NetApp SolidFire Reference Architecture with Veeam Backup & Replication 9.5

Transcription

1 Technical Report NetApp SolidFire Reference Architecture with Erik Kemp, NetApp; Stefan Renner, Veeam (EMEA); and Shawn Lieu, Veeam (Americas) September 2017 TR-4634 In partnership with Abstract This document outlines the reference architecture and best practices for using NetApp SolidFire storage in a environment.

2 TABLE OF CONTENTS 1 Executive Summary The Challenge The Solution Introduction About NetApp About Veeam Reference Architecture Overview Use Cases Deployment Scenarios Veeam Cloud Connect Backup General Overview System Requirements NetApp SolidFire Arrays SolidFire SF-Series Hardware Key SolidFire Features Enterprise Reliability and Availability Benefits of SolidFire Scale-Out Storage Element OS Software Overview Solution Architecture Requirements VMware System Requirements Hyper-V System Requirements Veeam Sizing Requirements Design Considerations Direct SAN Access Mode Data Restore in Direct SAN Access Mode Off-Host Backup (Hyper-V) Components SolidFire Volume Configuration Guidelines SolidFire Quality of Service NetApp SolidFire Reference Architecture with

3 9.2 SolidFire Storage Configuration Guidelines for Backup & Replication 9.5 Backup Repositories SolidFire Volumes as VMware Datastores Directly Connected SolidFire Volume (File System Backups) SolidFire Host Configuration Guidelines Host Connectivity SolidFire Plug-In for VMware vcenter SolidFire VSS Hardware Provider SolidFire VSS Provider Introduction SolidFire VSS Provider Overview SolidFire VSS Provider Requirements Active IQ Summary Where to Find Additional Information NetApp SolidFire Documentation Software Documentation Appendix A: QoS Tuning Scripts libsf.ps pre-backup.ps post-backup.ps LIST OF TABLES Table 1) System requirements for Cloud Gateway Table 2) System requirements for the Veeam backup server Table 3) System requirements for the WAN Accelerator Table 4) Veeam backup server specifications (VMware) Table 5) Veeam Backup & Replication console (VMware) Table 6) Backup proxy server specifications (VMware) Table 7) Backup repository server specifications (VMware) Table 8) Veeam backup server specifications (Hyper-V) Table 9) Veeam Backup & Replication console (Hyper-V) Table 10) Off-host backup proxy server specifications (Hyper-V) Table 11) Backup repository server specifications (Hyper-V) Table 12) Veeam sizing requirements Table 13) components Table 14) SolidFire volume QoS for backup repositories NetApp SolidFire Reference Architecture with

4 Table 15) SolidFire volume QoS for VMware datastores Table 16) SolidFire volume QoS for directly connected SolidFire volumes Table 17) Hardware prerequisites Table 18) Software prerequisites Table 19) Supported configurations LIST OF FIGURES Figure 1) SolidFire as Veeam Backup & Replication repositories for backup and archiving....7 Figure 2) SolidFire as a Veeam Backup & Replication backup repository and an off-site cloud repository....8 Figure 3) SolidFire production storage with Veeam Backup & Replication to E-Series and EF-Series backup repositories....8 Figure 4) SolidFire production storage with Veeam Backup & Replication with an AltaVault backup repository to a public cloud and StorageGRID....9 Figure 5) SolidFire production storage with Veeam Backup & Replication and an AltaVault backup repository to a public and a private cloud....9 Figure 6) SolidFire as a Veeam Backup & Replication backup repository and off-premises backup repository for FAS production storage Figure 7) Simple deployment Figure 8) Advanced deployment Figure 9) Distributed deployment Figure 10) Veeam Cloud Connect backup Figure 11) Virtual deployment Figure 12) Physical deployment Figure 13) Element OS UI dashboard Figure 14) Data restore in direct SAN access mode Figure 15) Off-host backup process NetApp SolidFire Reference Architecture with

5 1 Executive Summary 1.1 The Challenge The enterprise IT landscape is going through a transformation as virtualization has become the norm. The rise of the cloud has set a new standard for infrastructure agility, leaving enterprises racing to catch up. Across the industry, a sea of new technologies is pointing the way to the next-generation data center (NGDC). Customers are moving away from hand-crafted, single-purpose data center designs and are embracing new technologies that deliver agility, scalability, and greater efficiency. The NGDC relies on the availability and resiliency of data. Enterprises require a complete data protection solution that is reliable, flexible, and easy to use. Virtualizing an environment provides an increased level of data availability, but meeting aggressive recovery point objectives (RPOs) and recovery time objectives (RTOs) becomes increasingly difficult. Traditional backup tools were not created for virtualized environments. This fact makes it hard for many organizations to take full advantage of virtualization, and many IT managers struggle with the following issues: Unreliable backups Recovery that takes too long High costs associated with managing backup data and secondary storage An inability to provide reliable, true backups for compliance purposes Lost productivity because of managerial complexity The need to scale backup operations for growth 1.2 The Solution To meet these challenges, Veeam and NetApp SolidFire have collaborated to offer high-performance storage with reliable data protection that is designed for virtualized environments. Veeam and SolidFire help you modernize your data protection strategy with a solution that is architected from the ground up to manage large amounts of data. The solution can also handle the increasing performance and availability demands of next-generation data centers. Veeam Backup & Replication unifies backup and replication into a single solution, increasing the value of backup and reinventing data protection for VMware vsphere and Microsoft Hyper-V virtual environments. The Veeam agentless design provides multiple backup options to meet your needs. Features such as source-side deduplication and compression, changed block tracking, parallel processing, and automatic load balancing provide fast and efficient backups. Together, Veeam and NetApp create an optimal staging area for backups, thereby reducing backup ingest bottlenecks and providing faster backups through parallel processing. Veeam Backup & Replication provides the following advantages: Granular recovery of virtual machines (VMs) and files, including Microsoft Exchange and SharePoint application items The ability to automatically verify every backup, VM, and replica Self-service recovery of VMs and guest files without a direct network connection to the VM, user permissions, or the need to deploy costly agents Instant VM recovery in as little as two minutes A choice to back up and recover what you need, where you need it, and when you need it, whether it is on site, on tape, or in the cloud Veeam and NetApp offer the right solution for performance, flexibility, and reliability, providing an impressive, modern disaster recovery solution for your vsphere or Hyper-V environment. 5 NetApp SolidFire Reference Architecture with

6 This document is a reference architecture for creating a collaborative backup and recovery solution on NetApp SolidFire with data protection software. 2 Introduction Veeam and NetApp jointly developed this reference architecture to guide successful Backup & Replication 9.5 deployments with NetApp SolidFire storage and to enable data and application availability. This solution is optimized for virtual environments, providing all-flash backup and recovery on scalable, flexible, performance-oriented SolidFire storage arrays. This solution provides you with superior data management for virtual environments and high availability while also making your data highly available. SolidFire arrays provide a high-performing backup repository to house Veeam-created backups. With this capability, the recovery technologies that are enabled through Veeam can satisfy stringent RTOs. Recovery technologies such as Instant VM Recovery, SureBackup, and On-Demand Sandbox can achieve their full potential by using backup repositories that are capable of high I/O. These technologies allow you to restore from your backups faster and enable capabilities such as automated recovery verification. These technologies can also use backup data as a testing environment. This capability has changed the way that users have used backups in the past because there are more benefits associated with having backups. Backups no longer sit idle, waiting for an emergency restore; instead, you can use your backups in many creative ways. This solution includes the following capabilities: Recovery of a failed VM in as little as two minutes Near-continuous data protection with built-in replication Fast, agentless item recovery and e-discovery for Microsoft Exchange, SharePoint, and Active Directory, along with transaction-level recovery of SQL Server databases Automatic recoverability testing of every backup and every replica Off-site backups made up to 50 times faster than the speed of standard file copy with built-in WAN acceleration Fast and secure cloud backups with Veeam Cloud Connect Deduplication and compression to minimize storage consumption Off-site recovery with one-click site failover and support for facilitated data center migrations with zero data loss 2.1 About NetApp NetApp is the data authority for hybrid cloud. We provide a full range of hybrid cloud data services that simplify management of applications and data across cloud and on-premises environments to accelerate digital transformation. Together with our partners, we empower global organizations to unleash the full potential of their data to expand customer touchpoints, foster greater innovation and optimize their operations. 2.2 About Veeam Veeam recognizes the new challenges that companies across the globe face in enabling the always-on business, one that must operate 24 hours a day, 7 days a week, and 365 days a year. To address this challenge, Veeam delivers availability for the modern data center by making sure that RTOs and RPOs are less than 15 minutes for all applications and data. Veeam corporate headquarters are located in Baar, Switzerland, and the main Americas office is located in Columbus, Ohio. 6 NetApp SolidFire Reference Architecture with

7 3 Reference Architecture Overview This section describes reference architectures that range from small environments that protect a few terabytes of data to large enterprise-size environments with petabytes of data under management. 3.1 Use Cases SolidFire as Veeam Backup & Replication Repositories for Backup and Archiving Figure 1 depicts a typical NetApp SolidFire and Veeam setup. Figure 1) SolidFire as Veeam Backup & Replication repositories for backup and archiving. SolidFire as a Veeam Backup & Replication Backup Repository and an Off-Site Cloud Repository SolidFire can serve as a destination for Veeam Cloud Connect replication. This configuration enables partners to offer multitenancy cloud disaster recovery for highly virtualized environments. It also provides guaranteed CPU, RAM, disk (Veeam), and IOPS (SolidFire) for customer failover. This joint solution provides simplicity and guaranteed performance in the cloud market. Figure 2 depicts SolidFire as a Veeam Backup & Replication backup repository and an off-site cloud repository. 7 NetApp SolidFire Reference Architecture with

8 Figure 2) SolidFire as a Veeam Backup & Replication backup repository and an off-site cloud repository. SolidFire Production Storage with Veeam Backup & Replication to E-Series and EF-Series Backup Repositories Figure 3 depicts SolidFire production storage with Veeam backup and replication to NetApp E-Series and EF-Series backup repositories. Figure 3) SolidFire production storage with Veeam Backup & Replication to E-Series and EF-Series backup repositories. SolidFire Production Storage with Veeam Backup & Replication and an AltaVault Backup Repository to a Public Cloud and StorageGRID Figure 4 depicts SolidFire production storage with Veeam Backup & Replication with a NetApp AltaVault backup repository to a public cloud and NetApp StorageGRID. 8 NetApp SolidFire Reference Architecture with

9 Figure 4) SolidFire production storage with Veeam Backup & Replication with an AltaVault backup repository to a public cloud and StorageGRID. SolidFire Production Storage with Veeam Backup & Replication and AltaVault Backup Repository to a Public Cloud and a Private Cloud Figure 5 depicts SolidFire production storage with Veeam Backup & Replication and an AltaVault backup repository to a public cloud and a private cloud. Figure 5) SolidFire production storage with Veeam Backup & Replication and an AltaVault backup repository to a public and a private cloud. SolidFire as a Veeam Backup & Replication Backup Repository for FAS Production Storage Figure 6 depicts Veeam integration with a NetApp FAS-series production storage array, with newly created backups going to a SolidFire array for storage. To provide disaster recovery, backups can also be sent off the premises to another backup repository (another SolidFire system). Veeam provides a backup copy job for such scenarios. This job can be used for backups off the premises or for long-term archiving by using Veeam s built-in Grandfather-Father-Son (GFS) type retention. 9 NetApp SolidFire Reference Architecture with

10 Figure 6) SolidFire as a Veeam Backup & Replication backup repository and off-premises backup repository for FAS production storage. Leveraging Veeam s backup copy job architecture is important for achieving the last level of protection. The off-site copy provides safeguards for an entire data center level disaster. Veeam also provides an optional WAN acceleration component that can help reduce bandwidth utilization. This component can play a huge role in environments that have active-active sites or that have low available bandwidth to start with. The forever-incremental nature of the backup copy job enables only incremental change to data to be transmitted off site after the initial copy. Preseeding options are available for the initial transfer of data for environments that need it. 3.2 Deployment Scenarios This section describes the various deployment scenarios for Veeam Backup & Replication on NetApp SolidFire storage. Simple Deployment In a simple deployment scenario, one instance of Veeam Backup & Replication is installed on a physical or virtual Windows-based machine. This installation is referred to as a backup server. In a simple deployment (Figure 7), the backup server performs the following roles: It functions as a management point, coordinates all jobs, controls their scheduling, and performs other administrative activities. It acts as the default backup proxy for handling job processing and for transferring backup traffic. All services necessary for the backup proxy functionality are installed locally on the backup server. It is used as the default backup repository. During installation, Veeam Backup & Replication checks volumes of the machine on which you installed the product and identifies the volume with the largest amount of free disk space. Veeam Backup & Replication then creates a backup folder on this volume for use as the default backup repository. It is used as a mount server and guest interaction proxy. 10 NetApp SolidFire Reference Architecture with

