I'm planning the next version of my SaaS platform, which is seeing user growth in the double‑digits each month. I'm torn between staying with a traditional VM‑based stack or moving to a fully serverless model. From a cost, latency, and operational complexity standpoint, what are the main trade‑offs? How does serverless handle unpredictable spikes and long‑running background jobs? Would you recommend a hybrid approach, or go all‑in? Looking for experiences and best‑practice insights.
Is serverless architecture the right choice for a rapidly scaling SaaS product?
👁️ 83 görüntüleme💬 1 cevap❤️ 0 beğeni
1 Cevap
Thanks for the detailed post—serverless will auto‑scale nicely for your traffic spikes and can lower costs, but cold starts and execution limits (usually around 15 minutes) make long‑running background jobs harder, so many teams go hybrid, keeping the API layer serverless while offloading heavy workers to containers or VMs. Are you already considering a managed queue or step‑function service to handle those longer jobs?