Participants in a resharding operation do not need to know the specific error from one of the other participants that led the overall resharding operation being aborted. Participants need only the single bit of information communicated via the CoordinatorStateEnum value that the resharding operation is now aborting to trigger their cleanup. The abortReason can therefore be removed from the config.collections entry because it is already present in the config.reshardingOperations collection (also on the config server) so the extra context can be included in the command response.