Let JuiceFS do your “offsite backup” for you


JuiceFS is a distributed POSIX file system for the cloud that stores data in your own public cloud object store while presenting a full POSIX filesystem interface to your tools, languages and platforms.

Regardless if you have data in a in-house solution or public cloud, you can use JuiceFS to perform offsite backup. Just follow our documentation to mount JuiceFS to your host computer or in your public cloud host. Then use standard backup tools, such as rsync, to copy the backup data directly into the JuiceFS filesystem. JuiceFS transmissions are encrypted and will automatically transfer large files in parallel.  JuiceFS also handles unreliable public networks – if you’re stuck behind an overloaded or temperamental firewall your data will still get through.

If you use JuiceFS to store data or do backups already, we have a powerful feature that allows you to easily perform an offsite backup: Replication.  JuiceFS Replication will asynchronously copy all written data to another specified object store (on any cloud vendors or service regions).

As an example, assume your main business is in AWS East, and data needs to be backed up to Azure West. You only need to create a file system in AWS East, and then enable replication (and select Azure West). All data written to AWS S3 (including before enabling the copy function) will be automatically copied to Azure Blob Storage. When you need to perform data recovery in the Azure West area, it will read data directly from Azure Blob storage, which is fast and does not require traffic charges.

Another important feature of the enterprise version of JuiceFS is Global Data Mirroring, which can help you achieve near-real-time data mirroring (read-only) at ultra-long distances, such as mirroring from the United States to China, or vice versa. Compared with the data replication function mentioned above, data mirroring also proposes a read-only mirroring of metadata to ensure good performance for ultra-distant mirrored data. We found in testing that from AWS East to Tencent Cloud Shanghai, metadata is only delayed in seconds, and most of the data is synchronized within 30 seconds. Occasionally an encrypted connection is blocked by a wall and data synchronization is delayed. It also proactively repairs synchronization to ensure correct and consistent data access (slightly slower when accessed). The current data mirroring function is only available to enterprise customers, please contact us for more details.

We believe that JuiceFS is the best data backup solution in the market today, because it:

  1. Supports over 13 public cloud vendors and 69 regions and our platform is continually growing
  2. Ensures strong consistency, and 99.95% high availability (Enterprise Edition is 99.99%)
  3. Supports encryption in-transit, Enterprise Edition supports storage encryption
  4. Is compatible with POSIX, can be mounted to the VM through FUSE, and the user experience is consistent with using a local disk
  5. Is server-less, fully maintained by us and the cloud provider, no maintenance required from customer
  6. Provides rich monitoring that can be integrated in the customer’s own monitoring system
  7. Supports replication, which can help you backup or migrate your data to another vendor or region easily
  8. Provides recycle bin to effectively prevent accidental deletion
  9. Saves time and cost. Taking into account both maintenance cost and storage cost, JuiceFS can save 50% to 80% over other solutions.

By the way, if you are using Linux or Mac, JuiceFS can be directly mounted on your own computer. The above method is also suitable for backing up your personal data. We also provide 1TB permanent free space (You only need to pay for object storage from your cloud vendor).

Have another usage scenario you’d like to share?  Reach us at hello at juicedata.io, or speak to us via Intercom widget on the JuiceFS website ( https://juicefs.io ).  The widget is in the lower right corner.

Haven’t signed up for JuiceFS yet?  Get a 1TB filesystem for free here: https://juicefs.io

JuiceFS News: Kubernetes support, faster performance, small file support, new vendors: Microsoft Azure and Netease Cloud

Here are some of the most recent updates to JuiceFS:

  • New algorithms to significantly increase random write performance
  • Copy-free fast file merging – large performance benefits for big data analysis scenarios
  • Client-side cache sharing to improve read performance of hot data
  • Kubernetes Flex storage volume support
  • New cloud vendors: Microsoft Azure and Netease Cloud(cn)

For a more detailed update log, see our version update page.

We recommended re-mounting to upgrade to the latest version to obtain these improvements.

In addition, we have upgraded the documentation center, where you can find detailed architecture introductions, getting started guides, ways to use JuiceFS in Hadoop and Spark, solutions for using JuiceFS in Kubernetes, and more.

If you have any questions about using the JuiceFS or new challenges in data storage, hail us online at the green icon in the bottom right corner of the website.  Or email us: hello at juicedata.io

Welcome to JuiceFS!


JuiceFS is a POSIX shared file system designed for the cloud and contains three aspects:

  1. Designed for the Cloud: JuiceFS was born entirely for cloud computing, popularly known as Cloud Native. We believe we’ve created the best storage solution in the cloud computing space. We  support all the global mainstream cloud platforms in the west and the east, including Amazon Web Services (AWS), Google Cloud Platform (GCP), Microsoft Azure, Digital Ocean, Backblaze B2, Ali Cloud, Tencent Cloud, UCloud, Qingyun, Jinshan Cloud, NetEase Cloud, and Qiniu storage platform. Engineers using these cloud platforms can easily get started with JuiceFS. JuiceFS saves data in the account of the customer’s own object store – giving you complete control of your data. Data security is guaranteed, storage scale is as flexible as the underlying platform and does not require additional maintenance.
  2. Compatible with POSIX: Unlike object storage, JuiceFS uses the oldest and most widely used POSIX APIs in the storage domain, (ie, open, read, write, and close, etc).  POSIX is supported by almost all programming languages and data management platforms. At the OS level, JuiceFS is implemented via FUSE and is directly attached to the host and looks just like a local filesystem. A large number of existing applications and tools can be directly accessed without any modification, just as simple as using a local disk.
  3. Consistent Multi-Client Sharing: This is the biggest difference between JuiceFS and many cloud drives. JuiceFS can be mounted at the same time by multiple machines (supporting more than 1000 nodes without a single VPC limit) while reading and writing, and provides guaranteed strong consistency.

What is the value of these three points? An almost unlimited capacity storage system that can be accessed from anywhere.  JuiceFS has a low learning curve, no/super-low maintenance, and makes many distributed systems challenges simply go away.  Some example use cases include:

  • Data backup and recovery: because POSIX is the favorite interface for operation and maintenance engineers and none of them.
  • A full-scale data store for Hadoop or Spark clusters, including data warehouses such as Hive.
  • A shared storage volume in a container cluster to easily persist and share container data.
  • Replace existing NAS or NFS setups with unlimited capacity cloud storage.
  • Data mirroring: simplifying the challenge of moving data across multiple locations.
  • A global shared space to facilitate the exchange of engineers or hosts, eliminating the need for copying files and having versions get out of sync.
  • and many, many more uses.

If you have any data challenges, our expert storage team is here to help. Please email us at: hello at juicedata.io

To start a free 1TB trial, signup at: https://juicefs.io