FAQ

Questions about running compute jobs with ON? You’ll find the answers here.
Don’t see yours? Reach out!
What if my service crashes or stops?

A crashed or stopped service keeps its slot, ports, and remaining paid time reserved until it expires, it isn't torn down early. From the management page you can hit Restart (alongside Edit and Prolong) to bring it back without losing your endpoint or paid time.

What if my launch fails?

Live container logs on the management page show why, most often a rejected launch flag, a gated repo that needed a token, or a quantization the GPU can't run.

What types of workloads can I run on an Ocean Node?
You can run containerized compute jobs like embeddings, model inference jobs, data cleanup, batch processing, and fine-tune model workloads that finish within the job window and produce outputs you can download.
How can I check the status of my compute job?
To check your job status and history, open the Ocean Network dashboard, go to Profile, and view Job Status to see all current and past jobs. Users can also check job status in Ocean Orchestrator and Ocean-cli
How do paid compute jobs work?
Paid jobs charge per run based on resource time and environment selection
Do I need to set up servers as a compute user?
No. This is no server setup and serverless GPU compute: you choose a preferred Ocean Node with the resources you need, submit a containerized job, and get results back without managing servers or infrastructure
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.