on 11-16-201602:16 PM - last edited on 02-17-201712:14 PM by aluciani
This post was authored by Rohit Goyal, Product Marketing Manager, at Nutanix. Revised on 2/17 to reflect latest functionality.
Continuing the 10 things you need to know blog series, the next step is Acropolis Block Services. We announced Acropolis Block Services (ABS) with the 4.7 release (June 2016). Since then we have continued to enhance its functionality with additional performance improvements and increased qualified client support based on customer demand. Let’s uncover the “ten things you need to know” about ABS:
ABS makes it possible to implement shared infrastructure that addresses both hyperconvergence and block-level iSCSI storage for physical servers in a single solution without compromising high availability, scalability, or performance.
ABS enables IT to fully utilize existing server infrastructure even with workloads that remain on bare-metal servers due to licensing constraints, legacy application portability, or existing investments.
Utilize ABS for efficient backup and recovery techniques with simplified cloning of production databases through simplified Prism operations.
Nutanix continually seeks to add additional support for a range of operating systems and hypervisors:
Application performance seamlessly scales with the size of the Nutanix cluster, eliminating the need for fork-lift upgrades. Adding additional nodes increases performance and capacity together and doesn’t require any client-side reconfiguration.
Keep operations intact while you add capacity. Online LUN resizing now available in AOS 5.0, will make it easy to increase LUN sizes as needed while minimizing disruption.
Dynamic load balancing, also available in AOS 5.0, will avoid performance bottlenecks by automatically rebalancing traffic across the cluster.
AOS 5.0 also introduces high security with CHAP authentication and IP/IQN based whitelisting designed to ensure that only authorized clients can access specific iSCSI LUNs.
High availability applications such as Microsoft Windows Server Failover Clustering can easily fail-over/fail-back iSCSI LUNs in a matter of seconds.
ABS can be used to export Nutanix storage for use with applications like Oracle databases including Oracle RAC, Microsoft SQL Server, and IBM DB2 running outside of the Nutanix cluster on bare-metal or virtualized servers. This makes it easy to take advantage of the Nutanix web-scale architecture and migrate these apps to hyperconvergence and enterprise clouds at your own pace.
This blog includes express and implied forward-looking statements concerning product features and technology that are under development or in process, capabilities of such product features and technology, and our plans to introduce product features, including support for certain third-party solutions, in a future release. These forward-looking statements are not historical facts, and instead are based on our current expectations, estimates, opinions and beliefs. The accuracy of such forward-looking statements depends upon future events, and involves risks, uncertainties and other factors beyond our control that may cause these statements to be inaccurate and cause our actual results, performance or achievements to differ materially and adversely from those anticipated or implied by such statements, including, among others: failure to develop, or unexpected difficulties or delays in developing, new product features or technology on a timely or cost-effective basis; the introduction, or acceleration of adoption of, competing solutions; a shift in industry or competitive dynamics or customer demand; and other risks detailed in our registration statement on Form S-1, as amended, filed with the Securities and Exchange Commission. These forward-looking statements speak only as of the date of this blog and, except as required by law, we assume no obligation to update forward-looking statements to reflect actual results or subsequent events or circumstances. Any future product or roadmap information is intended to outline general product directions, and is not a commitment, promise or legal obligation for Nutanix to deliver any material, code, or functionality. This information should not be used when making a purchasing decision. Further, note that Nutanix has made no determination as to if separate fees will be charged for any future product enhancements or functionality which may ultimately be made available. Nutanix may, in its own discretion, choose to charge separate fees for the delivery of any product enhancements or functionality which are ultimately made available.
byaluciani07-08-201605:22 AM - edited 07-08-201601:24 PM
Our introductory post on ABS covered the design of ABS and the use cases for which it was designed. In this follow-on article, we will walk ABS through the paces of a real world deployment and day-two operations.
To make things interesting, we’ll assume a Nutanix cluster serving virtual machines (web & application tier) with spare capacity and performance resources. The administrator has chosen to serve these resources out to an external bare metal server running Oracle.
Let’s start with the initial deployment using ABS.
1. Provision ABS Storage on Nutanix and Whitelist Oracle
In the storage tab of the Prism management interface, provision the storage in a couple of clicks by creating a volume group (a grouping of LUNs or volumes) and whitelist each Oracle server. This enables discovery of the newly created LUNs.
2. Discover the Storage on the Oracle Server(s)
On the bare metal server(s), issue an iSCSI discovery to make the storage available to Oracle. Nutanix ABS LUNs, no matter how many have been created, are discovered over a single virtual IP called the Data Services IP. Oracle then layers ASM on the LUNs, readying them for I/O.
