What to check before moving backend work to serverless

Match the runtime to the work

Short, stateless request handling is a natural fit. Long-running jobs, heavy startup work and connection-hungry workloads need more care. The platform should simplify the system rather than move its complexity somewhere less visible.

Design for the limits you actually have

Understand execution time, memory, region placement, concurrency and database connections before launch. These are architecture inputs, not details to discover during an incident.

• Measure cold-start sensitivity
• Plan database connection behaviour
• Keep background jobs separate where needed
• Add logs and traces before production

export default defineEventHandler(async (event) => {
  const body = await readBody<{ email?: string }>(event)

  if (!body.email) {
    throw createError({ statusCode: 400, message: 'Email is required' })
  }

  return await subscribers.add(body.email)
})