Commit Latency vs Throughput
Commit Latency vs Throughput
Analysis of how commit latency scales with achieved throughput.
Values shown as mean ± standard deviation across seeds.
num_tables = 1
| Throughput (c/s) | Success Rate (%) | P50 Latency (s) | P95 Latency (s) | P99 Latency (s) | Mean Retries |
|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 1.38 ± 0.56 | 4.45 ± 1.37 | 1.1 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 2.44 ± 0.75 | 9.72 ± 0.54 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.34 ± 0.00 | 4.64 ± 1.04 | 22.64 ± 3.20 | 1.2 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.35 ± 0.00 | 9.89 ± 0.89 | 45.58 ± 2.49 | 1.5 |
| 2.1 ± 0.0 | 100.0 ± 0.0 | 0.36 ± 0.00 | 10.95 ± 1.00 | 53.33 ± 3.03 | 1.7 |
| 2.8 ± 0.0 | 99.7 ± 0.1 | 0.40 ± 0.00 | 15.00 ± 1.62 | 71.11 ± 2.25 | 2.0 |
| 4.0 ± 0.0 | 97.4 ± 0.2 | 0.49 ± 0.00 | 15.61 ± 1.60 | 94.81 ± 3.61 | 2.7 |
| 5.9 ± 0.0 | 72.4 ± 0.6 | 0.77 ± 0.01 | 1.78 ± 0.00 | 34.74 ± 2.82 | 4.4 |
| 6.8 ± 0.0 | 41.7 ± 0.2 | 0.90 ± 0.00 | 1.75 ± 0.01 | 1.94 ± 0.01 | 5.0 |
| 7.7 ± 0.0 | 18.7 ± 0.1 | 0.96 ± 0.01 | 1.75 ± 0.00 | 1.89 ± 0.01 | 5.3 |
num_tables = 2
| Throughput (c/s) | Success Rate (%) | P50 Latency (s) | P95 Latency (s) | P99 Latency (s) | Mean Retries |
|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.32 ± 0.00 | 1.15 ± 0.39 | 4.50 ± 1.94 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 3.12 ± 0.76 | 11.42 ± 2.09 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 4.47 ± 0.54 | 22.40 ± 1.63 | 1.2 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.34 ± 0.00 | 8.80 ± 0.91 | 46.12 ± 6.26 | 1.4 |
| 2.0 ± 0.0 | 100.0 ± 0.0 | 0.34 ± 0.00 | 11.75 ± 0.67 | 56.07 ± 5.79 | 1.5 |
| 2.7 ± 0.0 | 100.0 ± 0.0 | 0.34 ± 0.00 | 14.59 ± 0.84 | 72.78 ± 5.31 | 1.7 |
| 4.1 ± 0.0 | 99.9 ± 0.0 | 0.36 ± 0.00 | 22.54 ± 0.79 | 111.44 ± 5.54 | 2.1 |
| 7.8 ± 0.0 | 95.3 ± 0.1 | 0.48 ± 0.00 | 14.54 ± 1.00 | 146.34 ± 4.23 | 3.3 |
| 11.6 ± 0.0 | 71.2 ± 0.3 | 0.72 ± 0.00 | 1.59 ± 0.01 | 1.79 ± 0.01 | 4.8 |
| 13.8 ± 0.0 | 34.0 ± 0.1 | 0.86 ± 0.01 | 1.62 ± 0.00 | 1.77 ± 0.00 | 5.6 |
num_tables = 5
| Throughput (c/s) | Success Rate (%) | P50 Latency (s) | P95 Latency (s) | P99 Latency (s) | Mean Retries |
|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.32 ± 0.00 | 0.86 ± 0.21 | 3.82 ± 1.45 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 2.53 ± 0.66 | 9.13 ± 2.32 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 4.95 ± 1.29 | 21.63 ± 3.80 | 1.2 |
| 1.7 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 9.71 ± 2.26 | 44.98 ± 7.32 | 1.4 |
| 2.1 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 11.82 ± 1.57 | 56.01 ± 2.50 | 1.5 |
| 2.8 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 15.83 ± 1.58 | 78.58 ± 7.63 | 1.6 |
| 4.1 ± 0.0 | 100.0 ± 0.0 | 0.34 ± 0.00 | 24.87 ± 1.85 | 109.62 ± 8.14 | 1.8 |
| 8.1 ± 0.0 | 99.8 ± 0.0 | 0.35 ± 0.00 | 48.60 ± 3.00 | 238.45 ± 13.76 | 2.5 |
| 15.1 ± 0.0 | 92.6 ± 0.1 | 0.40 ± 0.00 | 1.05 ± 0.01 | 51.67 ± 3.79 | 2.9 |
| 28.8 ± 0.1 | 73.7 ± 0.2 | 0.71 ± 0.00 | 1.55 ± 0.00 | 1.72 ± 0.00 | 5.0 |
num_tables = 10
| Throughput (c/s) | Success Rate (%) | P50 Latency (s) | P95 Latency (s) | P99 Latency (s) | Mean Retries |
