It’s time to set the record straight: cloud caching and tiering are not the same.
Given the uptick in hybrid cloud deployments, it’s more important than ever for IT admins to understand the differences between the two technologies.
Tiering and caching are two often-misconstrued methods for data management, each of which combines the benefits of local and cloud storage. Both tiering and caching shift the bulk of data from on-premises storage to the cloud while providing seamless local access.
This is where the similarity ends, as caching and tiering have radically different philosophies. With tiering, your live data lives at the edge and stale data is migrated to the cloud. With cloud caching, all your data lives in the cloud and caches at edge locations provide fast local access.
Don’t Confuse “Copying Data” with “Moving Data”
Both caching and tiering transfer data between the local, on-premises storage tier and the cloud storage tier. A key distinction, however, is that caching copies the data between tiers, while tiering moves the data.
Tiering is an “edge-centric” approach. It moves portions of locally stored data based on predefined criteria from the edge to a lower cost/performance tier (cloud) and retrieves that data upon demand. The local storage is viewed as the primary storage, while the cloud is seen as a place for archiving “cold” data at a reduced cost. Each piece of data is stored in a single tier at any point in time.
In contrast, caching is a cloud-centric approach, in which the cloud hosts the “gold copy” of all your data. On-premises caching devices hold a local copy of frequently accessed data for fast and efficient access.
Common Hybrid Cloud Storage Use Cases
Now that we’ve defined the key attributes of caching and tiering, let’s see how these methods stack up vis-a-vis four use cases for hybrid cloud storage, as defined by Gartner in its new hybrid cloud storage Market Guide.
- Burst for capacity provides infinite and elastic storage capacity expansion for edge devices, spilling excess data into a low-cost cloud storage tier. Since cloud storage is elastic and you pay only for the capacity you use, making it particularly cost efficient for capacity bursting. This is the only use case supported by both tiering and caching.
- Disaster recovery backs up local data to the cloud to enable recovery and business continuity. By storing all your data in highly resilient and redundant cloud storage, caching provides inherent disaster recovery abilities, and more importantly, instant recovery. In case of disaster, a new caching device can be launched anywhere within minutes to immediately provide data access while the cache is being warmed up in the background. Tiering, on the other hand, only keeps cold data in the cloud – i.e., protecting local data is out of scope and requires a separate backup solution.
 Gartner, “Market Guide for Hybrid Cloud Storage, ” May 3, 2021
- Burst for compute is used when the dataset is created locally and needs to be accessed in the cloud for processing or analytics. For example, a visual effects company could launch 1000 servers in the cloud for eight hours to perform rendering of 3D models created by a team of artists working locally. Tiering isn’t suitable for this use case because live data processing (i.e., rendering) cannot take place in the cloud. Caching stores both hot and cold data in the cloud, unleashing the high-performance compute capabilities of the cloud for data analysis and processing.
- Data orchestration is used in hybrid cloud deployments to attain a consolidated view of data in multiple clouds using a single protocol or interface. Consider an enterprise that wants to present a single view of data that can be read and written from multiple edge and cloud locations, move data between locations, or manage access using a single namespace. Tiering doesn’t support this use case, as only cold data is managed in the cloud. Cloud caching, on the other hand, exposes a global multi-cloud file system consolidating the data of any number of backend storage clouds and edge locations into a single namespace accessible from anywhere.
Bottom Line – Cloud Caching vs. Tiering
Caching and tiering are distinct methods for managing data transfer between multiple edge and cloud locations. With tiering, your live data lives at the edge, and stale data is moved to the cloud. With cloud caching, all your data lives in the cloud and is cached at the edge for fast access.
While tiering enables organizations to reduce storage costs, it is limited to the “capacity bursting” use case. Caching is a superior solution for hybrid cloud architectures as it addresses not only capacity cloud bursting but the whole gamut of use cases, including disaster recovery, compute bursting, and data orchestration.