11 Figure 7) Simple deployment. If you plan to back up and replicate only a small number of VMs or to evaluate Veeam Backup & Replication, this configuration is adequate. Veeam Backup & Replication is ready for use right out of the box. As soon as it is installed, you can start using the solution to perform backup and replication operations. To balance the load of backing up and replicating your VMs, you can schedule jobs at different times. Advanced Deployment In large-scale virtual environments with numerous jobs, the load on the backup server is heavy. In this case, Veeam and NetApp recommend that you use the advanced deployment scenario (Figure 8), which moves the backup workload to dedicated backup infrastructure components. The backup server here functions as a manager for deploying and maintaining the backup infrastructure components. An advanced deployment includes the following components: Virtual infrastructure servers. VMware vsphere hosts used as the source and the target for backup, replication, and VM copy. Backup server. A configuration and control center of the backup infrastructure. Backup proxy. A data-mover component used to retrieve VM data from the source datastore, process it, and deliver it to the target. Backup repository. A location used to store backup files, VM copies, and auxiliary replica files. Dedicated mount servers. A component required to restore VM guest OS files and application items to the original location. Dedicated guest interaction proxies. Components used to deploy the runtime process in Microsoft Windows VMs. 11 NetApp SolidFire Reference Architecture with

12 Figure 8) Advanced deployment. With the advanced deployment scenario, you can easily meet your current and future data protection requirements. You can expand your backup infrastructure horizontally in a matter of minutes to match the amount of data that you want to process and the available network throughput. Instead of increasing the number of backup servers or constantly tuning job scheduling, you can install multiple backup proxies and repositories and distribute the backup workload among them. The installation process is fully automated, which simplifies deploying and maintaining the backup infrastructure in your virtual environment. In virtual environments with several proxies, Veeam Backup & Replication dynamically distributes backup traffic among these proxies. A job can be explicitly mapped to a specific proxy. Alternatively, you can let Veeam Backup & Replication choose the most suitable proxy. In this case, Veeam Backup & Replication checks the settings of available proxies and selects the most appropriate one for the job. The proxy server to be used should have access to the source and target hosts as well as to the backup repository to which files are written. The advanced deployment scenario can be a good choice for backing up and replicating off site. You can deploy a backup proxy in the production site and another one in the disaster recovery site closer to the backup repository. When a job is performed, backup proxies on both sides establish a stable connection, so this architecture also allows efficient data transport over a slow network connection or WAN. To regulate backup load, you can specify the maximum number of concurrent tasks per proxy and set up throttling rules to limit proxy bandwidth. The maximum number of concurrent tasks can also be specified for a backup repository in addition to the value of the combined data rate for it. Another advantage of the advanced deployment scenario is that it contributes to high availability. Jobs can migrate between proxies if one of them becomes overloaded or unavailable. Distributed Deployment The distributed deployment scenario (Figure 9) is recommended for large, geographically dispersed virtual environments with multiple backup servers that are installed across different sites. These backup servers are federated under Veeam Backup Enterprise Manager, an optional component that provides centralized management and reporting for these servers through a web interface. 12 NetApp SolidFire Reference Architecture with

13 Figure 9) Distributed deployment. Veeam Backup Enterprise Manager collects data from Veeam backup servers and enables you to run backup and replication jobs across the entire backup infrastructure through a single pane of glass. You can also edit and clone jobs by using a single job as a template. Veeam Backup Enterprise Manager also provides reporting data for various areas (for example, all jobs that were performed within the past 24 hours or 7 days, all VMs engaged in these jobs, and so on). Using indexed data consolidated on one server, Veeam Backup Enterprise Manager can search for VM guest OS files in VM backups created on all backup servers even if they are stored on backup repositories on different sites. These files can then be recovered with a single click. Searching for VM guest OS files is enabled through Veeam Backup Enterprise Manager itself. To streamline the search process, you can optionally deploy a Veeam Backup Search server in your backup infrastructure. With flexible delegation options and security roles, IT administrators can delegate the necessary file restore or VM restore rights to authorized personnel in your organization. For example, you can allow database administrators to restore Oracle or SQL Server VMs. If you use Veeam Backup Enterprise Manager in your backup infrastructure, you do not need to install licenses on every Veeam backup server that you deploy. Instead, you can install one license on the Veeam Backup Enterprise Manager server, and it is then applied to all servers across your backup infrastructure. This approach simplifies tracking license usage and license updates across multiple Veeam backup servers. In addition, VMware administrators can benefit from the Veeam plug-in for the vsphere web client, which can be installed by using Veeam Backup Enterprise Manager. Administrators can analyze cumulative information on used and available storage space; view statistics on processed VMs; and review success, warnings, and failure counts for all jobs. They can easily identify unprotected VMs and perform capacity planning for repositories, all directly from vsphere. 4 Veeam Cloud Connect Backup 4.1 General Overview Veeam Cloud Connect is designed specifically so that service providers can create and serve remote backup repositories. Service providers who are part of the Veeam Cloud & Service Provider (VCSP) program can use Veeam Cloud Connect to offer customers backup as a service (BaaS) and disaster recovery as a service (DRaaS). Veeam Backup & Replication (starting with version 8) can then consume these services from the customer s service provider of choice to send backups off site or to replicate VMs (versions 9 and 9.5). 13 NetApp SolidFire Reference Architecture with

14 With Veeam Cloud Connect, service providers can build their own Veeam-powered services, leveraging a technology built from the ground up to be scalable and to serve multiple tenants. Veeam Cloud Connect removes the barriers to deploying these services by greatly simplifying the deployment and customer onboarding process. No VPN Tunnels It is not easy to configure a VPN automatically, and it usually requires interaction between the service provider and the customer. Even when a VPN is properly configured, it requires ongoing monitoring and management to guarantee it is always up and running. With Veeam Cloud Connect, every connection happens directly over the internet using a single TCP\/UDP port protected by SSL\/TLS encryption. This configuration is possible because of a new and dedicated Veeam cloud gateway component. A cloud gateway is responsible for the transfer of all backup and replication traffic over the single port connection. The connection uses the public internet and guarantees complete confidentiality of the data traversing the connection. Multitenancy The second design principle is complete support for multitenancy. Service providers create competitive services by sharing their resources among their customers. Although multitenancy reduces costs, it cannot be allowed to reduce security. Each tenant must be completely isolated from others and in total control of their slice of the environment. Secure multitenancy is possible through Veeam Cloud Connect because of two different components: a cloud repository and a cloud host. For cloud-based backup workloads, service providers expose a cloud repository to their customers. The cloud repository creates an abstraction layer over an existing backup repository. Therefore, multiple customers can store backups inside the same shared repository with the same level of confidentiality they have in a dedicated storage environment. Cloud Gateways Cloud gateways are responsible for receiving external connections from customers and tunneling all data transmissions over a single TCP port (also UDP for DRaaS services) protected by an SSL certificate. A cloud gateway is a Windows service, so the best platform is a modern 64-bit OS like Windows Server 2012 R2. The correct sizing of a cloud gateway depends on the amount of traffic the service provider expects to receive and on the desired level of redundancy. WAN Acceleration WAN accelerators are optional components that can be deployed at the service provider to improve bandwidth utilization of remote backups and replicas sent by customers. Although Veeam Cloud Connect operations can be performed without WAN accelerators, these components are essential for service providers offering remote backup or replication services. Some customers already have Veeam WAN accelerator capabilities, but, to properly utilize them, the service provider must deploy and configure these components. WAN accelerators are enabled in Veeam Cloud Connect licenses issued to service providers, so there is no requirement for additional licensing when they are deployed for customers. System Requirements and Design Considerations The overall solution contains Veeam Backup & Replication (with a Cloud Connect license), a NetApp SolidFire storage system as a Veeam repository, a cloud gateway server, and an optional WAN accelerator (Figure 10). You can either run all components (management, the cloud gateway, the repository, and the WAN accelerator) on the same server, or you can deploy them on separate servers for scaling purposes. 14 NetApp SolidFire Reference Architecture with

