ConfigsApr 16, 2026·3 min read

Fission — Fast Serverless Functions on Kubernetes

Fission is an open-source serverless framework for Kubernetes that lets developers deploy functions in Python, Go, Node.js, and other languages with sub-second cold starts and no container builds required.

TL;DR
Fission deploys serverless functions in Python, Go, Node.js, and more on Kubernetes with sub-second cold starts and no container builds.
§01

What it is

Fission is an open-source serverless framework that runs on Kubernetes. It lets developers deploy functions in Python, Go, Node.js, Java, and other languages without building container images. Fission maintains pre-warmed containers (environments) that load function code at invocation time, achieving sub-second cold starts.

Fission targets teams running Kubernetes who want serverless capabilities without vendor lock-in. It brings the function-as-a-service model to any Kubernetes cluster, on-premises or cloud.

§02

How it saves time or tokens

Fission eliminates the Docker build step from the deploy cycle. Write a function, upload it, and it runs immediately on your existing Kubernetes cluster. Pre-warmed environments mean cold starts are sub-second instead of the multi-second delays common with container-based serverless. HTTP triggers, message queue triggers (Kafka, NATS), and timer triggers are configured declaratively.

§03

How to use

  1. Install the Fission CLI and deploy Fission to your Kubernetes cluster.
  2. Create an environment for your language: fission env create --name python --image fission/python-env.
  3. Deploy a function: fission function create --name hello --env python --code hello.py.
§04

Example

# Install Fission CLI
curl -Lo fission https://github.com/fission/fission/releases/latest/download/fission-cli-linux
chmod +x fission && sudo mv fission /usr/local/bin/

# Install on K8s cluster
kubectl create -k 'github.com/fission/fission/crds/v1?ref=v1.20.5'
kubectl create namespace fission
helm install fission fission-charts/fission-all --namespace fission

# Create a Python environment
fission env create --name python --image fission/python-env

# Deploy a function
echo 'def main():
    return "Hello from Fission"' > hello.py
fission function create --name hello --env python --code hello.py

# Create HTTP trigger
fission route create --method GET --url /hello --function hello

# Test
curl http://$FISSION_ROUTER/hello
§05

Related on TokRepo

§06

Common pitfalls

  • Fission requires a running Kubernetes cluster. It is not a standalone serverless platform. If you do not already run Kubernetes, the operational overhead may outweigh the benefits.
  • Pre-warmed environments consume cluster resources even when idle. Size your environment pool based on expected concurrency.
  • Function dependencies must be baked into the environment image or uploaded as packages. Large dependency sets require custom environment images.

Frequently Asked Questions

How does Fission achieve sub-second cold starts?+

Fission maintains pre-warmed container pools (environments) that are already running and ready. When a function is invoked, Fission loads the function code into an existing container rather than starting a new one from scratch.

What languages does Fission support?+

Fission supports Python, Go, Node.js, Java, .NET, Ruby, PHP, and Perl through official and community environments. You can also create custom environments for any language.

How does Fission compare to AWS Lambda?+

Fission runs on your own Kubernetes cluster, avoiding vendor lock-in and cloud costs. Lambda is fully managed with no infrastructure to maintain. Fission gives you more control; Lambda gives you less operational burden.

Does Fission support event triggers?+

Yes. Fission supports HTTP triggers, timer (cron) triggers, message queue triggers (Kafka, NATS), and Kubernetes watch triggers. You configure triggers declaratively via the CLI.

Can Fission auto-scale functions?+

Yes. Fission integrates with Kubernetes HPA (Horizontal Pod Autoscaler) to scale function pods based on request volume. It also supports pool-based scaling for the pre-warmed environment model.

Citations (3)

Discussion

Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.

Related Assets