streamComputeResult

Stream a compute result instead of fetching it through a URL.

New in v2.

Usage

const stream = await nautilus.streamComputeResult({ jobId: 'a-job-id' }) 
 
for await (const chunk of stream) { 
  process.stdout.write(chunk) 
} 

Preferable for large results: nothing is buffered in memory, and you can begin processing before the transfer finishes.

Returns

• Type: ComputeResultStream

An AsyncIterable<Uint8Array>.

Parameters

config

• Type: ComputeResultConfig

Same as getComputeResult: it streams the job's output result, outputs.tar (a tar archive), unless resultIndex names another.

Throws when the node does not know the job, when the job has not finished, or when it has no such result — the cases where getComputeResult returns undefined.