Nutanix Complete Cluster Date: May 2012 Authors: Ginny Roth, Tony Palmer, and Emil Flock

Transcription

1 ESG Lab Review Nutanix Complete Cluster Date: May 2012 Authors: Ginny Roth, Tony Palmer, and Emil Flock Abstract: This ESG Lab review documents hands-on testing of the Nutanix Complete Cluster, highlighting its ease of use, scalability, and support for enterprise-class data management in virtual environments. Server Virtualization Challenges Server virtualization is driving widespread and fundamental change as IT organizations struggle to meet exponentially increasing demand for network and storage resources in support of virtualized deployments. When asked for their preferred infrastructure mode, the preponderance of respondents (36%) 1 preferred a fully-integrated solution that addresses the challenges associated with hardware, tools, and processes that were not designed for virtual technology, including mapping storage and network resources to a virtual infrastructure. Figure 1. Preferred Infrastructures to Support Virtualization Source: Enterprise Strategy Group, Clearly, networking and storage needs, as well as integration, are creating headaches for IT organizations as they look to provide cost-effective solutions for virtual environments. Sizing bandwidth to support a virtual server environment is often an issue, with 24% of networking professionals citing it as a primary concern with their organization s virtual 1 Source: ESG Research Brief, Virtualized Computing Infrastructure Preferences, April The goal of ESG Lab reports is to educate IT professionals about emerging technologies and products in the storage, data management and information security industries. ESG Lab reports are not meant to replace the evaluation process that should be conducted before making purchasing decisions, but rather to provide insight into these emerging technologies. Our objective is to go over some of the more valuable feature/functions of products, show how they can be used to solve real customer problems and identify any areas needing improvement. ESG Lab's expert third-party perspective is based on our own hands-on testing as well as on interviews with customers who use these products in production environments. This ESG Lab report was sponsored by Nutanix.

2 ESG Lab Review: Nutanix Complete Cluster 2 usage. Additionally, storage administrators viewed capital costs of new storage infrastructure (36%) and scalability problems (25%) as top concerns when supporting virtualization. 2 In response to these concerns, integrated computing platforms have garnered attention. These solutions promise to alleviate networking and storage challenges associated with virtualization by combining services into one platform, eliminating the loosely coupled integration designs that exist today. The Solution: Nutanix Complete Cluster Nutanix introduced a new integrated, scale-out computing platform that delivers server, hypervisor, and storage components in a single system called the Nutanix Complete Cluster a seamless system that enables incremental scaleout from a single block to as many as needed, all managed as a single, unified system. Four server nodes are included in each 2U block, and are built to host and store virtual machines with a standard hypervisor running on each node. A Nutanix Controller Virtual Machine on each host manages storage for virtual machines on the host. Controller VMs work together to manage storage across the cluster as a pool using the Nutanix Distributed File System (NDFS). Standard VMware features like vmotion are supported by Nutanix Complete Cluster. Each server in the cluster contains a mix of SSD (PCIe and SATA) and SATA hard drives that can be pooled as a single storage resource. Virtual disks (vdisks) can be created from available storage and presented as a global iscsi target to any virtual machine without the need for a SAN. Nutanix also supports NFS to provide a unified pool of storage for virtual machines. Figure 2. Nutanix Complete Cluster 2 Source: ESG Research Report, The Evolution of Server Virtualization, November 2010.

3 ESG Lab Review: Nutanix Complete Cluster 3 Nutanix delivers the following features in a simple to deploy, integrated appliance: Command Center, a web-based management console that provides enterprise class features for managing both compute and storage in a simple to use interface. Distributed data replication that provides high availability for path and node failover. Heat-Optimized Tiering with configurable metrics to age and migrate data to lower storage tiers for optimal disk I/O performance. Support for snapshots to create point-in-time copies of data for quick recovery. Clone technology to provide rapid provisioning of virtual machines for multiple use cases such as virtual desktop deployment and production server copies for development environments. Support for VMware enterprise features such as vmotion, Dynamic Resource Scheduler (DRS), High Availability (HA), and server migration. ESG Lab validated the Complete Cluster solution s ease of use and its support for virtual environments during two days of hands-on testing at Nutanix headquarters in San Jose, California in June of 2011, and one day of follow-up testing in April Provisioning and VMware Integration Nutanix Complete Cluster solution is built from Nutanix Complete Blocks, which are Intel-based appliances that contain four nodes. Each of these nodes contains a combination of PCIe-SSD (Fusion-io), SATA SSD, and hard disk drives. The appliance tested by ESG Lab contained five SATA hard drives and one PCIe SSD drive for each node. Initially, the appliance configuration contained three nodes in the cluster, with a fourth on standby to add later to test scalability. As shown in Figure 3, testing was conducted in an environment that consisted of one Nutanix appliance and a laptop acting as an administrative console to connect to the appliance for management tasks. Figure 3. ESG Lab Test Bed ESG Lab Tested ESG Lab used a web browser to connect to Command Center and examined the features available for the cluster. As shown in Figure 4, the dashboard presents a comprehensive picture of the Complete Cluster: hosts are listed under both the storage and compute sections. With this configuration, ESG Lab was able to examine both the storage and virtual server resources associated with a host. In addition, multiple management tasks are available, including creating and editing storage pools and containers for logical storage allocation.

