SimpliVity Announces OmniStack Version 3.5
Last week, I had the opportunity to participate in SimpliVity’s first “SimpliVity Connect” conference where a number of partners, influencers, analysts, and star customers gathered to discuss what SimpliVity is doing in the market and what comes next for them. (Disclosure)
Unfortunately, the most exciting parts of the event to me personally were under NDA. However, some of the exciting news is announced this morning. SimpliVity is announcing the release of the next iteration of OmniStack software which brings a number of neat enhancements. They’re also announcing new tools, and support for a new hypervisor.
The theme of this release is the capability to scale to serve even larger customers. Many of the improvements are targeted at enterprise customers that require more flexibility and customizability out of the platform. There are quite a few enhancements in this release; I’m going to cover two that I think are the most interesting.
Intelligent Workload Optimizer
The Intelligent Workload Optimizer has – quite frankly – been a missing piece of the puzzle that competitors have already addressed. It’s good to see it added, but what’s even better is that the way it has been implemented is fundamentally different than the way others seem to have done it; I think the way SimpliVity is choosing to handle the problem is more efficient than other ways I’ve seen it addressed.
The Intelligent Workload Optimizer is about data locality. In a hyperconverged infrastructure environment, it’s possible that although the storage is inside of the servers, the data for a particular workload is not on the disks inside the node where the workload is running. This causes the storage system to have to traverse the network to access the data, which causes a dramatic increase in latency, even if it’s in the neighboring node. Network latency is substantially higher than bus latency, even in the best case.
The goal of Intelligent Workload Optimizer is to move workloads within the cluster in order to make sure that workloads run on the same node that is storing the data. This ensures the best storage performance without risking availability. The key difference between the OmniStack implementation and other data locality implementations is which component is migrated. In the case of other solutions I’m aware of, the data migrates to be local to the running workload. Conversely, Intelligent Workload Optimizer migrates the running workload to be local to the data. See Figure 1. Consider the difference between vMotion-ing a file server (IWO) and Storage vMotion-ing a file server’s disks (other solutions) and you’ll see why this is important.
Offering a REST API is all the rage for the purpose of supporting the automation and orchestration efforts of enterprise customers. With the big push for private or hybrid cloud models in the enterprise, the ability to programmatically interact with the platform is crucial if it’s going to be used to build an infrastructure providing cloud services.
The API being introduced can be browsed using a UI that is clean and intuitive. Sometimes the biggest barrier to levering APIs is documentation. In this case, it doesn’t seem like that should be any issue. My initial look at the API browser suggests that it will be very easy to understand and consume. See Figure 2.
OmniView – Predictive Analytics
I’ve been writing a lot recently about how analytics are the game changer for any platform (especially storage) over the next year. OmniView is not exactly a part of the 3.5 release, but it’s being announced in tandem with this release. With OmniView, SimpliVity is announcing capabilities to provide predictive insights into your environment including utilization and trending. It’s managed through a separate portal because the data is hosted in a SimpliVity data center. This allows SimpliVity to utilize it for other things as well, like fielding support calls and already having the data handy. They can also crunch the data from all customers to make smart decisions about roadmap and product direction.
Customers are increasingly asking for choice, and rightly so. With that in mind, most ISVs that began by making products that integrate with vSphere are being asked to make the same product integrate with Hyper-V (and others). SimpliVity is answering with support for Hyper-V coinciding with the release of Windows Server 2016. Note that there was a decision regarding development efforts to not support older versions of Hyper-V, so in order to take advantage of the SimpliVity OmniStack platform with Hyper-V, you’ll need to be running Server 2016. Nonetheless, this is an important announcement; in a recent survey we conducted (the full report is forthcoming), roughly 67% of respondents reported running vSphere, and 37% said that they were running Hyper-V. With almost 40% of the market leveraging Hyper-V in some form, it’s a platform that has to be addressed, and I’m glad to see that SimpliVity is doing that!
For more on the announcements, check out the official blog post on their website: