The commit of a chunk migration triggers the execution of a transaction with multiple conditional steps, which can summarised as:

• persist updated data in config.chunks
• If the placement of the parent collection is impacted by the migrated chunk(s), also persist updated data into config.placementHistory:
  • if prior information of the parent collection is available in config.placementHistory, fetch it, adapt it and re-insert it
  • otherwise, extract up-to-date placement data from config.chunks and persist it into config.placementHistory.

Such logic is currently distributed through several future sub-chains, which complicates the understanding of the whole process.

The objective of this ticket is to provide a more explicit alternative.