Commit Overhead vs Throughput
Commit Overhead vs Throughput
Analysis of commit protocol overhead as percentage of total transaction time.
Overhead = (commit_latency / total_latency) × 100
This represents time spent in retries, exponential backoff, and manifest I/O.
Values shown as mean ± standard deviation across seeds.
num_tables = 1
| Throughput (c/s) | Success Rate (%) | Mean Overhead (%) | P50 Overhead (%) | P95 Overhead (%) | P99 Overhead (%) |
|---|---|---|---|---|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.4 ± 0.0 | 0.4 ± 0.0 | 0.7 ± 0.0 | 0.8 ± 0.1 |
| 0.5 ± 0.0 | 100.0 ± 0.0 | 0.9 ± 0.0 | 0.8 ± 0.0 | 1.2 ± 0.0 | 1.4 ± 0.0 |
| 1.0 ± 0.0 | 100.0 ± 0.0 | 1.6 ± 0.0 | 1.5 ± 0.0 | 2.0 ± 0.1 | 2.2 ± 0.1 |
| 2.0 ± 0.0 | 100.0 ± 0.0 | 3.0 ± 0.0 | 2.9 ± 0.0 | 3.5 ± 0.0 | 3.8 ± 0.1 |
| 4.9 ± 0.0 | 99.8 ± 0.0 | 7.1 ± 0.0 | 7.0 ± 0.0 | 7.9 ± 0.0 | 8.3 ± 0.1 |
| 8.7 ± 0.1 | 89.1 ± 0.5 | 12.6 ± 0.1 | 12.5 ± 0.1 | 13.4 ± 0.1 | 13.8 ± 0.1 |
| 10.5 ± 0.0 | 53.3 ± 0.3 | 15.1 ± 0.0 | 15.0 ± 0.0 | 15.9 ± 0.0 | 16.2 ± 0.0 |
| 11.7 ± 0.0 | 23.2 ± 0.1 | 16.7 ± 0.0 | 16.6 ± 0.0 | 17.4 ± 0.0 | 17.7 ± 0.0 |
| 12.5 ± 0.0 | 12.1 ± 0.0 | 17.8 ± 0.0 | 17.8 ± 0.0 | 18.5 ± 0.0 | 18.8 ± 0.0 |
num_tables = 5
| Throughput (c/s) | Success Rate (%) | Mean Overhead (%) | P50 Overhead (%) | P95 Overhead (%) | P99 Overhead (%) |
|---|---|---|---|---|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.4 ± 0.0 | 0.4 ± 0.0 | 0.7 ± 0.0 | 0.8 ± 0.0 |
| 0.5 ± 0.0 | 100.0 ± 0.0 | 0.8 ± 0.0 | 0.8 ± 0.0 | 1.2 ± 0.1 | 1.3 ± 0.1 |
| 1.0 ± 0.0 | 100.0 ± 0.0 | 1.5 ± 0.0 | 1.5 ± 0.0 | 1.9 ± 0.1 | 2.1 ± 0.1 |
| 2.0 ± 0.0 | 100.0 ± 0.0 | 2.9 ± 0.0 | 2.9 ± 0.0 | 3.4 ± 0.1 | 3.6 ± 0.1 |
| 4.9 ± 0.0 | 100.0 ± 0.0 | 7.0 ± 0.1 | 7.0 ± 0.1 | 7.7 ± 0.1 | 8.1 ± 0.1 |
| 9.9 ± 0.1 | 100.0 ± 0.0 | 13.9 ± 0.1 | 13.9 ± 0.1 | 14.7 ± 0.1 | 15.2 ± 0.2 |
| 19.5 ± 0.1 | 98.9 ± 0.1 | 27.2 ± 0.1 | 27.2 ± 0.1 | 28.2 ± 0.2 | 28.6 ± 0.1 |
| 28.6 ± 0.0 | 57.4 ± 0.2 | 39.9 ± 0.0 | 39.9 ± 0.0 | 40.4 ± 0.0 | 40.6 ± 0.0 |
| 30.6 ± 0.0 | 30.2 ± 0.1 | 42.8 ± 0.0 | 42.7 ± 0.0 | 43.2 ± 0.0 | 43.3 ± 0.0 |
num_tables = 20
| Throughput (c/s) | Success Rate (%) | Mean Overhead (%) | P50 Overhead (%) | P95 Overhead (%) | P99 Overhead (%) |