15 Figure 10) Veeam Cloud Connect backup. Virtual Deployment If you have a virtual deployment (Figure 11) already in place, you can use its compute and network resources to deploy the Veeam Cloud Connect installation as a VM on top of the hypervisor. After your Windows VM is ready, you can map your SolidFire storage into the VM to store the cloud. Figure 11) Virtual deployment. Physical Deployment In the physical deployment scenario, you build a solution stack consisting of a server for compute, network switches, and SolidFire storage. This configuration allows you to keep the stack separate from other environments so that there is no contention for resources. 15 NetApp SolidFire Reference Architecture with

16 Figure 12) Physical deployment. 4.2 System Requirements This section describes the system requirements for the SolidFire reference architecture with Veeam Backup & Replication 9.5. Cloud Gateway Table 1) System requirements for the cloud gateway. Specification CPU Memory Disk space Network OS Requirement x86 or x86-64 processor OS requirements plus cloud gateway service requirements. A single connection from a tenant consumes approximately 512KB of memory. 1GB of memory in a cloud gateway can be used to receive up to 2,000 concurrent connections. 2GB for product installation and 4.5GB for Microsoft.NET Framework installation. 10GB per 100 VMs for a guest file system catalog folder (persistent data). Additional free disk space for the instant VM recovery cache folder. For nonpersistent data, NetApp recommends at least 10GB. 1Gbps LAN 32-bit and 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 16 NetApp SolidFire Reference Architecture with

17 Veeam Backup Server Table 2) System requirements for the Veeam backup server. Specification CPU Memory Disk space Network OS Software Requirement x86-64 processor 4GB RAM plus 500MB RAM for each concurrent job. Additionally, for users with tape installations (for file-to-tape jobs processing more than 1,000,000 files): 1.5GB RAM for file-to-tape backup for every 1,000,000 files 2.6GB RAM for file restore for every 1,000,000 files 1.3GB RAM for catalog jobs for every 1,000,000 files 2GB for product installation. 4.5GB for Microsoft.NET Framework installation. 10GB per 100 VMs for a guest file system catalog folder (persistent data). Additional free disk space for an instant VM recovery cache folder. For nonpersistent data, NetApp recommends at least 10GB. 1Gbps or faster for on-site backup and replication, and 1Mbps or faster for offsite backup and replication. High latency and reasonably unstable WAN links are supported. Only 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 The setup process performs a system configuration check to determine whether all prerequisite software is available on the machine on which you plan to install Veeam Backup & Replication. If some of the required software components are missing, the setup wizard offers to install the missing software automatically. This covers the following programs: Microsoft.NET Framework Microsoft Windows Installer 4.5 Microsoft SQL Server Management Objects Microsoft SQL Server System CLR Types Microsoft Visual C SP1 Redistributable Package The following software must be installed manually: Microsoft PowerShell 2.0 or later (optional) Firefox, Google Chrome, Microsoft Edge, or Microsoft Internet Explorer 10.0 or later 17 NetApp SolidFire Reference Architecture with

18 SQL Server database Local or remote installation of the following versions of Microsoft SQL Server (both full and express editions are supported): Microsoft SQL Server 2016 Microsoft SQL Server 2014 Microsoft SQL Server 2012 (Microsoft SQL Server 2012 Express Edition SP3 is included in the setup) Microsoft SQL Server 2008 R2 Microsoft SQL Server 2008 WAN Accelerator Table 3) System requirements for the WAN Accelerator. Specification CPU Memory Disk space Network OS Software Requirement x86-64 processor. Using multicore processors improves data processing performance and is highly recommended on WAN links faster than 10Mbps. 8GB RAM. Using faster memory (DDR3/DDR4) improves data processing performance. Disk space requirements depend on the WAN accelerator role. For more information, see WAN Accelerator Sizing. 1Gbps or faster for on-site backup and replication. 1Mbps or faster for off-site backup and replication. High latency and reasonably unstable WAN links are supported. Only 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 Microsoft Windows Vista SP2 The setup process performs a system configuration check to determine if all prerequisite software is available on the machine on which you plan to install Veeam Backup & Replication. If any of the required software components are missing, the setup wizard offers to install the missing software automatically. This covers the following software: Microsoft.NET Framework Microsoft Windows Installer 4.5 Microsoft SQL Server Management Objects Microsoft SQL Server System CLR Types Microsoft Visual C SP1 Redistributable Package The following software must be installed manually: Microsoft PowerShell 2.0 or later (optional) Firefox, Google Chrome, Microsoft Edge, or Microsoft Internet Explorer 10.0 or later 18 NetApp SolidFire Reference Architecture with

19 SQL Server database Local or remote installation of the following versions of Microsoft SQL Server (both full and express editions are supported): Microsoft SQL Server 2016 Microsoft SQL Server 2014 Microsoft SQL Server 2012 (Microsoft SQL Server 2012 Express Edition SP3 is included in the setup) Microsoft SQL Server 2008 R2 Microsoft SQL Server NetApp SolidFire Arrays 5.1 SolidFire SF-Series Hardware NetApp SolidFire storage systems are architected from the ground up to be the foundation for the wideranging data storage requirements of an NGDC. Such systems are equally adept at handling large sequential I/O for video, analytical, and backup applications as well as small random I/O for small and medium-sized enterprise mixed workloads. NetApp SolidFire SF-Series storage nodes are the building blocks for a SolidFire platform, allowing you to build a broad range of scale-out, all-flash storage systems that are highly available and easy to control. Mix and match any SolidFire node within a system to take advantage of the most current and costeffective flash systems available. SolidFire systems have the following capabilities: Scale a single system predictably without performance degradation 20TB to petabytes of effective capacity 200K to millions of IOPS The NetApp SolidFire Platform Compatibility Guarantee ensures hardware compatibility. Therefore, customers can seamlessly add new SF-Series nodes to expand their system with the latest flash technology, rather than being locked into a single generation of flash. This eliminates the hardware replacement, storage migration, and forklift upgrades that are common with many traditional enterprise storage systems, ultimately protecting customers investment in their infrastructure. 5.2 Key SolidFire Features Following are some of the important features of SolidFire: Automated management Scale-out Automatic distribution and load balancing Comprehensive API Complete management integrations Regular upgrades with no interruptions Tens of terabytes to multiple petabytes Nondisruptive scaling and no downtime Mixture of nodes in the same cluster Guaranteed performance Allocate storage performance independent of capacity Manage performance in real time without affecting other volumes 19 NetApp SolidFire Reference Architecture with

