end