[SERVER-32940] ReplicationCoordinator waitUntilOpTimeForReadUntil should return NotYetInitialized when oplog has not been created. Created: 26/Jan/18 Updated: 06/Dec/22 Resolved: 12/Apr/18 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Replication |
| Affects Version/s: | 3.7.1 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Matthew Russotto | Assignee: | Backlog - Replication Team |
| Resolution: | Duplicate | Votes: | 0 |
| Labels: | neweng | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||||||
| Assigned Teams: |
Replication
|
||||||||||||
| Operating System: | ALL | ||||||||||||
| Participants: | |||||||||||||
| Linked BF Score: | 16 | ||||||||||||
| Description |
|
If a read with afterClusterTime read concern is received after a heartbeat has set the cluster time but before the oplog has been created, the system will crash trying to access the oplog. Instead, we should return NotYetInitialized to the user. |
| Comments |
| Comment by Eric Milkie [ 26/Jan/18 ] |
|
This should be true for any command that specifies a readConcern prior to the replication subsystem being initialized (that is, after _finishLoadLocalConfig() has returned). |