3. Generate Load on the Oracle Instance
Because of the way that ABS is designed, the single virtual IP address for ABS keeps I/O spread uniformly across all nodes in the cluster. A single node is never overwhelmed. Indeed, most monster Oracle databases will run comfortably on an entry-level, three-node Nutanix cluster.
The advantages of ABS persist long after the initial deployment. Now let’s consider some of the most common storage management scenarios that benefit from the Nutanix architecture with ABS.
Avoiding forklift upgrades by making it easy to scale capacity and performance as needed is one of the greatest benefits that comes from deploying a Nutanix infrastructure. Instead of having to pre-purchase fibre-channel SAN capacity in anticipation of future growth (and then frequently guessing incorrectly), Nutanix makes it easy to “pay as you grow” by scaling the number of nodes on an as-needed basis. This advantage remains true for ABS deployments.
Prism uses machine learning techniques to help administrators predict how much runway is left in term of cluster capacity and performance based on how fast Oracle is growing. This runway can be displayed as part of Prism’s native 1-click planning capability.
As the runway shortens, it is then a simple task to use this facility to optimize the environment by either following the product’s recommendations or by adding nodes to plan for events anticipated by storage administrators.
Scaling the Cluster
Simply plug in one or more new nodes. After they are automatically discovered, it is a simple operation to allocate them to the existing cluster. The new resources will contribute as equal citizens serving I/O for a subset of LUNs, ensuring a well-balanced cluster. This means Oracle will be able to take full advantage of the newly added capacity and performance resources.
Provisioning an Oracle “Test” Virtual Machine
Quickly Clone the volume group hosting the LUNs and then attach them to one or more virtual machines. Clones are fast and take up only minimal additional capacity (deltas). Attaching the clones to virtual machines is a one-click operation.
Handling Node Failures
ABS intelligently handles failures by failing over LUNs on the failed node to surviving nodes. Failover occurs in the order of seconds so applications see a minor blip but no disruption. Oracle will lose no more than 1/nth of performance and capacity resources (n nodes in a cluster) on this failure. Contrast this with traditional dual-head storage architectures where approximately 50% of performance resources are lost when a storage head fails.
Upgrading the Cluster
Upgrade the Nutanix cluster through the 1-Click Upgrade dialog in Prism. Upgrades are non-disruptive and occur one node at a time in a rolling fashion. While the upgrade is proceeding, I/O is intelligently managed across nodes, ensuring applications receive optimal performance during the upgrade.
In our initial post, we covered the design and promise of ABS—especially how ABS enables Nutanix to act as a bridge to hyperconvergence by consolidating virtualized environments and bare-metal storage onto a single Nutanix fabric. In this second post, we examined how to deploy ABS in three easy steps and led you through several of the most common day-two advantages of ABS.
These can result in significant opex and capex savings that will be of interest to many IT organizations. If you’re curious to hear more about ABS performance, please check out our final blog in this series which presents proof of the high I/O performance possible with ABS in hyperconverged environments.
Acropolis Block Services (ABS) is available today as a generally available product. We hope you’ll check it out and let us know what you think!
This post is authored by Timothy Isaacs and Geoffrey Noer, Product Management Nutanix
byaluciani07-07-201612:10 AM - edited 07-08-201605:23 AM
In a perfect world, everything would run in virtual machines and leverage hyperconverged infrastructure to achieve the lowest possible capex and opex. But even as the Enterprise drive towards hyperconvergence continues to accelerate, a number of adjacent use cases stubbornly demand their own storage silos, often delaying the adoption of the hyperconverged infrastructure model in the process.
One such need is for general-purpose file services—especially home directories, user profiles, application logs and department shares. To address this, Nutanix recently released its Acropolis File Services (AFS) offering. Currently available as a technology preview, AFS enables these use cases to be consolidated onto Nutanix for the first time without the need for special-purpose filers.
Just as important is the need for customers to continue to run a subset of production applications bare metal. Addressing bare metal applications like Oracle RAC and Microsoft SQL Server is the driving force behind the newly released Acropolis Block Services (ABS). With ABS, Nutanix customers can easily plug their tier 0/1 applications into the same Nutanix fabric already running their virtualized workloads that are able to take full advantage of hyperconvergence. Think of ABS as a service that bridges the world of bare metal with the world of virtualization and in the process consolidates infrastructure across both.
Nutanix recommends that customers virtualize everything. However, some practical limitations, predominantly non-technical, force a subset of applications to run bare metal. For instance, database licensing schemes can make virtualizing databases expensive. Another example is High Performance Computing (HPC) where one application can often take over an entire server cluster at very high utilization rates, diminishing the value of server virtualization.