4 ESG Lab Review: Nutanix Complete Cluster 4 Figure 4. Nutanix Command Center ESG Lab tested the ability to create an NFS datastore for use by a new VMware virtual machine. Nutanix developed a distinct NFS implementation, Nutanix Distributed File System (NDFS), which uses the Nutanix storage controllers, one per node, to manage access to shared storage across the cluster. ESG Lab began the test by creating a new container from an existing storage pool. Figure 5 shows a dialog box to create a storage container, which will become a new NFS mount point for the four nodes in the cluster. ESG Lab created a 4TB container from storage pool sp1 with a replication factor of 2. The replication factor provides two copies of data for high availability. Figure 5. Create Container for NFS Datastore

5 ESG Lab Review: Nutanix Complete Cluster 5 Nutanix reduces the complexity of provisioning and registration of new datastores into vsphere down to a few simple steps. After accepting the container parameters, ESG Lab was presented a dialog box with the option of creating an NFS datastore for the container. ESG Lab accepted the option and, as Figure 6 shows, created an NFS datastore named NTNX-NFS. Note that the datastore is accessible to all four hosts in this block. ESG Lab used the vsphere client to verify the datastore was created and available for the ESX server. Figure 6. NFS Datastore Created and Mounted Next, ESG Lab tested the clone functionality of Nutanix Complete Cluster using VMware View. One Windows 7 virtual image configured as a 32-bit desktop with 1GB RAM and a 25GB hard drive was used to test cloning functionality. As Figure 7 shows, ESG Lab used two different methods to create 100 clones of a Windows 7 virtual machine. The View Composer linked clones option used the native VMware software to create replicas of the parent virtual machine that can share the same datastore. A linked clone is made from a snapshot of the parent. All files available on the parent at the moment of the snapshot continue to remain available to the linked clone. Ongoing changes to the virtual disk of the parent do not affect the linked clone, and changes to the disk of the linked clone do not affect the parent. The Full virtual machines option used VMware s API for Array Integration (VAAI) plug-in for vsphere to call Nutanix s native array-based snapshotting known as Nutanix Quickclones.

6 ESG Lab Review: Nutanix Complete Cluster 6 Figure 7. Create Virtual Machine Clones Figure 8 shows the cloning in progress for linked clones on the ESXi server. ESG Lab confirmed that the entire copy process for 100 clones lasted approximately 30 minutes using VMware cloning and 35 minutes using Nutanix snapshots. By comparison, during testing of this product in June 2011, it took 20 minutes to clone 25 virtual machines using the Nutanix snapshot function. In addition, the VAAI plugin functionality was not available, requiring CLI commands from the Nutanix virtual storage controller.

7 ESG Lab Review: Nutanix Complete Cluster 7 Figure 8. NFS Clones in Progress Figure 9 illustrates performance of the cluster during the linked cloning process, showing bandwidth speeds ranging from approximately 40 to 77 MB/s and IOPS ranging from 2200 to Figure 9. Performance Monitor During Cloning

