Which technique is recommended to reduce network latency for large payloads?

Which technique is recommended to reduce network latency for large payloads?

• A. Disable compression
• B. Increase fetch size
• C. Avoid streaming
• D. Compress large payload (gzip)

Answer: (D)

Reducing network latency for large payloads hinges on shrinking the data that must traverse the network. Compressing large payloads with gzip makes the data smaller, so the time to send it over the network drops. The saved transmission time often outweighs the CPU cost of compressing and decompressing, especially when the payload is large and highly compressible. In MuleSoft, enabling gzip compression for the payload reduces the amount of data sent in the response, speeding up perceived latency. The other options don’t directly cut the amount of data transmitted: disabling compression increases data size; increasing fetch size typically means more data is retrieved at once and can raise latency; avoiding or not streaming doesn’t necessarily reduce transfer time and mainly affects memory and flow control rather than latency. So compressing large payloads with gzip is the best approach.

Reducing network latency for large payloads hinges on shrinking the data that must traverse the network. Compressing large payloads with gzip makes the data smaller, so the time to send it over the network drops. The saved transmission time often outweighs the CPU cost of compressing and decompressing, especially when the payload is large and highly compressible. In MuleSoft, enabling gzip compression for the payload reduces the amount of data sent in the response, speeding up perceived latency. The other options don’t directly cut the amount of data transmitted: disabling compression increases data size; increasing fetch size typically means more data is retrieved at once and can raise latency; avoiding or not streaming doesn’t necessarily reduce transfer time and mainly affects memory and flow control rather than latency. So compressing large payloads with gzip is the best approach.