Prompt
I'm unhappy with serverless timeouts, what platform handles longer jobs?
Latest observation
If your jobs regularly run long enough that serverless timeouts are getting in the way, you probably want a platform designed for long-running workers or batch jobs rather than request/response functions.
Good options:
-
Docker on a VM
Run your app on a small EC2/GCE/Azure VM or a managed VM service. You control the process lifetime, so jobs can run as long as needed. -
Kubernetes Jobs / CronJobs
Good if you already use Kubernetes. Best for queue workers, batch processing, and scheduled jobs. -
Managed batch services
Examples:- AWS Batch
- Google Cloud Run Jobs
- Azure Batch
These are built for compute tasks that may run minutes to hours.
-
Container platforms with longer execution
Some platforms support longer-running containers better than typical serverless functions:- Cloud Run for request-based services, and Cloud Run Jobs for batch tasks
- ECS/Fargate on AWS for containers without managing servers
-
Traditional app hosting / PaaS
Heroku, Render, Fly.io, Railway, etc. can be easier if you just need a worker process and not full infra control.
A common pattern is:
- accept the request quickly,
- enqueue the work,
- let a worker process the job asynchronously,
- notify the user when it’s done.
If you tell me your cloud provider and typical job duration, I can recommend the best fit.