20 Guarantee performance to every volume with fine-grain quality-of-service (QoS) settings Global efficiency Inline and post-compression Automatic distribution of data (no hot spots) with always-on deduplication Global thin provisioning Data assurance NetApp SolidFire Helix RAID-less data protection Real-time replication (synchronous and asynchronous) Integrated backup and recovery (cloud) Complete data and performance availability regardless of system condition or application activity Data safeguarded with 256-bit encryption 5.3 Enterprise Reliability and Availability SolidFire also provides the following features that support enterprise reliability and availability: Data assurance To provide data redundancy, SolidFire maintains two copies of every data block on two separate nodes through a technology called SolidFire Helix that is built into the NetApp SolidFire Element OS operating software. This allows a cluster to sustain application performance after failures occur. It also takes away the need to have separate storage shelves with shared drive access, making the hardware less complex and less expensive. For example, SolidFire uses single-attach SSDs instead of more expensive enterprise-grade, dual-attach SSDs. Self-healing from failures Active pool. All resources in the system are always in the active pool. There is no need to have spare drives or spare nodes sitting idle in case of failure. Drive failure. If a drive fails, the system automatically restores full redundancy by redistributing copies of data by using a meshed rebuild process. There is no degraded mode operation and no performance penalty during a rebuild. The process typically completes in five minutes or less. Because of the speed with which full redundancy is restored, this feature provides a level of data protection that exceeds RAID 6 in a typical system. Node failure. Because data copies are distributed on separate nodes, all data remains accessible if a node fails. Connections to the failed node are automatically redirected to other nodes. As with a drive failure, full redundancy is restored quickly and automatically by making sure there are two copies of each block. Regardless of the failure mode drive, node, backplane, network failure, or software failure the recovery process is the same. Because the recovery workload is distributed across all nodes in the cluster, redundancy is restored quickly, and no single node or application workload takes a performance hit. The more nodes in the cluster, the faster the activity occurs and the lower the overall impact. Active support The SolidFire Element OS is the storage operating system of the next-generation data center and is included on every SF-Series appliance. This operating system provides all the features you demand from primary storage total reliability, all-flash performance, end-to-end security, and more plus all five criteria for delivering agile, predictable storage performance. SolidFire systems also have the following capabilities: Secure assist 24/7/365 worldwide availability Expert tier 3 support engineers 20 NetApp SolidFire Reference Architecture with

21 NetApp Active IQ 5.4 Benefits of SolidFire Scale-Out Storage Nondisruptive scale-out and scale-in. Add or remove nodes from a SolidFire cluster without disrupting service or compromising volume-level QoS settings. Data is automatically redistributed in the background across all nodes in the cluster, maintaining balance as the system grows. Instant resource availability. Newly added storage capacity and performance resources are instantly available to every volume within the system, eliminating the need to reallocate volumes over new drives. Simplified capacity planning. Initial implementations begin with a simple 4-node/4U cluster configuration and scale out easily with 1U node increments, allowing performance and capacity resources to be added as needed. Eliminate multiyear capacity and performance projections and scale on demand. Seamless generational upgrades. New nodes with more capacity and performance are added to the established cluster, and old nodes are removed, retired, or repurposed. No rebalancing, restriping, or volume reallocation is required. All QoS settings remain enforced. 5.5 Element OS Software The SolidFire Element OS continues to deliver the industry s most comprehensive enterprise feature set for consolidating multiple, mixed workloads onto an agile, secure, and automated infrastructure. By integrating with VMware Virtual Volumes (VVols), increasing Fibre Channel (FC) scalability, and enhancing multitenant network capabilities, the latest Element OS significantly improves flexibility and expands the range of use cases for a SolidFire all-flash scale-out system. Key functionality includes the following features: Integrated Virtual Volumes. By integrating SolidFire QoS with VMware Virtual Volumes, you can achieve highly granular control over storage performance on a per-vm basis. This functionality allows you to set minimum, maximum, and burst IOPS levels, providing capacity levels and guaranteed performance for even the most performance-sensitive VMs. You can change both capacity and performance dynamically without migrating data or affecting performance. Fibre Channel scalability. Connect to up to four FC nodes in a SolidFire cluster, and scale out to 40 storage nodes. Each FC node supports 500,000 IOPS, resulting in 2,000,000 IOPS for a four-fcnode cluster. Expansive VLAN features. You can tag default networks and support more flexibility in multitenant networks by enabling customers to use overlapping IP addresses on VLANs through separate routing tables, preserving limited IP address resources. Intuitive user interface (UI). The UI consolidates the storage system data into a single dashboard, saving customers time and resources. Only SolidFire offers a true shared-nothing architecture, which enables you to upgrade to the Element OS nondisruptively. SolidFire has offered 100% nondisruptive software and hardware upgrades since the SolidFire all-flash, scale-out storage platform was released for general availability in November The SolidFire Element OS is the storage operating system of the next-generation data center and is included on every SF-Series appliance. This operating system provides all the features you demand from primary storage total reliability, all-flash performance, end-to-end security, and more plus all the features needed to deliver agile, predictable storage performance: True scale-out architecture Guaranteed performance Automated management Data assurance 21 NetApp SolidFire Reference Architecture with

22 depicts the Element OS UI dashboard. Figure 13) Element OS UI dashboard Overview Veeam Backup & Replication is a data protection and disaster recovery solution for VMware vsphere and Microsoft Hyper-V virtual environments of any size or complexity. By combining all the necessary functions in one intuitive interface, Veeam Backup & Replication solves the most critical problems of virtualized infrastructure management. The solution also protects mission-critical VMs from both hardware and software failures. 6.2 Solution Architecture Veeam Backup & Replication is composed of the following three elements: Backup server Backup proxy Backup repository Veeam Backup Server The Veeam backup server is the Windows-based physical machine or VM on which Veeam Backup & Replication is installed. It is the core component in the backup infrastructure that fills the role of the configuration and control center. The Veeam backup server performs all types of administrative activities, including the following tasks: It coordinates backup, replication, recovery verification, and restore tasks. It controls job scheduling and resource allocation. It is used to set up and manage backup infrastructure components and to specify global settings for the backup infrastructure. 22 NetApp SolidFire Reference Architecture with