|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.31 ± 0.00 | 1.14 ± 0.33 | 4.64 ± 0.75 | 1.1 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.32 ± 0.00 | 2.90 ± 0.93 | 12.00 ± 3.07 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 4.86 ± 0.87 | 21.28 ± 2.50 | 1.2 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 9.25 ± 1.65 | 44.68 ± 1.95 | 1.4 |
| 2.1 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 11.22 ± 1.14 | 53.49 ± 7.41 | 1.4 |
| 2.7 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 15.38 ± 2.21 | 71.99 ± 8.48 | 1.6 |
| 4.1 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 24.66 ± 1.30 | 114.59 ± 7.41 | 1.8 |
| 8.2 ± 0.1 | 100.0 ± 0.0 | 0.34 ± 0.00 | 48.00 ± 3.30 | 231.87 ± 7.58 | 2.2 |
| 15.5 ± 0.0 | 95.2 ± 0.0 | 0.34 ± 0.00 | 0.96 ± 0.02 | 142.38 ± 3.31 | 2.6 |
| 36.5 ± 0.0 | 97.4 ± 0.1 | 0.45 ± 0.00 | 1.12 ± 0.01 | 1.52 ± 0.01 | 3.3 |
num_tables = 20
| Throughput (c/s) | Success Rate (%) | P50 Latency (s) | P95 Latency (s) | P99 Latency (s) | Mean Retries |
|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.29 ± 0.01 | 0.85 ± 0.46 | 4.35 ± 1.13 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.31 ± 0.00 | 2.50 ± 0.60 | 10.25 ± 3.30 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.32 ± 0.00 | 4.91 ± 1.10 | 24.62 ± 5.16 | 1.2 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 8.46 ± 1.06 | 42.37 ± 3.05 | 1.3 |
| 2.1 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 11.57 ± 1.10 | 54.17 ± 6.00 | 1.4 |
| 2.7 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 15.73 ± 0.86 | 81.85 ± 3.03 | 1.5 |
| 4.1 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 22.59 ± 1.22 | 108.15 ± 6.04 | 1.7 |
| 8.2 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 54.14 ± 4.55 | 252.54 ± 11.43 | 2.1 |
| 15.8 ± 0.1 | 97.1 ± 0.1 | 0.33 ± 0.00 | 37.69 ± 1.47 | 239.20 ± 11.53 | 2.5 |
| 37.0 ± 0.1 | 98.6 ± 0.1 | 0.35 ± 0.00 | 0.68 ± 0.00 | 0.94 ± 0.01 | 2.5 |
num_tables = 50
| Throughput (c/s) | Success Rate (%) | P50 Latency (s) | P95 Latency (s) | P99 Latency (s) | Mean Retries |
|---|
| 0.2 ± 0.0 | 100.0 ± 0.0 | 0.24 ± 0.01 | 0.66 ± 0.47 | 3.54 ± 1.05 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.29 ± 0.00 | 2.30 ± 0.96 | 10.82 ± 2.63 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.31 ± 0.00 | 4.33 ± 0.74 | 19.39 ± 2.44 | 1.2 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.32 ± 0.00 | 9.02 ± 0.35 | 42.52 ± 3.53 | 1.3 |
| 2.1 ± 0.0 | 100.0 ± 0.0 | 0.32 ± 0.00 | 11.85 ± 1.43 | 57.42 ± 1.33 | 1.4 |
| 2.7 ± 0.0 | 100.0 ± 0.0 | 0.32 ± 0.00 | 14.45 ± 1.66 | 75.20 ± 4.07 | 1.5 |
| 4.1 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 22.54 ± 0.39 | 115.93 ± 2.58 | 1.6 |
| 8.2 ± 0.0 | 100.0 ± 0.0 | 0.33 ± 0.00 | 52.94 ± 2.83 | 250.78 ± 11.04 | 2.0 |
| 16.0 ± 0.1 | 98.9 ± 0.0 | 0.33 | 85.53 ± 3.12 | 403.19 ± 10.18 | 2.4 |
| 36.9 ± 0.1 | 98.6 ± 0.1 | 0.33 ± 0.00 | 0.52 ± 0.00 | 0.66 ± 0.00 | 2.2 |
Notes
- Latencies reported in seconds (converted from milliseconds)
- Values shown as mean ± standard deviation across multiple seeds
- Throughput = commits per second during steady-state window
- Success rate = percentage of transactions that committed successfully
- Mean retries = average number of retry attempts per committed transaction
