Deploy Any App to Alibaba Cloud SAE in Two Simple Steps

Serverless Application Engine (SAE) is Alibaba Cloud's zero‑code, containerized hosting platform that offers stable, elastic service for thousands of enterprises, simplifying Kubernetes adoption with a low learning curve.

Two‑step deployment: prepare the code repository and create the application

Step 1: Prepare the code repository – Fork an open‑source project such as Excalidraw to your own Git account (any Git source like GitHub, Gitee, GitLab, or Codeup can be used).

Excalidraw is a popular open‑source sketching board on GitHub. GitHub address: https://github.com/excalidraw/excalidraw License: MIT License

Step 2: Create the application and bind the source repository

When creating the app, fill in the application name, HTTP listening port (use 3000 for Excalidraw), request timeout (set to 60 seconds), and keep other settings at their defaults.

HTTP listening port: 3000 (the default port used by Excalidraw)

Request timeout: 60 seconds (to allow the service to start up)

Bind the source repository, select the deployment mode “continuous deployment from source”, and configure the following:

Build type: “system auto‑detect and build image” – no Dockerfile required.

Start command: the command that launches the Excalidraw service. cd ./excalidraw-app && vite --host After completing the configuration, choose “skip advanced settings and create the app”, which triggers the build and deployment of the application version.

SAE supports multiple Git source types, automatic image building, various trigger modes for continuous deployment, an internal image registry (with optional push to a personal ACR), and customizable start commands and working directories.

Visualization of the process

During the build, logs and timing for each stage are observable. Once the build succeeds, a new application version is created; you can switch all traffic to the new version via the version list.

Code security

Each build runs in a fully isolated, disposable runtime environment. Network isolation ensures security between runtimes, and the build environment is destroyed after use, leaving no trace of the source code.

Performance and experience

SAE optimizes cloud‑side build time to approach local build speed through multi‑level caching:

Runtime optimization + image cache enables cold starts under 3 seconds.

High‑frequency dependencies are baked into the runtime image for instant reuse.

Medium‑frequency dependencies are cached in OSS for fast retrieval without external network access.

Low‑frequency dependencies are cached in the image after the first pull; subsequent builds reuse the cache.

SAE 2.0 capability upgrades

Elastic capability: scaling latency reduced to the hundred‑millisecond level, support for scaling down to zero, and idle‑billing where CPU is not charged when there are no requests.

Developer experience: zero‑migration, instant app creation, one‑click gray release, and enriched platform engineering features such as S2A to boost development efficiency.

Applicable scenarios: seamless migration for open‑source microservices (e.g., XXL‑Job, ElasticJob, Spring Task) and new support for web‑application types, covering both instance hosting and traffic hosting.