Nexenta Blog

Global Leader in Software Defined Storage

storage

Raise Your SDS IQ (2 of 6): Practical Review of Scale-Up Software-Only SDS

By

by Michael Letschin, Field CTO

This is the second of six posts (the last one was Scale-up Vendor-Defined “SDS”) where we’re going to cover some practical details that help raise your SDS IQ and enable you to select the SDS solution that will deliver Storage on Your Terms.  The second SDS flavor in our series is Scale-up Software-Only SDS.

Scale-up Software-Only SDS is just that – SDS benefits delivered via a software-only approach; the software sits on industry standard servers (think Cisco, Dell, Supermicro) and can leverage a variety of JBODs (like Fujitsu, Supermicro, Quanta).  The end result is similar to what you get with Scale-up Vendor-Defined “SDS” — one or two commodity head nodes with JBOD behind it – but with two big differences: much better cost efficiency because it’s more vendor agnostic, and greater opportunity for scale up by leveraging big disks for the JBODs (like Supermicro’s 90-bay). Both provide REST-based management with rich APIs via front-end software, to enable easy, automatic provisioning and management of storage. Unlike Vendor-Defined, Scale-up Software-Only SDS offers good scalability and excellent cost efficiency and flexibility because of it’s software-only approach.

Scale-up Software-Only SDS is an excellent option for companies running virtual machines or the legacy apps that help run their business. It’s a good choice if you have accounting applications and legacy hardware because its performance in this use case is excellent. Unlike vendor-defined options, this one gives you the flexibility to design your solution as needed; and that applies both to the solution you need today, as well as creating flexibility for the future.. Plus, you can choose the most cost-effective commodity-priced machines rather than paying a premium for proprietary hardware – we’ve seen as much as 50-80% cost difference between identical solutions of the different types.  Scale-up Software-Only SDS is limited to one or two head nodes, but you can put larger drives behind them, so you can scale up pretty big, but obviously scale out is limited.

Overall grade: A-

See below for a typical build and the report card:

Screen Shot 2016-05-24 at 10.11.24 AM

It’s time to raise your SDS IQ

By

by Michael Letschin, Field CTO

If you’re like other storage buyers – you’re going to invest in a solution, you want storage on your terms – optimized for your organization, its requirements, now and in the future. When it comes to distinguishing the wealth of Software Defined Storage (SDS) solutions from one another, you probably have a better shot of telling monkeys apart (note: there are 260 species of monkeys). Even respected analysts like Gartner, IDC, 451 Research and TechTarget have different SDS definitions – SDS must be software only, SDS can be hyperconverged, SDS is open source, or SDS can be hardware-based.   What most people seem to agree on is that SDS enables storage services through a software interface, and often runs on commodity hardware, enabled by the decoupling of storage software and hardware.

Yet that still doesn’t help answer the question, what meets YOUR needs? It may seem a little unconventional for a vendor blog, but our goal in this series (expect another six blogs after this one) is to give you some practical information on SDS types – what are the flavors, what works best where, how different SDS types rate against common use cases, and what you should select to bring up your organization’s SDS IQ.

We’re going to cover six types of SDS solutions:

  • Scale-up Vendor-Defined “SDS”
  • Scale-up Software Only
  • Scale Out
  • Hyperconverged
  • Virtual Storage Appliance
  • Containerized

Review our report cards to see whether your favorite SDS solution made the grade – we’ll look at each type and rate them on four critical categories: flexibility, scalability, performance, and cost; we’ll suggest the best use cases for each solution, and even share a few vendors to look at. We’ll be using a 5 point grading system:

  • A: Excellent; well-rounded and recommended solution
  • B: Very Good; above average solutions, especially for certain use cases
  • C: Passing; improvement needed for overall usage
  • D: Close fail; almost passing, solution with numerous gaps
  • F: Failing, not a workable solution

What’s on your SDS wish list?

To help you raise your SDS IQ, it’s helpful to start by doing your homework – what’s on your SDS wish list? For example, making sure you’re still in charge of managing drives, so you can handle predictable drive failures. Many organizations also want policy-based provisioning using REST-based APIs, specifically thin provisioning and scripted storage solution. Tiering is also often a must-have for SDS because of its ability to match data with storage types and maximize your return on investment. You might also be looking for SDS that’s independent of hypervisors. Hyperconvergence expands the portfolio of solutions even further. Take a few minutes to think through what matters most, and we’ll help you figure out how to get it.

Watch this space for the first review in our series – Scale-up, Vendor-Defined “SDS”

Questions from the Field: How do you define Software-defined?

By

By Michael Letschin, Field CTO

If you’re in IT, and attending the usual industry events, you can’t help but notice the explosion of companies, from those making software to those making hardware, claiming to have Software-defined solutions.  I even bumped into someone (we’ll leave him nameless) who claimed that his company’s flash controller was “software-defined”, because after all, it was software that defined how the hardware should be managed; right?  Right …

Yes, software needs hardware, but that doesn’t make the resulting solution Software-defined.  While there are an increasing number of Software-defined solutions out there, it’s still a bit of the Wild West, and buyers best beware.  Have some healthy caution as you explore solutions and understand how your vendor defines Software-defined.  Getting that base-level understanding is important, because the solutions that flow from the definition have different characteristics that either will or won’t work with your use case.