|---|---|---|---|---|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.3 ± 0.0 | 0.3 ± 0.0 | 0.6 ± 0.0 | 0.8 ± 0.1 |
| 0.5 ± 0.0 | 100.0 ± 0.0 | 0.6 ± 0.0 | 0.7 ± 0.0 | 1.1 ± 0.0 | 1.2 ± 0.0 |
| 1.0 ± 0.0 | 100.0 ± 0.0 | 1.3 ± 0.1 | 1.5 ± 0.0 | 1.9 ± 0.1 | 2.1 ± 0.1 |
| 2.0 ± 0.0 | 100.0 ± 0.0 | 2.8 ± 0.0 | 2.8 ± 0.0 | 3.3 ± 0.1 | 3.5 ± 0.2 |
| 4.9 ± 0.0 | 100.0 ± 0.0 | 6.9 ± 0.0 | 6.8 ± 0.0 | 7.5 ± 0.1 | 7.9 ± 0.1 |
| 9.9 ± 0.1 | 100.0 ± 0.0 | 13.8 ± 0.1 | 13.8 ± 0.1 | 14.6 ± 0.1 | 15.0 ± 0.2 |
| 19.8 ± 0.1 | 100.0 ± 0.0 | 27.5 ± 0.1 | 27.5 ± 0.1 | 28.5 ± 0.1 | 29.0 ± 0.2 |
| 35.2 ± 0.0 | 71.7 ± 0.1 | 48.8 ± 0.0 | 48.8 ± 0.0 | 49.2 ± 0.0 | 49.5 ± 0.0 |
| 38.5 ± 0.0 | 38.4 ± 0.1 | 53.3 ± 0.0 | 53.3 ± 0.0 | 53.7 ± 0.0 | 53.9 ± 0.0 |
num_tables = 50
| Throughput (c/s) | Success Rate (%) | Mean Overhead (%) | P50 Overhead (%) | P95 Overhead (%) | P99 Overhead (%) |
|---|---|---|---|---|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.4 ± 0.0 | 0.4 ± 0.0 | 0.7 ± 0.0 | 0.8 ± 0.0 |
| 0.5 ± 0.0 | 100.0 ± 0.0 | 0.6 ± 0.0 | 0.7 ± 0.0 | 1.1 ± 0.0 | 1.3 ± 0.0 |
| 1.0 ± 0.0 | 100.0 ± 0.0 | 1.0 ± 0.0 | 1.4 ± 0.0 | 1.8 ± 0.0 | 2.0 ± 0.0 |
| 2.0 ± 0.1 | 100.0 ± 0.0 | 2.2 ± 0.1 | 2.7 ± 0.1 | 3.2 ± 0.1 | 3.4 ± 0.0 |
| 4.9 ± 0.1 | 100.0 ± 0.0 | 6.5 ± 0.1 | 6.8 ± 0.1 | 7.4 ± 0.1 | 7.7 ± 0.2 |
| 9.8 ± 0.1 | 100.0 ± 0.0 | 13.4 ± 0.1 | 13.6 ± 0.1 | 14.3 ± 0.1 | 14.8 ± 0.3 |
| 19.7 ± 0.2 | 100.0 ± 0.0 | 27.0 ± 0.2 | 27.1 ± 0.2 | 28.0 ± 0.2 | 28.4 ± 0.2 |
| 38.5 ± 0.1 | 78.9 ± 0.3 | 53.2 ± 0.0 | 53.2 ± 0.0 | 53.7 ± 0.1 | 53.9 ± 0.1 |
| 42.8 ± 0.0 | 43.1 ± 0.1 | 59.0 ± 0.0 | 59.0 ± 0.0 | 59.4 ± 0.0 | 59.5 ± 0.0 |
Interpretation
- Low overhead (<10%): System operating efficiently, minimal contention
- Medium overhead (10-30%): Moderate contention, acceptable performance
- High overhead (30-50%): High contention, commit protocol becoming significant
- Very high overhead (>50%): Commit protocol dominates, system saturated
Values shown as mean ± standard deviation across multiple seeds. At saturation, overhead can exceed 50%, meaning more time is spent retrying commits than executing transactions!