8 ESG Lab Review: Nutanix Complete Cluster 8 In addition to these performance metrics, Nutanix included a number of reporting metrics to allow administrators to actively monitor the health of a cluster. ESG Lab examined the metrics collected and displayed through the reporting utility, which include storage capacity, tier usage, provisioned capacity, IOPS, latency, and bandwidth utilization. Figure 10 shows an alert history that reports on thresholds that are exceeded by the system. For example, Nutanix tracks disk usage when it exceeds 75% and device temperatures over 70 degrees Celsius. Figure 10. Nutanix Alert History Nutanix Colossus Testing ESG Lab audited the results of collaboration between Nutanix and VMware on a performance test (dubbed Project Colossus 3 ) that started with 300 desktops and scaled smoothly up to 3,000. They used the Reference Architecture Workload Simulator (RAWC) tool to create workloads that simulated typical end-user behavior and employed VMware View 5 linked clone technology to provision the virtual machines. The RAWC tool is designed to characterize storage, network, and compute workloads, simulating multiple tasks that users perform, such as Internet browsing, archiving files with compression tools, composing and sending , using various Microsoft Office applications, and opening PDF files with Adobe Reader. A system designated as the controller builds the test configurations and stores the shares necessary for execution. View Client sessions are established, and scripts are executed on each desktop to simulate active desktop users. Three different types of virtual desktop users task, knowledge, and power were simulated. For example, a knowledge worker was assumed to average between 9-12 IOPS, with a 20 IOPS peak during boot and logon. The results showed less than three-tenths of a second difference between the slowest and fastest application times as the environment scaled from 300 to 3,000 virtual desktops, proving the ability to grow capacity rapidly without adversely affecting performance. 3 Source: Project Colossus White Paper, November 2011.

9 ESG Lab Review: Nutanix Complete Cluster 9 Why This Matters Integrated platforms offer the potential to greatly decrease the cost of deploying virtual technology. A scale-out design allows customers to pay as they go not only for server resources, but also for networking and storage equipment. Additionally, IT organizations can begin to realize true OPEX savings by simplifying and centralizing the management of disparate resources such as storage, networking, and servers. Nutanix has addressed all these concerns with an integrated compute platform that is easy to manage and that scales rapidly to meet the heavy demands of virtual environments. Storage and networking functions require no advanced training to manage, and the virtual machines immediately recognize storage for rapid provisioning. ESG Lab was able to verify that provisioning storage, whether as an iscsi target or NFS datastore, was simple and provided a unified pool of storage for multiple virtual machines. In addition, the cloning process was simple to deploy and completed quickly for 100 virtual machines. ESG Lab took a special interest in the Nutanix/VMware collaboration on Project Colossus. Scaling from 300 to 3,000 desktops with variations of only three-tenths of a second in application performance proves that Nutanix s scale-out capability is flexible and robust.

10 ESG Lab Review: Nutanix Complete Cluster 10 Management and High Availability Migrating virtual machines from one host to another is a required function in VMware environments, functionality that Nutanix Complete Cluster supports natively. During testing performed in June 2011, ESG Lab examined migration of a Win7-Desktop-0 virtual image from node to ; as Figure 11 shows, the migration was successful and took less than a minute to complete. ESG Lab verified the migration with Command Center and observed that Win7- Desktop-0 was now listed as a compute resource on Figure 11. Migrate Virtual Machines Another supported function for VMware is the Dynamic Resource Scheduler (DRS). DRS automates live migrations through vmotion when resources become taxed. Nutanix Complete Cluster supports DRS by supporting VMware vmotion. ESG Lab tested DRS by using a virtual image named CPU-Test. When powered up, CPU-Test started a program that created 100% load on host Figure 12 shows the performance tab on the host, illustrating the high workload. As the task window shows, Win7-Desktop1 began to migrate as soon as CPU utilization was 100%.

11 ESG Lab Review: Nutanix Complete Cluster 11 Figure 12. Automatic Migration Using Dynamic Resource Scheduler Nutanix Complete Cluster is designed to respond to path and node failures. First, ESG Lab examined path failover, as shown in Figure 13. Path failure occurs when the data path to a particular node becomes unavailable. ESG Lab tested the effectiveness of path failover by disabling one of the four controller virtual machines associated with the four nodes. Concurrently, ESG Lab started an application on a desktop running on the associated node. The remaining controller virtual machines took over the lost data path and caused no disruption to the application. Figure 13. Path Failover

12 ESG Lab Review: Nutanix Complete Cluster 12 ESG Lab tested node failover by powering down node in the appliance. The node was configured with five virtual images, which began to immediately migrate to other nodes in the cluster as shown in Figure 14. ESG Lab verified the images migrated successfully by opening the consoles in vcenter for the affected machines. Figure 14. Node Failover Figure 15 shows the migration of VMs subsequent to a node failure. ESG Lab verified the images migrated successfully by opening the consoles in vcenter for the affected machines. Figure 15. Server Migration After Failed Node

