Sharing a filesystem across multiple machines – cluster or distribute?

When you start using technologies such as load balancers and clustering software, you often end up in a situation where you need the same files on multiple servers. While you could simply copy the files, what if you could mount the files on each of the servers, sharing the filesystem across the systems without the SPOF that an NFS server introduces? One of the easiest ways to do this is to use Gluster.

Gluster, also known as GlusterFS, is an open source distributed filesystem that provides scalable and flexible storage for large volumes of data. Initially developed by Gluster Inc., it is now maintained by the open source community. Gluster uses a distributed architecture to create a single and unified filesystem that can span across multiple servers and storage devices. This approach allows you to aggregate the storage capacity of multiple servers and present it as a single, well-structured filesystem to users and applications. It has a wide range of applications, such as data storage, backup, and content delivery.

Key features and concepts of Gluster include the following:

• Scalability: Adding more storage servers to the cluster allows Gluster to easily accommodate growing data storage needs while scaling horizontally.
• Redundancy: Gluster ensures data availability by replicating data across multiple nodes for redundancy and fault tolerance.
• Flexibility: Gluster supports various storage options, including local disks, NAS, and cloud storage. It can be customized to fit specific use cases and technologies.
• Filesystem abstraction: It provides users and applications with a standard filesystem interface, making integration into existing systems relatively easy.
• Data distribution: Data is distributed across the cluster in a way that improves both performance and reliability. Data can be distributed evenly or based on specific criteria.
• Automatic healing: Gluster has a self-healing feature that automatically detects and repairs data inconsistencies or corrupted files.

Gluster is often used in environments where large-scale, distributed storage is required, such as web servers, cloud computing, big data analytics, and media streaming services. It provides a cost-effective and flexible solution for managing data across a network of servers and storage devices.

Getting ready

For this recipe, you will need two Oracle Linux 8 systems, each with access to YUM repos. For this exercise, we will call them gluster1 and gluster2. They are identical systems, each with 8 GB RAM, 4 vCPUs, and 100 GB of drive space. The filesystems have 50 GB in /, 5 GB in /home, and 8 GB in swap. The remaining disk space is unallocated. Additionally, for this example, each node will have a 100 GB LUN used for storing Gluster data.