Nexmark performance degradation with nightly-20231205

[cyliu0]

Describe the bug

The q4,q9,q102 performance dropped a lot. Meanwhile, the q5-many-windows performance grew a lot.

+-----------------------------------------------------------------+--------------+------------+---------------------+-----------------------------+
| BENCHMARK NAME                                                  | EXECUTION ID | STATUS     | FLUCTUATION OF BEST | FLUCTUATION OF LAST 10 DAYS |
+-----------------------------------------------------------------+--------------+------------+---------------------+-----------------------------+
| nexmark-q4-blackhole-medium-1cn-affinity                        |        16068 | Negative   | -23.90%             | -13.15%                     |
| nexmark-q4-blackhole-medium-1cn                                 |        16069 | Negative   | -33.26%             | -22.21%                     |
| nexmark-q9-blackhole-medium-1cn                                 |        16095 | Negative   | -33.45%             | -24.22%                     |
| nexmark-q9-blackhole-medium-1cn-affinity                        |        16097 | Negative   | -20.60%             | -11.47%                     |
| nexmark-q102-blackhole-medium-1cn                               |        16098 | Negative   | -44.18%             | -31.17%                     |
| nexmark-q102-blackhole-medium-1cn-affinity                      |        16099 | Negative   | -34.39%             | -25.20%                     |
| nexmark-q5-many-windows-blackhole-medium-1cn                    |        16116 | Positive   | 64.52%              | 73.89%                      |
| nexmark-q5-many-windows-blackhole-medium-1cn-affinity           |        16117 | Positive   | 81.68%              | 71.02%                      |

Error message/log

No response

To Reproduce

No response

Expected behavior

No response

How did you deploy RisingWave?

No response

The version of RisingWave

nightly-20231205

Additional context

No response

[cyliu0]

Looks like the performance recovered. @wangrunji0408

[wangrunji0408]

We have identified the problematic PR and reverted it in #13822.