Batch vs Parallel Requests Bench
Compare performance of sequential, parallel, and batched request execution strategies
Configuration

Number of requests per batch

Max parallel requests

Randomness in request duration

Recommendation

For this configuration, batched execution often provides a good balance between parallelism and resource usage.

Fastest Method

Batched

78.4% faster than slowest

Total Time

871ms

Performance Comparison
Sequential
4029ms
5.0 req/s
Parallel
1069ms
18.7 req/s
Batched
871ms
23.0 req/s
Sequential
4029ms

Executes one request at a time, waiting for each to complete before starting the next

0ms4029ms

Throughput

5.0 req/s

Avg Latency

201ms

Min Latency

163ms

Max Latency

239ms

Parallel
1069ms

Executes multiple requests simultaneously, limited by concurrency setting

0ms1069ms

Throughput

18.7 req/s

Avg Latency

201ms

Min Latency

168ms

Max Latency

235ms

Batched
871ms

Divides requests into batches, executing each batch in parallel, then moving to the next batch

0ms871ms

Throughput

23.0 req/s

Avg Latency

194ms

Min Latency

163ms

Max Latency

238ms

馃弳 Winner!

78.4% faster than the slowest method