13 ESG Lab Review: Nutanix Complete Cluster 13 Next, ESG Lab analyzed the snapshot capabilities of Nutanix Complete Cluster. Snapshots preserve data at any point in time for virtual machines in the cluster. In vcenter, ESG Lab opened the console to the Win7-GoldImage virtual machine and created a text file on the desktop. Using the CLI on the controller virtual machine, ESG Lab created a snapshot of the virtual machine called Win7-snap1. As shown in Figure 16, using the Command Center console, ESG Lab verified the successful completion of the snapshot. Next, ESG Lab deleted the text file on the Win7-GoldImage desktop. Returning to the CLI, ESG Lab restored the snapshot to a new clone name, which created a new iscsi target. ESG Lab then opened the vcenter console, removed the current iscsi target for Win7-GoldImage, and added the new iscsi target created by the snapshot restore. Using the VMware console for the restored image, ESG Lab verified that the restored snapshot contained the text file. Figure 16. Nutanix Complete Cluster Snapshots Finally, ESG Lab examined growing the cluster to accommodate additional ESX hosts. To accomplish this, ESG Lab first added a host in vcenter in the vsphere client and then logged into the Nutanix Controller VM to issue the necessary cluster grow commands. Using the Nutanix command line interface (NCLI) on the controller virtual machine, ESG Lab used the command host add svm-ip= to add the host to the cluster. ESG Lab then used Command Center to verify creation of the new host and added its SSD and SATA drives to the existing storage pool. Why This Matters The promise of integrated computing can only be realized if converged solutions support enterprise-class data management functions. Customers need to be able to recover quickly from data, node, and path failures that can have a severe impact on company revenue. In addition, the ability to work seamlessly with virtual management software helps reduce the costs associated with maintaining disparate management tools. Nutanix Complete Cluster enables combines multiple enterprise functions such as cloning, snapshots, cluster management, HA failover, and virtual machine migration to provide the protection and fault tolerance required by today s demanding businesses. ESG Lab found the enterprise-class features effective in delivering a high performance and highly available virtual environment.

14 The Bigger Truth ESG Lab Review: Nutanix Complete Cluster 14 The adoption of server virtualization requires significant changes to existing storage infrastructure. Storage groups are playing catch up as they design new networked storage solutions to meet the demands that the high I/O and throughput densities of virtualization creates. While virtualization promises to deliver savings in capital expenditures, these savings can be wiped out by the costs of upgrading to faster storage systems needed to respond to these requirements. According to ESG research, simplified management, reduced deployment times, and a better economic model have all been identified as benefits of integrated platforms as users enjoy ease of management (44%), faster deployments (37%), improved TCO (35%), and less time required for accelerated hardware and software integration (33%). Respondents also envision benefits associated with interoperability, application performance, and service and support. 4 Nutanix delivers a solution for virtual environments that simplifies the deployment of scale-out compute and storage resources while providing enterprise-class data management functions. Customer looking to build a virtual desktop environment rapidly and easily will find the solution offered by Nutanix to be a perfect fit. ESG Lab was able to confirm effective implementation of cluster services, snapshots, and cloning in addition to support for multiple VMware functions such as vmotion, Dynamic Resource Scheduler, and HA failover. With new support for the NFS protocol, path failover, and VAAI plugins, Nutanix has extended its interoperability for a variety of customer environments. ESG was not able to test dynamic node addition and removal for its cluster service but verified that Nutanix plans to make the capability available within the coming month. ESG recommends that Nutanix include cluster grow and shrink services to its GUI Management Console to reduce reliance on command line interface operations. In addition, the added capabilities of alerting and configurable parameters would be a welcome improvement to the reporting system. Virtualization will continue to be a disruptive force for IT, requiring new ways of thinking about compute and storage environments as scalability and elasticity become paramount. Nutanix is at the forefront of this emerging trend with a solution that is both simple and versatile to deploy and manage with the ability to scale easily for demanding virtual desktop environments. Nutanix s innovative integration of both storage and compute resources liberates organizations to elastically grow and adapt infrastructure resources to meet rapidly-changing business objectives. All trademark names are property of their respective companies. Information contained in this publication has been obtained by sources The Enterprise Strategy Group (ESG) considers to be reliable but is not warranted by ESG. This publication may contain opinions of ESG, which are subject to change from time to time. This publication is copyrighted by The Enterprise Strategy Group, Inc. Any reproduction or redistribution of this publication, in whole or in part, whether in hard-copy format, electronically, or otherwise to persons not authorized to receive it, without the express consent of the Enterprise Strategy Group, Inc., is in violation of U.S. copyright law and will be subject to an action for civil damages and, if applicable, criminal prosecution. Should you have any questions, please contact ESG Client Relations at (508) Source: ESG Research Brief, Integrated Computing Trends, March 2011.