A more efficient implementation of the multi-round block partitioner

[zjma]

Description

Implemented PartitionerV2 which is ~2.5x faster than the default partitioner.
Also included:

• Partitioner selection via envvars.
• A sharded execution bug fix.
• Giving the last executor shard extra threads to handle merged last round.
• Minor refactoring in aptos-block-partitioner.

[github-actions]

✅ Forge suite compat success on aptos-node-v1.6.2 ==> 98156ea4feba78e71659cd7606c4dae953e709f6

Compatibility test results for aptos-node-v1.6.2 ==> 98156ea4feba78e71659cd7606c4dae953e709f6 (PR)
1. Check liveness of validators at old version: aptos-node-v1.6.2
compatibility::simple-validator-upgrade::liveness-check : committed: 4345 txn/s, latency: 7362 ms, (p50: 6500 ms, p90: 11200 ms, p99: 16300 ms), latency samples: 178160
2. Upgrading first Validator to new version: 98156ea4feba78e71659cd7606c4dae953e709f6
compatibility::simple-validator-upgrade::single-validator-upgrade : committed: 1771 txn/s, latency: 16472 ms, (p50: 19000 ms, p90: 22200 ms, p99: 22300 ms), latency samples: 92100
3. Upgrading rest of first batch to new version: 98156ea4feba78e71659cd7606c4dae953e709f6
compatibility::simple-validator-upgrade::half-validator-upgrade : committed: 1768 txn/s, latency: 15630 ms, (p50: 19300 ms, p90: 21900 ms, p99: 22300 ms), latency samples: 91960
4. upgrading second batch to new version: 98156ea4feba78e71659cd7606c4dae953e709f6
compatibility::simple-validator-upgrade::rest-validator-upgrade : committed: 3342 txn/s, latency: 9305 ms, (p50: 10200 ms, p90: 13000 ms, p99: 13700 ms), latency samples: 133700
5. check swarm health
Compatibility test for aptos-node-v1.6.2 ==> 98156ea4feba78e71659cd7606c4dae953e709f6 passed
Test Ok

• Grafana dashboard
• Humio Logs
• Test runner output
• Test run is land-blocking

[github-actions]

✅ Forge suite realistic_env_max_load success on 98156ea4feba78e71659cd7606c4dae953e709f6

two traffics test: inner traffic : committed: 6296 txn/s, latency: 6219 ms, (p50: 6000 ms, p90: 7800 ms, p99: 12300 ms), latency samples: 2726200
two traffics test : committed: 100 txn/s, latency: 2999 ms, (p50: 2900 ms, p90: 3800 ms, p99: 4600 ms), latency samples: 1760
Latency breakdown for phase 0: ["QsBatchToPos: max: 0.224, avg: 0.212", "QsPosToProposal: max: 0.190, avg: 0.157", "ConsensusProposalToOrdered: max: 0.634, avg: 0.594", "ConsensusOrderedToCommit: max: 0.539, avg: 0.517", "ConsensusProposalToCommit: max: 1.168, avg: 1.110"]
Max round gap was 1 [limit 4] at version 1350390. Max no progress secs was 4.022127 [limit 10] at version 1350390.
Test Ok

• Grafana dashboard
• Humio Logs
• Test runner output
• Test run is land-blocking

[github-actions]

✅ Forge suite framework_upgrade success on aptos-node-v1.5.1 ==> 98156ea4feba78e71659cd7606c4dae953e709f6

Compatibility test results for aptos-node-v1.5.1 ==> 98156ea4feba78e71659cd7606c4dae953e709f6 (PR)
Upgrade the nodes to version: 98156ea4feba78e71659cd7606c4dae953e709f6
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 4136 txn/s, latency: 7490 ms, (p50: 7800 ms, p90: 10400 ms, p99: 14400 ms), latency samples: 165460
5. check swarm health
Compatibility test for aptos-node-v1.5.1 ==> 98156ea4feba78e71659cd7606c4dae953e709f6 passed
Test Ok

• Grafana dashboard
• Humio Logs
• Test runner output
• Test run is land-blocking