The goal of this ticket is to fix the known retryability issues and consistency issues that can happen when update/delete commands (or Batch/BulkWrite commands containing update/delete) execute concurrently with chunk migrations on a sharded cluster.

As of today (Apr 11 2025), here is the list of known issues

• SPM-3673: Fix updateOne by _id behavior re. chunk migrations for non-retryable writes
• SPM-1631: updateMany without shard key
• SERVER-54019: Retryable write can report it matched and modified more documents than it actually did upon retry, due to interceding chunk migration
• SERVER-8537: Incorrect result numbers for partial updates if there is an error
• SERVER-103725: Retryable write that completed with n=0 or failed with non-retryable error should not execute again on retry

Fixing the issues with multi:true writes in the list above will likely depend on SPM-1153.

Other related/overlapping tickets: SERVER-20361, SERVER-38897, SERVER-47371