tarun11Mavani commented on PR #18760:
URL: https://github.com/apache/pinot/pull/18760#issuecomment-4705040889

   ### Performance Validation (JMH)
   
   Ran `BenchmarkFunnelCountAggregation` locally (JDK 21, 1 fork, 3 warmup / 5 
measurement iterations of 5s each) on a 10,000-row block with 4 funnel steps.
   
   **Single-key path (existing queries — must show zero regression):**
   
   | Strategy | Throughput (ops/s) | Error (±) |
   |---|---|---|
   | bitmap | 275.2 | 17.3 |
   | set | 276.6 | 11.9 |
   | theta_sketch | 1663.3 | 125.7 |
   | partitioned | 273.9 | 11.4 |
   | partitioned_sorted | 3299.4 | 1222.0 |
   
   **Multi-key path (new feature):**
   
   | Strategy | Throughput (ops/s) | Error (±) |
   |---|---|---|
   | bitmap | 372.0 | 25.2 |
   | set | 370.5 | 25.3 |
   | theta_sketch | 288.3 | 7.7 |
   | partitioned | 371.6 | 21.8 |
   | partitioned_sorted | 928.9 | 5.1 |
   
   Single-key throughput is unchanged from baseline (upstream master without 
multi-key code) — the separate `addSingleKey`/`addMultiKey` method split 
ensures the hot path carries no additional overhead. Multi-key is expectedly 
slower due to composite-ID computation per row, but well within acceptable 
bounds for the added functionality.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to