Blobfuse is an open source project developed to provide a virtual filesystem backed by the Azure Storage. It uses the libfuse open source library (fuse3) to communicate with the Linux FUSE kernel module, and implements the filesystem operations using the Azure Storage REST APIs. BlobFuse lets Linux workloads use Azure Blob Storage like a local file system for several scenarios, including AI/ML training and checkpointing, HPC simulations, Kubernetes stateful workloads, big data analytics/preprocessing, and large-scale backup and archiving. BlobFuse supports standard file operations with caching, offers health monitoring and blob filtering, and can preload containers or folders into the local cache for faster access.

BlobFuse supports two operating modes: a caching mode that works by caching data locally on VM nodes, ideal for repeatedly accessed data that fits on VM local storage, and a streaming mode that reads large files in chunks directly from storage for big AI/ML, genomics, or HPC workloads. This article helps you decide which mode is best suited for your workloads.

Detailed documentation of BlobFuse is here.
Please submit any issues/feature requests/questions here.

You can install BlobFuse from Microsoft repositories for Linux by using simple commands to install the BlobFuse package. If no package is available for your distribution and version, you can build the binary from source code. Refer to the instructions for details.

You can mount a container by using the mount command. You can either include your desired configuration settings as command line parameters or provide a configuration file that contains your settings. Refer the page for details.

By default, BlobFuse logs warnings to the system log. However, you can route logs to a local directory, change which types of information appear in logs, or disable logs entirely by changing the default configuration. Refer the page for details.

Health monitor is a tool that you use to monitor mount activities and resource usage. This article describes what data you can obtain, and how to enable health monitor and view output reports.

To see a list of commands, type blobfuse2 -h and then press the ENTER key. To learn about a specific command, just include the name of the command (For example: blobfuse2 mount -h).

Refer the page for limitations and known issues.
For troubleshooting common issues, refer this page.

This project is licensed under MIT.

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.