A High-Performance, Cloud-Native, Distributed File System

Built for developers, Easy to use

• Using like a local disk

  # Format JuiceFS file system
  > juicefs format redis://your-redis-host:6379/1 myjfs
  
  # Mount JuiceFS file system
  > juicefs mount -d redis://your-redis-host:6379/1 /mnt/juicefs
  
  # Check mount point status
  > df -h /mnt/juicefs
  Filesystem      Size   Used  Avail Capacity iused    ifree %iused  Mounted on
  JuiceFS:myjfs  1.0Pi    0Bi  1.0Pi     0%       0 10485760    0%   /mnt/juicefs
  
  # Copy dataset from home to JuiceFS
  > cp -r ~/dataset /mnt/juicefs/

• Simple development, not dependent on SDK

  # Python example.py
  path = '/mnt/juicefs/dataset/days.txt'
  days_file = open(path, 'r')
  days = days_file.read()
  
  new_path = '/mnt/juicefs/new_days.txt'
  new_days = open(new_path, 'w')
  
  title = 'Days of the Week\n'
  new_days.write(title)
  print(title)
  
  new_days.write(days)
  print(days)
  
  days_file.close()
  new_days.close()

• For Kubernetes PV

  apiVersion: v1
  kind: PersistentVolumeClaim
  metadata:
    name: web-pvc
  spec:
    accessModes:
      - ReadWriteMany
    resources:
      requests:
        storage: 10Pi
    storageClassName: juicefs-sc
  ---
  apiVersion: apps/v1
  kind: Deployment
  metadata:
    name: nginx-run
  spec:
    selector:
      matchLabels:
        app: nginx
    template:
      metadata:
        labels:
          app: nginx
      spec:
        containers:
          - name: nginx
            image: linuxserver/nginx
            ports:
              - containerPort: 80
            volumeMounts:
              - mountPath: /config
                name: web-data
        volumes:
          - name: web-data
            persistentVolumeClaim:
              claimName: web-pvc
  
  > kubectl get sc
  NAME                 PROVISIONER                RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
  juicefs-sc           csi.juicefs.com            Retain          Immediate           false                  88m

• Using like HDFS

  # Shell
  > hadoop fs -ls jfs://myjfs
  
  # Hive
  CREATE TABLE IF NOT EXISTS person
  (
    name STRING,
    age INT
  ) LOCATION 'jfs://myjfs/tmp/person';

• Access via S3 API

  # Start S3 Gateway for JuiceFS file system
  > juicefs gateway redis://localhost:6379/1 localhost:9000
  
  # Access ​​JuiceFS file system via AWS CLI
  > aws --endpoint-url http://localhost:9000 s3 ls s3://myjfs