Consolidate Bare Metal Applications onto the Nutanix Fabric
When apps are deployed using ABS, customers are taking advantage of the benefits of hyperconvergence and virtualization while also collapsing what would have been a bare metal silo onto the Nutanix fabric. Such consolidation increases efficiency through higher utilization levels, directly lowering capex and opex as a result. Practically, this means buying less infrastructure less often. Just as importantly, management costs can be decreased because of the single point of management that Nutanix offers, coupled with advanced tooling, automation and orchestration capabilities. In addition to spending less on process and third party management tools, simplified management also means being able to redeploy valuable and overburdened IT staff to other business-critical needs.
ABS Use Cases
ABS for Three-tier Applications
A popular deployment mode involves consolidating a three-tier application onto a single Nutanix cluster. This allows the production database tier to run on bare metal servers while both the web tier and the application tier operate in a fully virtualized infrastructure.
ABS for Database Development and Test
Database dev & test environments usually run in virtualized environments. Here the purpose of ABS is to use Nutanix native clones to clone a production bare metal database already using ABS for its storage. The clones are then attached to virtual machines for the purposes of dev & test.
ABS for Infrastructure Consolidation along the Application Development Lifecycle
Consolidation also allows customers to optimize costs across the application development lifecycle. This is accomplished by standardizing on the same product and process starting with development and test (QA) leading all the way to staging and production. Go virtual or go bare metal? Here, the answer is simply “yes.” Leverage virtualization for development and test databases, then run the staging and production databases bare-metal using ABS.
ABS for Server Investment Protection
When server and storage amortization cycles are misaligned (storage is being retired but servers are not), ABS can be used as a bridge to a fully hyperconverged environment. Start by replacing the storage being retired by plugging your existing virtualized or bare metal server investment into the Nutanix cluster with ABS. When the servers are also ready for retirement, it is then a simple matter to move the applications onto Nutanix running as hyperconverged virtual machines. And since the application’s data remains unchanged on the cluster, moving them is an easy, copy-free operation.
ABS for Legacy Applications
Most legacy applications live out their lives undisturbed with no plans for virtualization until a major rewrite. Like the last use case, ABS can function as a bridge, serving these bare metal legacy applications until that rewrite takes place. At that point they can be run on Nutanix the traditional way as hyperconverged virtual machines.
In order for ABS to succeed in a bare metal server context, Nutanix engineers focused on high performance, high availability, and flexibility. The result is a simple and broadly applicable storage solution fully integrated with the current Nutanix feature set and capability.
All nodes in the Nutanix cluster work in parallel to deliver aggregate, scale-out performance. If a single node pushes 100,000 random I/O, then three nodes will push 300,000. This performance is aggregated and can be delivered to a single monster application running bare metal.
Highly Available & Durable
In the unlikely event of a node failure, iSCSI failover ensures that the ABS services that had been running on the failed node will quickly be taken over by a surviving node. This failover process occurs in the order of seconds, within the thresholds for target applications. Data durability is inherited from the platform by maintaining two or three data copies along with several low-level data integrity checks—a norm on the Nutanix platform. Five 9’s of availability or higher can be achieved as a result.
Fully Leverages Nutanix Feature Set
ABS deployments benefit from the same comprehensive Nutanix feature set that customers already love in virtualized environments. Snapshots of ABS block data can be taken frequently, with multiple copies retained, replicated, recovered on DR, backed up to the cloud, deduped, compressed, and erasure coded. Best of all, it can easily all be managed from Prism, providing a single point of management for the Nutanix solution.
ABS storage is served over iSCSI, a popular and broadly accepted block storage protocol, which does away with the need for fiber channel networks and the complexity that accompanies them. With Nutanix, block storage can be provisioned in a few clicks and intelligent defaults mean that changes are usually minimal. Day-two operations like scaling out, updating software and servicing hardware are also accomplished in a few clicks and not different from operations on the Nutanix platform today. And unlike fibre channel SANs, no client-side changes are required to scale out—simply grow the size of the Nutanix cluster to scale the application I/O performance.
The strong value of hyperconvergence can be put into question when one or two production applications benefit from continuing to run as bare metal deployments. ABS alleviates this concern, making it easy to serve those bare metal applications with ABS while simultaneously attaining the advantages of hyperconvergence for the majority of the applications in a fully virtualized environment.
ABS addresses multiple use cases including production bare metal databases and database dev & test. It can also smooth the entire application development lifecycle by consolidating the storage on Nutanix and using a combination of virtualized and bare metal access to the data. Finally, legacy server and bare metal application deployment investments can be protected until those servers are retired and legacy applications rewritten. The simplicity and ease of use of ABS is shown in real-world deployments and day-two operations, which we will detail in a later blog.
We hope you enjoyed this post. Acropolis Block Services (ABS) is available today as a generally available product. Take it through the paces and let us know what you think!
This post is authored by Timothy Isaacs and Geoffrey Noer, Product Management Nutanix