23 In addition to its primary functions, a newly deployed Veeam backup server also acts as a default backup proxy and the backup repository (it manages data handling and data storing tasks). The Veeam backup server uses the following services and components: Veeam Backup Service. A Windows service that coordinates all the operations that Veeam Backup & Replication performs, such as backup, replication, recovery verification, and restore tasks. The Veeam Backup Service runs under the local system account or an account that has the local administrator permissions on the backup server. Veeam broker service. Interacts with the virtual infrastructure to collect and cache the virtual infrastructure topology. Jobs and tasks query information about the virtual infrastructure topology from the broker service, which accelerates job and task performance. Veeam Backup Catalog/Guest Catalog service. Manages guest OS file system indexing for VMs and replicates system index data files to enable search through guest OS files. Index data is stored in the Veeam backup catalog, which is a folder on the backup server. The Veeam guest catalog service running on the backup server works in conjunction with search components that are installed on Veeam Backup Enterprise Manager and (optionally) a dedicated Microsoft Search Server. Veeam mount service. Mounts backups and replicas for file-level access, browsing the VM guest file system, and restoring VM guest OS files and application items to the original location. Veeam backup proxy services. In addition to dedicated services, the backup server runs a set of data-mover services. For details, see the section Backup Proxy. Veeam Backup & Replication configuration database. Stores data about the backup infrastructure, jobs, sessions, and so on. The database instance can be on a SQL Server that is installed either locally on the same machine where the backup server is running or remotely. Veeam Backup & Replication console. Provides the application UI and allows user access to the application s functionality. Veeam Backup PowerShell snap-in. An extension for Microsoft Windows PowerShell 2.0. Veeam Backup PowerShell adds cmdlets that enable you to perform backup, replication, and recovery tasks through the PowerShell CLI or run custom scripts. In this way, you can fully automate Veeam Backup & Replication operation. Backup Proxy When Veeam Backup & Replication is first installed, the Veeam backup server coordinates all job activities and handles data traffic. Therefore, when you perform a backup, replication, VM copy, or VM migration job or perform restore operations, VM data is moved from the source to the target through the Veeam backup server. This scenario is acceptable for virtual environments in which few backup jobs are performed. In large-scale environments, however, the workload on the Veeam backup server is significant. To take the workload off the Veeam backup server, Veeam Backup & Replication uses backup proxies. A backup proxy is an architecture component that sits between the data source and the target and is used to process jobs and deliver backup traffic. Backup proxy tasks include retrieving VM data from production storage. Tasks also include compressing the data and sending it to the backup repository (for example, if you run a backup job) or to another backup proxy (for example, if you run a replication job). As a data handling task is assigned to the backup proxy, the Veeam backup server becomes the point of control for dispatching jobs to proxy servers. The role of a backup proxy can be assigned to a dedicated Windows Server (physical or virtual) in your virtual environment. You can deploy backup proxies in both the primary site and remote sites. To optimize the performance of several concurrent jobs, you can use multiple backup proxies. In this case, Veeam Backup & Replication distributes the backup workload between available backup proxies. By using backup proxies, you can easily scale your backup infrastructure up and down according to your demands. Backup proxies run lightweight services that take a few seconds to deploy, and deployment is 23 NetApp SolidFire Reference Architecture with

24 fully automated. Veeam Backup & Replication installs the necessary components on a Windows-based server when you add it to the product console. As soon as you assign the role of a backup proxy to the added server, Veeam Backup & Replication starts the required services on it. The primary role of the backup proxy is to provide an optimal route for backup traffic and to enable efficient data transfer. Therefore, when deploying a backup proxy, you must analyze the connection between the backup proxy and the storage with which it is working. Depending on the type of connection, the backup proxy can be configured in one of the following ways (starting with the most efficient): A machine used as a backup proxy should have direct access to the storage on which VMs reside or to the storage to which VM data is written. In this way, the backup proxy retrieves data directly from the datastore, bypassing the LAN. The backup proxy can be a VM with HotAdd access to VM disks on the datastore. This type of proxy also enables LAN-free data transfer. If neither of the preceding scenarios is possible, you have alternatives. You can assign the role of the backup proxy to a machine on the network that is closer to the source or closer to the target storage with which the proxy works. In this case, VM data is transported over the LAN by using the Network Block Device (NBD) protocol. Depending on the type of backup proxy and your backup architecture, the backup proxy can use one of the following data transport modes: direct SAN access, virtual appliance, or network. If the VM disks are on the SAN storage and the SAN storage is added to the Veeam Backup & Replication console, the backup proxy can also use the backup from storage snapshots mode. You can select the transport mode or let Veeam Backup & Replication automatically choose it. The backup proxy uses the following services and components: Veeam Installer Service. An auxiliary service that is installed and started on any Windows Server after it is added to the list of managed servers in the Veeam Backup & Replication console. This service analyses the system and installs and upgrades necessary components and services depending on the role that is selected for the server. Veeam Transport. Responsible for deploying and coordinating executable modules that act as data movers and that perform the main job activities on behalf of Veeam Backup & Replication. These activities include communicating with VMware Tools, copying VM files, performing data deduplication and compression, and so on. Backup Repository A backup repository is a location that Veeam Backup & Replication jobs use to store backup files, copies of VMs, and metadata for replicated VMs. Technically, a backup repository is a folder on the backup storage. By assigning different repositories to jobs and by limiting the number of parallel jobs for each one, you can balance the load across your backup infrastructure. 7 Requirements 7.1 VMware System Requirements Veeam Backup Server (VMware) Table 4 lists the specifications and requirements of the Veeam backup server with VMware. 24 NetApp SolidFire Reference Architecture with

