Table of Contents
File storage is still a very important part of every enterprise data storage infrastructure, and there is no surprise that now, more than ever, users ask for file services in the cloud too.
Cloud providers neglected to add file services to their product portfolio at the beginning, concentrating on block and object storage. Albeit these two types of storage cover many use cases, and new applications can be developed to use them without needing file storage, the reality is that there are several circumstances where files are preferable:
- Lift and Shift: With more and more enterprises opting for the public cloud as their primary IT infrastructure, it is very common now to see “lift and shift” migrations. In this scenario, the users want to replicate the same services they had in the on-premises data center, including (POSIX)-compliant file systems, data services, and all other enterprise features they are accustomed to.
- Simplicity: File storage is the most user-friendly storage, and many developers still prefer this familiar interface to others because it gives them the ability to build even more portable applications while simplifying sharing machine- and human-generated data.
- Performance: Albeit object storage performance is improving by leaps and bounds, file systems still provide the best combination of performance, usability, and scalability for many workloads. It is still the primary interface for the majority of big data, artificial intelligence / machine learning (AI/ML), and high performing computing (HPC) applications, and usually offers data services such as snapshots to improve data management operations.
In the last few years, file systems have also become more cloud-friendly, showing better integrations with object storage. This brought several advantages to end users:
- Better scalability. Policy-driven tiering mechanisms enable us to move cold data to Amazon’s Simple Storage Service (S3)-compatible storage and save precious resources in the high-performance tier.
- Best combination of speed and $/GB. Some file storage gateways are specifically designed to work with an object storage back-end.
- Simplify data migrations and synchronization. Many file storage systems can replicate data to remote file or object stores, in the cloud or on-premises. This allows us to synchronize and serve data sets across different infrastructures to optimize compute-data vicinity, hence latency.
- Disaster Recovery (DR). Syncing data to a remote object store enables the user to leverage a cheaper storage repository in the cloud and populate a file system only if necessary.
These functionalities are particularly important now that vendors are optimizing their file storage for flash memory and access speed so that the user can build a multi-tier infrastructure to optimize $/GB as well. Furthermore, users are looking at hybrid and multi-cloud for their infrastructure strategy, and this storage infrastructure design perfectly fits in this context.
With this in mind, it is easy to understand the reasons behind the success of file storage in the cloud, especially if it can be integrated with on-premises and cloud storage as well.
A Key Criteria Report analyzes the most important features of a technology category to understand how they impact an enterprise and its IT organization. Features are grouped into four categories:
- Table Stakes
- Key Criteria
- Critical Impact of Features on the Metrics
- Near-term Game-changing Technology
The goal is to help organizations assess capabilities and build a mid-to-long-term infrastructure strategy. In a mature technology, the solutions are divided into three target market categories: enterprise, high-performance, and specialized solutions. In a mature market, these differ in their characteristics and how they can be integrated with existing infrastructures. That said, the assessment is more dependent on the specific user’s needs and not solely on the organization’s vertical.
Table stakes are system characteristics and features that are important when choosing the right solution. They include architectural choices that depend on the size of the organization, its requirements, the expected growth over time, and the types of workloads. Table stakes are mature, and the implementation of these features will not add any business advantage nor significantly change the total cost of ownership (TCO) or return on investment (ROI) of the infrastructure.
Key criteria features really differentiate one solution from another. Depending on real user needs, they have a positive impact on one or more of the metrics mentioned. Therefore, implementation details are essential to understanding the benefits relative to the infrastructure, processes, or business. Following table stakes and key criteria, aspects like architectural design and implementation regain importance and need to be analyzed in great detail. In some cases, the features described in the Key Criteria section are the core solution, and the rest of the system is designed around them. This could be an important benefit for organizations that see them as a real practical advantage, but it also poses some risks in the long term. In fact, over time, the differentiation introduced by a feature becomes less relevant and falls into the table-stakes group, while new system capabilities introduce new benefits or address new needs, with a positive impact on the metrics like efficiency, manageability, flexibility and so on.
Key criteria bring several benefits to organizations of all sizes with different business needs. The section is organized to give the reader a brief description of the specific functionality or technology, its benefits in general terms, and what to expect from a good implementation. In order to give a complete picture, we also include examples of the most interesting implementation currently available in the market.
Critical Impact of Features on the Metrics
Technology, functionality, and architecture designs that have demonstrated their value are adopted by other vendors, become a standard, and lose their status as a differentiator. Initially, the implementation of these key criteria was crucial for delivering real value, perhaps with some trade-offs. The most important metrics for the evaluation of a technology solution include:
- Manageability and ease-of-use
This section documents the impact that individual features have on the metrics at the moment of report publication. Each feature is scored from 1 to 5, with a score of five having the most impact on an enterprise. This is not absolute and should always be verified with the organization’s requirements and use case. Strategic decisions can then be based on the impact each metric can have on the infrastructure, system management, and IT processes already in place with particular emphasis on ROI and TCO.
Near-Term Game-Changing Technology
In this report section, we analyze the most interesting technologies on the horizon over the next twelve to eighteen months. Some are already present in some form but usually as part of niche products or for addressing very specific use cases. In either case, at this stage, the implementations available are not mature enough to be grouped in key criteria. Yet when implemented correctly and efficiently, the technologies can really make a difference to the metrics.
Over time, game-changing features become key criteria, and the cycle repeats. Therefore, to get the best ROI, it is important to check what vendors are offering today and what they plan to release in the near future.
The Key Criteria Report is part of a series of documents aimed at giving the reader tools to understand technology more fully, evaluate it, and explore the market to find the best solutions for their organization.
In this context, and to get a complete view of the state of the solutions available in the market, the reader should consider the following documents:
- Key Criteria for Evaluating Hosted Kubernetes report is an introduction to the technology. It defines the necessary evaluation metrics, the key criteria that may be used to evaluate new solutions, and the impact of the latter on the former. It is dedicated to those end-users that are approaching a new technology for the first time or want an update on the latest evolution.
- GigaOm Radar for Hosted Kubernetes offers a brief 360° view of the market, including market and technical positioning of most notable vendors, a short introduction of their solutions and differentiation, including a high-level graphic comparison of the vendors.
- Vendor Profiles for Hosted Kubernetes are easy-to-read deep-dive documents that cover a single vendor regarding the solutions described in the other reports. They provide more details on the solution, how the vendor approached the key criteria, and the impact that its solutions have on the evaluation metrics. This document helps end-users to get a quick but complete evaluation of a single vendo