Cloud storage is the abstraction, pooling, and sharing of storage resources through the internet. Cloud storage is facilitated by IT environments known as clouds, which enable cloud computing—the act of running workloads within a cloud environment. Access cloud storage doesn't require an intranet connection (that’s known as network-attached storage) or a direct connection to storage hardware (that's known as direct-attached storage).
There are 3 types of cloud storage: public cloud storage, private cloud storage, and hybrid cloud storage. There are also 3 ways to format this storage: As blocks, files, or objects. Each format has its pros and cons (blocks are faster, files are easier to understand, and objects work best with quick moving workloads), but some software-defined cloud storage products can combine all 3 formats into a unified, easy-to-deploy solution.
Many organizations are discovering that traditional storage methods can be the bottleneck that slows their agility and scalability. This has led to the development of containers, which allow applications to scale rapidly, be more reliable, and offer better performance than more conventional means or methods.
Data can be acted upon as short-term memory or archived as long-term memory. The short-term memory is handled by random-access memory (RAM), which is responsible for processing and remembering all requests and actions during the time a computer processes specific computations (known as tasks). Once all computations are complete, the data can be stored as long-term memory among different storage volumes, some of which may exist as clouds.
At the base of every cloud storage volume are storage resources abstracted from physical hardware. Virtualization technology is one way of abstraction, and it can take a dozen different servers (commodity or proprietary) and abstract the storage space from all of them. All this virtual storage space can be pooled together into something called a data lake that users can access as a single repository. If those data lakes are connected to the internet, then you’ve created cloud storage.
Block storage splits a single storage volume (like a cloud storage node) into individual instances known as blocks. It's a fast, low latency storage system ideal for high performance workloads.
Object storage involves pairing a piece of data with unique identifiers known as metadata. Since objects are uncompressed and unencrypted, they can be accessed very quickly at huge scale—making them ideal for cloud-native applications.
File storage is the dominant technology used on NAS systems and is responsible for organizing data and representing it to users. Its hierarchical structure allows us to navigate data from top to bottom easily, but increases processing time.
Public cloud storage
Public cloud storage is the storage of data among pools of resources that are abstracted from hardware not owned by the end user. Since there are certain innate risks that come with not owning or managing the systems that are storing your data, many organizations are using containers to move workloads and applications among public cloud environments. Persistent storage solutions (like Red Hat® OpenShift® Data Foundation ) help keep stateful applications from failing and losing all their data.
Private cloud storage
Private cloud storage is the storage of data among pools of resources abstracted from resources solely dedicated to the end user—usually within the user’s firewall and sometimes on premise. Since setting up an enterprise-scale private cloud manually can be less efficient in the long run than using existing software, companies are using platforms like OpenStack® to digitally transform pools of virtual resources into private clouds.
Hybrid cloud storage
Hybrid cloud storage is the storage of data among multiple cloud environments with some degree of workload portability, orchestration, and management among them. While the public and private cloud environments that make up a hybrid cloud are separate entities, migrating data between them is facilitated by a complex network of local-area networks (LANs), wide area networks (WANs), application programming interfaces (APIs), virtual private networks (VPNs), or containers. This separate—yet connected—architecture allows enterprises to store data in any environment and move data between them as desired.
There are important differences between cloud computing and virtualization that can cause some enterprises to favor virtual storage over cloud storage. It may have to do with compliance and regulatory guidelines, but those industries still need to be able to provide cloud-like storage scalability across virtual deployments.
Take organizations with distributed operations, such as the banking industry. Remote branches must access traditional networks through 4 tiers: network, compute, storage, and service-area networks or NAS. And while the ideal solution would be to put an air-conditioned datacenter in the basement of that branch office, it’s not realistic due to space and/or personnel constraints. So, some virtual storage solutions (like Red Hat Hyperconverged Infrastructure) combine 2 tiers to streamline virtual storage deployment and management.
Why choose Red Hat for cloud?
Because we not only help you get to the cloud—we can help you be productive in the cloud. Our open source technologies bring a consistent foundation to any cloud deployment: public, private, hybrid, or multi.
Why choose Red Hat for storage?
Because we take software-defined storage—the inherently open way enterprises are meeting today's storage challenges—one step further by making it open source, giving you total freedom to decide where and how to store your data.