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 | 0.57 ± 0.01 | 1.25 ± 0.22 | 3.89 ± 1.52 | 1.1 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.58 ± 0.00 | 2.53 ± 0.46 | 11.89 ± 1.79 | 1.2 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.59 ± 0.00 | 6.14 ± 1.25 | 23.93 ± 2.11 | 1.4 |
| 1.6 ± 0.0 | 99.7 ± 0.1 | 0.75 ± 0.03 | 9.11 ± 0.92 | 41.96 ± 5.00 | 2.1 |
| 2.0 ± 0.0 | 98.6 ± 0.2 | 0.86 ± 0.01 | 11.26 ± 1.67 | 53.85 ± 4.63 | 2.5 |
| 2.5 ± 0.0 | 93.6 ± 0.4 | 1.12 ± 0.01 | 13.75 ± 0.94 | 66.93 ± 6.96 | 3.3 |
| 3.1 ± 0.0 | 75.2 ± 0.4 | 1.47 ± 0.01 | 11.46 ± 1.74 | 71.98 ± 9.18 | 4.2 |
| 3.7 ± 0.0 | 44.8 ± 0.1 | 1.78 ± 0.02 | 3.73 ± 0.10 | 56.98 ± 5.77 | 4.9 |
| 4.1 ± 0.0 | 24.8 ± 0.1 | 1.87 ± 0.02 | 3.53 ± 0.01 | 34.79 ± 3.96 | 5.3 |
| 4.4 ± 0.0 | 10.7 ± 0.0 | 1.92 ± 0.02 | 3.47 ± 0.01 | 16.79 ± 3.99 | 5.4 |
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 | 0.57 ± 0.01 | 1.03 ± 0.11 | 2.68 ± 0.44 | 1.1 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.58 ± 0.00 | 1.70 ± 0.42 | 6.31 ± 1.17 | 1.2 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.59 ± 0.00 | 2.49 ± 0.58 | 11.38 ± 2.03 | 1.3 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.65 ± 0.00 | 5.22 ± 0.34 | 21.59 ± 1.57 | 1.7 |
| 2.0 ± 0.0 | 100.0 ± 0.0 | 0.69 ± 0.01 | 6.99 ± 0.81 | 26.82 ± 2.49 | 2.0 |
| 2.7 ± 0.0 | 99.4 ± 0.1 | 0.81 ± 0.01 | 8.23 ± 0.65 | 36.08 ± 3.05 | 2.6 |
| 3.9 ± 0.0 | 93.8 ± 0.2 | 1.13 ± 0.02 | 10.86 ± 0.71 | 51.03 ± 2.75 | 3.8 |
| 5.0 ± 0.0 | 60.8 ± 0.2 | 1.71 ± 0.01 | 9.43 ± 0.75 | 53.57 ± 3.20 | 5.5 |
| 5.4 ± 0.0 | 32.9 ± 0.1 | 1.91 ± 0.02 | 7.87 ± 0.74 | 53.21 ± 2.96 | 5.9 |
| 5.7 ± 0.0 | 13.9 ± 0.0 | 1.92 ± 0.01 | 6.35 ± 0.38 | 53.78 ± 3.44 | 5.9 |
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 | 0.56 ± 0.00 | 0.81 ± 0.04 | 1.33 ± 0.18 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.57 ± 0.01 | 0.98 ± 0.05 | 2.48 ± 0.29 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.59 ± 0.00 | 1.52 ± 0.25 | 4.78 ± 0.63 | 1.3 |
| 1.7 ± 0.0 | 100.0 ± 0.0 | 0.63 ± 0.01 | 2.34 ± 0.35 | 9.04 ± 0.47 | 1.6 |
| 2.1 ± 0.0 | 100.0 ± 0.0 | 0.66 ± 0.00 | 3.08 ± 0.16 | 12.47 ± 0.72 | 1.8 |
| 2.7 ± 0.0 | 100.0 ± 0.0 | 0.70 ± 0.00 | 3.80 ± 0.37 | 15.40 ± 1.28 | 2.1 |
| 4.1 ± 0.0 | 99.6 ± 0.1 | 0.79 ± 0.01 | 5.20 ± 0.37 | 22.70 ± 1.36 | 2.8 |
| 7.3 ± 0.0 | 88.8 ± 0.1 | 1.20 ± 0.02 | 8.29 ± 0.81 | 37.74 ± 2.02 | 4.5 |
| 9.8 ± 0.0 | 59.5 ± 0.2 | 1.58 ± 0.01 | 8.61 ± 0.34 | 45.60 ± 3.29 | 5.7 |
| 11.5 ± 0.0 | 28.0 ± 0.1 | 1.86 ± 0.01 | 7.92 ± 0.14 | 48.87 ± 1.08 | 6.2 |