25 Table 4) Veeam backup server specifications (VMware). Specification CPU Memory Disk space Network OS Software SQL Server database Requirement Modern x86-64 processor (minimum 2 cores or vcpus) 4GB RAM plus 500MB RAM for each concurrent job 2GB for product installation. 4.5GB for Microsoft.NET Framework installation. 10GB per 100 VMs for a guest file system catalog folder (persistent data). Additional free disk space for the instant VM recovery cache folder (for nonpersistent data, NetApp recommends at least 10GB). 1Gbps or faster for on-site backup and replication. 1Mbps or faster for off-site backup and replication. High latency and reasonably unstable WAN links are supported. Only 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows Server 2003 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 The setup process performs a system configuration check to determine if all prerequisite software is available on the machine on which you plan to install Veeam Backup & Replication. If some of the required software components are missing, the setup wizard can install missing software automatically. This covers the following software: Microsoft.NET Framework Microsoft Windows Installer 4.5 Microsoft SQL Server Management Objects Microsoft SQL Server System CLR Types Microsoft Visual C SP1 Redistributable Package The following software must be installed manually: Microsoft PowerShell 2.0 or later (optional) Firefox, Google Chrome, Microsoft Edge, or Microsoft Internet Explorer 10.0 or later Local or remote installation of the following versions of Microsoft SQL Server (both full and express editions are supported): Microsoft SQL Server 2016 Microsoft SQL Server 2014 Microsoft SQL Server 2012 (Microsoft SQL Server 2012 Express Edition SP3 is included in the setup) Microsoft SQL Server 2008 R2 Microsoft SQL Server NetApp SolidFire Reference Architecture with

26 If you plan to back up VMs running Windows Server 2012 R2 and later, and data deduplication is enabled for some VM volumes, NetApp recommends that you deploy the Veeam Backup & Replication console and mount server on a machine running the same or later version of Windows Server with the data deduplication feature enabled. Otherwise, some types of restore operations for these VMs (such as Windows File Level Recovery) might fail. Due to its limitations, SQL Server Express Edition can be used only for evaluation purposes or in a smallscale production environment. For environments with many VMs, you must install a fully functional commercial version of SQL Server. You cannot deploy Veeam Backup & Replication and Veeam Backup Enterprise Manager configuration databases on Microsoft SQL AlwaysOn availability groups. Veeam Backup & Replication Console (VMware) Table 5 lists the specifications and requirements of the Veeam Backup & Replication console (VMware). Table 5) Veeam Backup & Replication console (VMware). Specification CPU Memory Disk space Network OS Software Requirement Modern x86 processor (minimum 2 cores or vcpus). The use of multicore processors improves data processing performance and enables more tasks to be processed concurrently by the backup proxy. 2GB RAM 500MB for product installation 4.5GB for Microsoft.NET Framework installation 1Mbps connection to the backup server. High latency and low bandwidth affect UI responsiveness. Only 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows Server 2003 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 Microsoft.NET Framework (included in setup) Windows Installer 4.5 (included in setup) Microsoft PowerShell 2.0 or later (optional) Firefox, Google Chrome, Microsoft Edge, or Microsoft Internet Explorer 10.0 or later Backup Proxy Server (VMware) Table 6 lists the specifications and requirements of the backup proxy server (VMware). 26 NetApp SolidFire Reference Architecture with

27 Table 6) Backup proxy server specifications (VMware). Specification CPU Memory Disk space Network OS Software Requirement Modern x86 processor (minimum 2 cores or vcpus). The use of multicore processors improves data processing performance, and allows more tasks to be processed concurrently by the proxy. 2GB RAM plus 200MB for each concurrent task. Using faster memory (DDR3/DDR4) improves data processing performance. 300MB 1Gbps or faster for on-site backup and replication. 1Mbps or faster for off-site backup and replication. High latency and reasonably unstable WAN links are supported. Both 32-bit and 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows Server 2003 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 Microsoft Windows Vista SP2 For a vsphere 5.5/6.0 backup proxy server running on Windows Server 2008 or earlier or a vsphere 6.5 backup proxy server running on Windows Server 2008 R2 or earlier: Microsoft Visual C SP1 Redistributable Package (x64). The installation package can be downloaded from Note: To protect VMs running on ESXi 5.5 and later, you must deploy backup proxies on machines running a 64-bit version of Windows. VDDK 5.5 and later versions do not support 32-bit versions of Windows. Backup Repository Server (VMware) Table 7 lists the specifications and requirements of the backup repository server (VMware). Table 7) Backup repository server specifications (VMware). Specification CPU Memory Disk space Requirement x86 processor (x86-64 recommended) 4GB RAM, plus up to 2GB RAM (32-bit OS) or up to 4GB RAM (64-bit OS) for each concurrent job depending on the backup chain s length and the backup files sizes 200MB for Veeam Backup & Replication components and sufficient disk space to store backup files and replicas. NetApp recommends high-rpm drives and a RAID 10 configuration. 27 NetApp SolidFire Reference Architecture with

28 Network OS 1Gbps or faster for on-site backup and replication, and 1Mbps or faster for offsite backup and replication. High latency and reasonably unstable WAN links are supported. Both 32-bit and 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows Server 2003 SP2 Microsoft Windows 8.x Microsoft Windows 7 SP1 Microsoft Windows Vista SP2 Linux (Bash shell, SSH, and Perl are required). Check the full list of required Perl modules at A 64-bit edition of Linux must be able to run 32-bit programs. Pure 64-bit Linux editions are not supported (Perl installation must support 32-bit variables). Note: If you plan to use a Windows backup repository with data deduplication, make sure that you set up Windows Server correctly. For more information, see the Veeam Knowledge Base. 7.2 Hyper-V System Requirements Veeam Backup Server (Hyper-V) Table 8 lists the specifications and requirements of the Veeam backup server with Microsoft Hyper-V. Table 8) Veeam backup server specifications (Hyper-V). Specification CPU Memory Disk space Network Requirement x86-64 processor 4GB RAM plus 500MB RAM for each concurrent job 2GB for product installation 4.5GB for Microsoft.NET Framework installation 10GB per 100 VMs for the guest file system catalog folder (persistent data) 1Gbps or faster for on-site backup and replication. 1Mbps or faster for off-site backup and replication. High latency and reasonably unstable WAN links are supported. 28 NetApp SolidFire Reference Architecture with

