By: Paul Harris, SoftIron Regional Director, Sales (APAC)
We’ve said it before and we’ll say it again: since when did software-defined mean that hardware doesn’t matter? Managed service providers (MSPs) - and the resellers that service them - are driving demand for a ‘task-specific’ approach when it comes to their underlying storage platform infrastructure. And by ‘task-specific’ we mean purpose-built hardware, designed for software-defined storage, and not compute. We’re witnessing a genuine paradigm shift in the expectations of what commodity hardware can deliver, versus the dawning realisation of the value that can be extracted across the board from this novel approach in the data center.
If the “hardware doesn’t matter” mantra holds true anywhere, you would expect it to be within managed service provision, where the service itself is the “brand” being sold, rather than the underlying hardware. So why is there an increasing interest in our “task-specific” approach?
Several common themes are emerging amongst the many challenges and choices MSPs are faced with today. Here, we highlight the top 4 issues confronting them, and share some insight into why a “task-specific” or “purpose-built” hardware approach for software-defined storage, is gaining popularity.
Dis-aggregating hardware and software platforms
If you’re a service provider (or a reseller selling into one) that is trying to build an architecture where flexibility is the goal, then we’ll often see a design approach that favors dis-aggregating the hardware and software platforms. A “software-defined” approach checks this box. You have the choice of a proprietary software-defined architecture, but that would limit your hardware options to those approved for the software, thereby limiting any genuine future flexibility by tying yourself to a particular vendor and their software platform. This is one of the core reasons that MSPs look to open source as the solution, and the de-facto standard for open source, software-defined storage, is Ceph (pronounced “sef”, if you’re new to this world).
Getting value from all the tools in the “Ceph Swiss Army Knife”
It’s not unusual for service providers to offer a variety of services ranging from backup and disaster recovery, to coverage for “burst” capability when demand quickly escalates. As well as data orchestration across different deployment scenarios and a range of more standard storage offers based on file, block or object storage needs. Ceph’s adaptability, flexibility and infinite scalability make it an ideal contender because of its “swiss army knife-like” ability to handle multiple scenarios and linear scalability. We find that MSPs turn to Ceph as a strategic decision, but then struggle to make the most of it by running it on generic, commodity hardware in their testing. This results in becoming limited by the complexity of its many options in deployment and use, which leads them to very quickly seek alternate methods in which they can simplify Ceph adoption profitably and at scale.
A Secure, Resilient, Scalable Architecture
When your entire business - and the business of your customers - rely on the quality of your service and adherence to customer SLAs, then the resilience of your chosen architecture is paramount. In fact, Data Center Knowledge recently reported that “the risk of an unreliable data infrastructure impacting the economy has become so great that the British government is considering laws to force it to become more resilient.”
Ceph, again, scores well here with its inherent resilience due to the way in which it distributes and replicates data, ensuring no single point of failure. As the number of customers scale and the demands increase, the linear nature by which it allows more nodes to be added is ideal, scaling performance linearly as required.
When it comes to hardware, however, it’s a different story. A highly available, resilient architecture needs intricate alignment between the hardware and software layers – something unachievable with “bare metal” appliances running open-source software. Consider a failed drive, for example, where the replacement needs to be fast, painless, and achieved with zero service disruption. Our task-specific approach enables granular levels of control, allowing admins to visually locate the exact caddy and specific drive that require attention. It also alerts Ceph that a change is about to take place, stopping it from re-configuring when the caddy of drives is temporarily “lost”. When your hardware and software “layers” aren’t talking to each other, you’ll never achieve this level of integration.
The Ability to Make Good Margins
Last, but by no means least, the architecture has to enable MSPs to make a fair margin on their service, from day one. SoftIron’s competitive pricing, over and above the strategic choices highlighted here, directly translate into margin through:
- Multiple product offers on the same platform enable flexibility to differentiate price points of different services and to flex the offer as needs change
- Reduced management complexity needing fewer, less highly skilled staff
- Linear scalability eliminating future “forklift” upgrades of hardware as the business grows
- Integrated support of both hardware and software, eliminating blame games and any “finger-pointing”, speeding resolution of any issue identified
- The ability to move to another hardware vendor’s solution and still run Ceph – ensuring a long term competitive environment for future business
- Ceph’s flexible interfaces enabling the MSP to differentiate in the ways they implement provisioning of services, billing, and multi-tenancy into their architecture
SoftIron’s HyperDrive Storage appliance was recently selected by Enterprise Management Associates (EMA), a leading IT and data management research and consulting firm, to receive a Top 3 Award in their “EMA Top 3 Enterprise Decision Guide 2020” report. HyperDrive was selected as a leading storage solution in the “Hybrid Cloud Management – Enterprise Data Services” category for the awards. The report is available here.
