[SERVER-63163] exclude TTL reads from operation resource consumption calculation Created: 01/Feb/22 Updated: 22/May/23 Resolved: 22/May/23 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | Eric Milkie | Assignee: | Backlog - Storage Execution Team |
| Resolution: | Won't Do | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||
| Assigned Teams: |
Storage Execution
|
||||
| Participants: | |||||
| Description |
|
Since the TTL reader runs every minute, it can very quickly accumulate index read units in the operationMetrics. We should decide how we want to count these. |
| Comments |
| Comment by Lyon Chen [ 02/May/22 ] |
|
After conversation with Product, we are lowering the RPU price and fixed a related TTL issue, so we are okay to close this issue as Won't Do. cc chris.shum@mongodb.com kevin.jernigan@mongodb.com vishal.dhiman@mongodb.com |
| Comment by Louis Williams [ 01/Feb/22 ] |
|
One option would be to only accumulate metrics when a TTL pass has deleted data. If it did not, then we should not include the reads. Another, slightly more complex, option would be to keep track of the oldest key in each index and only schedule the TTL deleter when we expect there to be data to delete. |