Attu

Attu is an all-in-one milvus administration tool. With Attu, you can dramatically reduce the cost of managing milvus.

Installation Guides

Before you begin, make sure that you have Milvus installed on either Zilliz Cloud⁠ or your own server⁠.

Compatibility

Running Attu from Docker

Here are the steps to start a container for running Attu:

docker run -p 8000:3000 -e MILVUS_URL={milvus server IP}:19530 zilliz/attu:v2.4

Make sure that the Attu container can access the Milvus IP address. After starting the container, open your web browser and enter http://{ Attu IP }:8000 to view the Attu GUI.

Optional Environment Variables for Running Attu Docker

To run the Docker container with these environment variables, use the following command:

Attu SSL Example

docker run -p 8000:3000 \
-v /your-tls-file-path:/app/tls \
-e ATTU_LOG_LEVEL=info  \
-e ROOT_CERT_PATH=/app/tls/ca.pem \
-e PRIVATE_KEY_PATH=/app/tls/client.key \
-e CERT_CHAIN_PATH=/app/tls/client.pem \
-e SERVER_NAME=your_server_name \
zilliz/attu:dev

Custom Server Port Example

This command lets you run the docker container with host networking, specifying a custom port for the server to listen on

docker run --network host \
-v /your-tls-file-path:/app/tls \
-e ATTU_LOG_LEVEL=info  \
-e SERVER_NAME=your_server_name \
-e SERVER_PORT=8080 \
zilliz/attu:dev

Running Attu within Kubernetes

Before you begin, make sure that you have Milvus installed and running within your K8's Cluster⁠. Note that Attu only supports Milvus 2.x.

Here are the steps to start a container for running Attu:

kubectl apply -f https://raw.githubusercontent.com/zilliztech/attu/main/attu-k8s-deploy.yaml

Make sure that the Attu pod can access the Milvus service. In the example provided this connects directly to my-release-milvus:19530. Change this based on the Milvus service name. A more flexible way to achieve this would be to introduce a ConfigMap. See this example for details.

Install Desktop application

If you prefer to use a desktop application, you can download the desktop version of Attu⁠.

Build Docker Image Locally

• Dev: yarn run build:dev
• Release: yarn run build:release

FAQ

• I can't log into the system

  Make sure that the IP address of the Milvus server can be accessed from the Attu container. #161⁠

• TBD

More Screenshots

Data view	collection tree

create collection	vector search

✨ Contributing Code

Thank you for your interest in contributing to Attu! Here's how you can build Attu locally to contribute code, test out the latest features, or try out an open PR:

Build the Server

1. Fork and clone the Attu repository.
2. Navigate to the server directory by running cd server in the terminal.
3. Install dependencies by running yarn install.
4. To start the server in development mode, run yarn start.
5. Create a new branch for your PR by running git checkout -b my-branch.

Build the Client

1. Fork and clone the Attu repository.
2. Navigate to the client directory by running cd client in the terminal.
3. Install dependencies by running yarn install.
4. To start the server in development mode, run yarn start.
5. Create a new branch for your PR by running git checkout -b my-branch.

Submitting a Pull Request

1. Make changes and ensure that tests pass.
2. Commit changes and push to your fork.
3. Create a Pull Request targeting the main branch of Attu.

We appreciate your contributions to Attu, regardless of size. Thanks for supporting the project!

❓ Do you have any questions or problems?

If you encounter any bugs or want to request a new feature, please create a GitHub issue⁠. It's important to check if someone else has already created an issue for the same problem before submitting a new one.

Userful links

Here are some helpful resources to get you started with Milvus:

• Milvus documentation⁠: Here, you can find detailed information on how to use Milvus, including installation instructions, tutorials, and API documentation.
• Milvus python SDK⁠: The Python SDK allows you to interact with Milvus using Python. It provides a simple and intuitive interface for creating and querying vectors.
• Milvus Java SDK⁠: The Java SDK is similar to the Python SDK but designed for Java developers. It also provides a simple and intuitive interface for creating and querying vectors.
• Milvus Go SDK⁠: The Go SDK provides a Go API for Milvus. If you're a Go developer, this is the SDK for you.
• Milvus Node SDK⁠: The Node SDK provides a Node.js API for Milvus. If you're a Node.js developer, this is the SDK for you.
• Feder⁠: Feder is a JavaScript tool designed to aid in the comprehension of embedding vectors.

Community

💬 Join our vibrant community on the Milvus Discord where you can share your knowledge, ask questions and engage in meaningful conversations. It's not just about coding, it's about connecting with other like-minded individuals. Click the link below to join now!