ChubaoFS

Overview

ChubaoFS (储宝文件系统) is a distributed file system for cloud native applications. It has the following features:

• scale-out metadata management

• strong replication consistency for both append and random write

• specific storage optimizations for large and small files

• multiple volumes

• POSIX-compatible

In particular, ChubaoFS can be used to provide persistent storage volumes for online applications and database services orchestrated by Kubernetes.

Document

https://chubaofs.readthedocs.io/en/latest/

https://chubaofs.readthedocs.io/zh_CN/latest/

Build

Build latest version of chubaofs is simply as following:

$ git clone http://github.com/chubaofs/chubaofs.git
$ cd chubaofs
$ make

If the build is successful, cfs-server and cfs-client will be found in directory build/bin

Docker

Under the docker directory, a helper tool called run_docker.sh is provided to run ChubaoFS with docker-compose.

To start a minimal ChubaoFS cluster from scratch, note that /data/disk is arbitrary, and make sure there are at least 30G available space.

$ docker/run_docker.sh -r -d /data/disk

If client starts successfully, use mount command in client docker shell to check mount status:

$ mount | grep chubaofs

Open http://127.0.0.1:3000 in browser, login with admin/123456 to view grafana monitor metrics.

Or run server and client seperately by following commands:

$ docker/run_docker.sh -b
$ docker/run_docker.sh -s -d /data/disk
$ docker/run_docker.sh -c
$ docker/run_docker.sh -m

For more usage:

$ docker/run_docker.sh -h

License

Licensed under the Apache License, Version 2.0. For detail see LICENSE and NOTICE.

Reference

Reference to cite when you use the system in a research paper or tech report:

Haifeng Liu, et al., CFS: A Distributed File System for Large Scale Container Platforms. SIGMOD‘19, June 30-July 5, 2019, Amsterdam, Netherlands. https://dl.acm.org/citation.cfm?doid=3299869.3314046