So, how do we define Software-defined?  Well, we are Nexenta after all, the storage software company; we would say that the only true SDS solutions are ones where the software is hardware agnostic, architecturally flexible, and able to deliver business agility (along with the usual storage software features).  But don’t take our word for it, for a deeper dive into SDS definitions, check out George Crump’s latest blog “What Exactly is Software Defined Storage“.  The important takeaway here is that the definition of SDS matters when you’re trying to solve a problem – it defines the benefits that the solution is capable of delivering to you.

Next week we’ll be starting a blog series on how to Raise Your SDS IQ, where we’ll walk through the six different types of SDS, as the industry defines them, and explain where and why they excel, and fall down.  So, watch this space as we work to build out your buyer’s toolkit; that way you won’t be the guy (or gal) with the knife at the gun fight ;).

Using the Right Storage Protocol for the Right Use Case

By

By Michael Letschin, Field CTO

IT professionals have no shortage of storage protocols to choose from, such as NFS, SMB, Fibre Channel (FC), iSCSI and Object. “Experts” are writing books about which protocol is best, usually taking the side of a vendor with a particular axe to grind. The truth is they each have their sweet spot. The key is to make sure that your storage solution is flexible enough to support all your data center’s needs at the same time.

Virtualization

In most data centers the virtual infrastructure supports the majority of the business critical applications and workloads. The virtualization platform of choice, at least for now, is VMware. While FC is still very prevalent in VMware environments and VVOLs makes FC more flexible, we believe NFS is the ideal choice for most VMware environments. Let’s face it, VMs are essentially files and what better way to store files than a protocol designed specifically for file based data like NFS. The advantages of NFS are well documented but the key is that NFS provides a much easier mapping of a VM to its datastore. You can now make decisions, like what tier of storage to place a VM on, at a discrete VM level.

Mission Critical Applications

Many environments, for a variety of reasons, choose not to virtualize certain mission critical applications. They may already be clustered or there may be too many performance concerns. For these situations, many data centers will leverage a block protocol like FC or iSCSI. If the high performance storage requires low latency access, then FC is ideal, but iSCSI can hold its own for other applications where latency is not critical. Again, your storage software should give you the flexibility to choose any or all of these protocols as it makes sense.

Files

Managing file data, or unstructured data for those who want to sound cool, is one of the biggest challenges facing IT. And just like applications not all this data is equal. Most IT professionals immediately think of user data here, created by office productivity applications. It needs to be put on moderately performing storage but not the fastest storage since most users today are connecting via WiFi or even broadband. You need to keep it a long time because users never want you to delete their files. For this data, assuming most users are running Windows, SMB is the protocol of choice.

Another type of file data comes from machines like cameras, recording devices and sensors. It can range in size from trillions of very small sensor files to a few very large files from video cameras. The industry will tell you that object storage is the way to go here, and it very well may be. But we encourage you to use NFS first. It takes a lot of data to exceed the maximum potential of a modern NFS server. Again, the storage solution should not force you to make a choice.

At the other end of the file spectrum is high performance data that you need to access rapidly or a process that needs to write data quickly. For this, NFS is ideal. It is a high performance file system and with the appropriate use of flash delivers the performance that these use cases demand.

Conclusion

If you noticed, NFS is most appropriate in the majority of the use cases but not all of them. We think the storage solution you use should not also force you into a specific protocol. You should choose the one that makes sense for your specific use cases. And that’s why we built NexentaStor.

Questions from the Field: Hyperconvergence

By

By Michael Letschin, Field CTO

Having the most complete portfolio of Software-Defined Storage solutions in the industry is something that Nexenta pride’s itself on, but with that comes questions about all sorts of other storage technologies when I am out talking with our customers and partners.  Their questions range from trying to understand the latest trends like enterprise containers to the impact of the Internet of  Things and augmented reality, but more often than not their questions are about how some of the newer datacenter technologies will help their business.  These technologies range from Software-Defined Networking to public cloud and of course hyperconverged solutions.  Sometimes we integrate very well, for instance with the VMware vCloud Air where we can run inside the public cloud and be a DR target for our existing customers, at a public cloud price point.  In other environments we have to explain that many solutions are not one size fits all.  Hyperconverged falls into that camp and George Crump, an analyst for Storage Switzerland, has a great write-up here – Is Hyperconverged worth the Hype? – on the pros and cons to that market.  I think a key takeaway is that if you’re looking at a new project or a new datacenter with fixed needs, then the simplicity of hyperconverged could be the answer; but if you are growing a datacenter or expect unpredictable growth there are some caveats: the inability to separate storage and compute as you grow can result in over-buying, and losing the benefits you get from virtualization and consolidation.  In those cases, the idea of a traditionally isolated compute and storage solution has real benefit.  Utilizing new technologies like Software-Defined Storage to give you the flexibility of choosing the right hardware for you when you need it gets the enterprise closer and closer to the dream of a next generation or Software-Defined Datacenter.

For more on hyperconverged, check out Is Hyperconverged worth the Hype? by George Crump at Storage Swiss.