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 | 0.55 ± 0.00 | 0.75 ± 0.02 | 0.98 ± 0.08 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.57 ± 0.01 | 0.85 ± 0.02 | 1.60 ± 0.11 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.59 ± 0.00 | 1.13 ± 0.13 | 2.96 ± 0.34 | 1.3 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.63 ± 0.00 | 1.68 ± 0.18 | 5.28 ± 0.57 | 1.6 |
| 2.1 ± 0.0 | 100.0 ± 0.0 | 0.65 ± 0.00 | 1.79 ± 0.04 | 6.09 ± 0.44 | 1.7 |
| 2.7 ± 0.0 | 100.0 ± 0.0 | 0.69 ± 0.00 | 2.38 ± 0.15 | 8.34 ± 0.42 | 2.0 |
| 4.1 ± 0.0 | 99.9 ± 0.0 | 0.75 ± 0.00 | 3.22 ± 0.23 | 11.81 ± 0.43 | 2.5 |
| 7.7 ± 0.0 | 93.4 ± 0.2 | 1.02 ± 0.01 | 5.09 ± 0.16 | 21.45 ± 0.86 | 4.1 |
| 10.9 ± 0.0 | 66.7 ± 0.2 | 1.30 ± 0.00 | 6.27 ± 0.24 | 28.41 ± 1.00 | 5.3 |
| 13.1 ± 0.0 | 32.0 ± 0.1 | 1.46 ± 0.00 | 6.39 ± 0.22 | 31.06 ± 0.90 | 5.9 |
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 | 0.53 ± 0.01 | 0.71 ± 0.03 | 0.85 ± 0.08 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.56 ± 0.00 | 0.78 ± 0.01 | 1.13 ± 0.13 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.58 ± 0.00 | 0.91 ± 0.03 | 1.60 ± 0.16 | 1.2 |
| 1.7 ± 0.0 | 100.0 ± 0.0 | 0.62 ± 0.00 | 1.13 ± 0.02 | 3.03 ± 0.24 | 1.5 |
| 2.0 ± 0.0 | 100.0 ± 0.0 | 0.65 ± 0.00 | 1.26 ± 0.03 | 3.36 ± 0.19 | 1.7 |
| 2.7 ± 0.0 | 100.0 ± 0.0 | 0.68 ± 0.00 | 1.55 ± 0.09 | 4.55 ± 0.25 | 1.9 |
| 4.1 ± 0.0 | 99.9 ± 0.0 | 0.73 ± 0.00 | 2.01 ± 0.07 | 6.44 ± 0.44 | 2.4 |
| 7.8 ± 0.0 | 94.4 ± 0.2 | 0.94 ± 0.00 | 3.11 ± 0.12 | 11.23 ± 0.38 | 3.9 |
| 11.3 ± 0.0 | 69.1 ± 0.2 | 1.19 ± 0.00 | 3.93 ± 0.11 | 15.32 ± 0.79 | 5.2 |
| 13.5 ± 0.0 | 32.9 ± 0.1 | 1.33 ± 0.00 | 4.37 ± 0.14 | 17.91 ± 0.78 | 5.8 |
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 | 0.45 ± 0.01 | 0.67 ± 0.01 | 0.76 ± 0.02 | 1.0 |
| 0.4 ± 0.0 | 100.0 ± 0.0 | 0.53 ± 0.00 | 0.74 ± 0.01 | 0.88 ± 0.03 | 1.1 |
| 0.8 ± 0.0 | 100.0 ± 0.0 | 0.56 ± 0.00 | 0.80 ± 0.00 | 1.03 ± 0.04 | 1.2 |
| 1.6 ± 0.0 | 100.0 ± 0.0 | 0.60 ± 0.00 | 0.93 ± 0.02 | 1.53 ± 0.17 | 1.5 |
| 2.0 ± 0.0 | 100.0 ± 0.0 | 0.63 ± 0.00 | 1.03 ± 0.02 | 1.86 ± 0.15 | 1.6 |
| 2.8 ± 0.0 | 100.0 ± 0.0 | 0.67 ± 0.00 | 1.13 ± 0.02 | 2.15 ± 0.06 | 1.8 |
| 4.1 ± 0.0 | 99.9 ± 0.0 | 0.72 ± 0.00 | 1.41 ± 0.02 | 2.99 ± 0.13 | 2.3 |
| 7.8 ± 0.0 | 95.0 ± 0.2 | 0.89 ± 0.00 | 1.98 ± 0.02 | 5.20 ± 0.13 | 3.8 |
| 11.5 ± 0.0 | 70.0 ± 0.1 | 1.12 ± 0.01 | 2.39 ± 0.03 | 7.10 ± 0.16 | 5.1 |
| 13.7 ± 0.0 | 33.3 ± 0.1 | 1.25 ± 0.00 | 2.54 ± 0.03 | 8.11 ± 0.17 | 5.8 |
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
