[SERVER-41166] Allow the router to send commitTransaction to all participants in parallel for single-write-shard transactions whose read-only participants have already made the reads durable Created: 15/May/19 Updated: 12/Dec/23 |
|
| Status: | Open |
| Project: | Core Server |
| Component/s: | Sharding |
| Affects Version/s: | None |
| Fix Version/s: | 4.3 Desired |
| Type: | Improvement | Priority: | Major - P3 |
| Reporter: | Esha Maharishi (Inactive) | Assignee: | Backlog - Cluster Scalability |
| Resolution: | Unresolved | Votes: | 0 |
| Labels: | ShardedTxn:FutureOptimizations, sharding-common-backlog | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Assigned Teams: |
Cluster Scalability
|
| Backwards Compatibility: | Fully Compatible |
| Participants: |
| Description |
|
If the router sees all the read-only shards have already committed their read Timestamp (by monitoring their last committed OpTimes through their ReplicaSetMonitors), the router could send commitTransaction to the read-only shards in parallel with sending commitTransaction to the write-shard, and only wait for the write-shard's response before returning the outcome to the client. |