29 OS Software SQL Server database Only 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows Server 2003 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 The setup process performs a system configuration check to determine if all prerequisite software is available on the machine on which you plan to install Veeam Backup & Replication. If some of the required software components are missing, the setup wizard can install missing software automatically. This task covers the following software: Microsoft.NET Framework Microsoft Windows Installer 4.5 Microsoft SQL Server Management Objects Microsoft SQL Server System CLR Types Microsoft Visual C SP1 Redistributable Package The following software must be installed manually: Microsoft PowerShell 2.0 or later (optional) Firefox, Google Chrome, Microsoft Edge, or Microsoft Internet Explorer 10.0 or later System Center Virtual Machine Manager 2016, 2012 R2, or 2012 or 2008 R2 Admin UI (optional, needed to register the System Center Virtual Machine Manager server with Veeam Backup & Replication infrastructure) Remote Desktop Protocol (RDP) client version 7.0 and later installed on the backup server (required to open the VM console during SureBackup recovery verification of Microsoft Hyper-V VMs). The RDP client is preinstalled on Windows 7 and Windows Server 2008 R2 OS and later. You can download the RDP client from Local or remote installation of the following versions of Microsoft SQL Server (both full and express editions are supported): Microsoft SQL Server 2016 Microsoft SQL Server 2014 Microsoft SQL Server 2012 (Microsoft SQL Server 2012 Express Edition is included in the setup) Microsoft SQL Server 2008 R2 Microsoft SQL Server 2008 Microsoft SQL Server 2005 Note: If you plan to back up VMs running Windows Server 2012 R2 and later and data deduplication is enabled for some VM volumes, NetApp recommends that you deploy the Veeam Backup & Replication console and mount server on a machine running the same or a later version of Windows Server. The data deduplication feature should be enabled. Otherwise, some types of restore operations for these VMs (such as Windows File Level Recovery) might fail. 29 NetApp SolidFire Reference Architecture with

30 Due to its limitations, SQL Server Express Edition can be used only for evaluation purposes or in a small-scale production environment. For environments with many VMs, you must install a fully functional commercial version of SQL Server. You cannot deploy Veeam Backup & Replication and Veeam Backup Enterprise Manager configuration databases on Microsoft SQL AlwaysOn availability groups. Veeam Backup & Replication Console (Hyper-V) Table 9 lists the specifications and requirements of the Veeam Backup & Replication console. Table 9) Veeam Backup & Replication console (Hyper-V). Specification CPU Memory Disk space Network OS Software Requirement Modern x86 processor (minimum 2 cores or vcpus). The use of multicore processors improves data processing performance and enables more tasks to be processed concurrently by the backup proxy. 2GB RAM 500MB for product installation 4.5GB for Microsoft.NET Framework installation 1Mbps connection to the backup server. High latency and low bandwidth affect user interface responsiveness. Only 64-bit versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows Server 2003 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 Microsoft.NET Framework (included in the setup) Windows Installer 4.5 (included in the setup) Microsoft PowerShell 2.0 or later (optional) Firefox, Google Chrome, Microsoft Edge, or Microsoft Internet Explorer 10.0 or later RDP client version 7.0 and later (required to open the VM console during SureBackup recovery verification of Microsoft Hyper-V VMs). The RDP client is preinstalled on Windows 7/Windows Server 2008 R2 OS and later. You can download the RDP client from Off-Host Backup Proxy Server (Hyper-V) Table 10 lists the specifications and requirements for an off-host backup proxy server with Hyper-V. 30 NetApp SolidFire Reference Architecture with

31 Table 10) Off-host backup proxy server specifications (Hyper-V). Specification CPU Memory Disk space Network OS Software Requirement Modern x86 processor (minimum 2 cores or vcpus). Using multicore processors improves data processing performance, and permits more tasks to be processed concurrently by proxy. 2GB RAM plus 200MB for each concurrent task. Using faster memory (DDR3/DDR4) improves data processing performance. 300MB 1Gbps or faster for on-site backup and replication. 1Mbps or faster for off-site backup and replication. High latency and reasonably unstable WAN links are supported. Microsoft Windows Server 2016 with Hyper-V role enabled Microsoft Windows Server 2012 R2 with Hyper-V role enabled Microsoft Windows Server 2012 with Hyper-V role enabled Microsoft Windows Server 2008 R2 SP1 with Hyper-V role enabled Volume Shadow Copy Service (VSS) hardware provider that supports transportable shadow copies. The VSS hardware provider is typically distributed as a part of the client components supplied by the storage vendor. Backup Repository Server (Hyper-V) Table 11 lists the specifications and requirements of a backup repository server with Hyper-V. Table 11) Backup repository server specifications (Hyper-V). Specification CPU Memory Disk space Network Requirement x86 processor (x86-64 recommended) 4GB RAM, plus up to 2GB RAM (32-bit OS) or up to 4GB RAM (64-bit OS) for each concurrent job depending on the backup chain s length and the backup files sizes 200MB for Veeam Backup & Replication components and sufficient disk space to store backup files and replicas. NetApp recommends high-rpm drives and a RAID 10 configuration. NetApp recommends a 1Gbps LAN for on-site backup and replication and a 1Mbps WAN for off-site backup and replication. High latency and reasonably unstable WAN links are supported. 31 NetApp SolidFire Reference Architecture with

32 OS Both 32-bit and 64-bit (recommended) versions of the following operating systems are supported: Microsoft Windows Server 2016 Microsoft Windows Server 2012 R2 Microsoft Windows Server 2012 Microsoft Windows Server 2008 R2 SP1 Microsoft Windows Server 2008 SP2 Microsoft Windows Server 2003 SP2 Microsoft Windows 10 Microsoft Windows 8.x Microsoft Windows 7 SP1 Microsoft Windows Vista SP2 Linux SSH and Perl are required (check the full list of required Perl modules at A 64-bit edition of Linux must be able to run 32-bit programs. Pure 64-bit Linux editions are not supported. 7.3 Veeam Sizing Requirements For Veeam sizing requirements, see Table 12. Table 12) Veeam sizing requirements. Veeam Sizing Requirements Veeam backup proxy Veeam backup server Veeam repository Virtual. 1 vcpu for every 200 VMs (recommended minimum: 2 vcpus), 4GB RAM, 300MB disk space for installed components, and a VMXNET3 network interface card (NIC). No CPU or memory reservations are required. 4GB of RAM plus 500MB of RAM for each concurrent job. Disk space: 2GB for the product, plus 10GB per 100 VMs for a guest file system catalog and at least 10GB for a VM recovery cache folder. No CPU or memory reservations are required. Additional sizing considerations should be applied if the backend SQL Server is deployed on this server. For additional guidance, see the Veeam documentation. The Veeam repository can be collocated with the backup server role for small deployments or on a dedicated server (physical or virtual Windows or Linux server, NAS device, or dedicated backup appliance). The repository requires sufficient free space to store all backup job data. If a VM is used, NetApp recommends using the VMware Paravirtual SCSI (PVSCSI) controller for the disk or disks that store backup data. vsphere 5.5 enables the use of disks larger than 2TB, which might be advantageous for a repository server. Physical. 1 CPU core for every 250 VMs, 2GB of RAM for each concurrent Virtual Machine Disk (VMDK) backup. 32 NetApp SolidFire Reference Architecture with

33 8 Design Considerations 8.1 Direct SAN Access Mode Veeam proxy servers directly connect to the storage fabric because this configuration is the fastest way to perform backups with Backup & Replication. Fast backup speeds are critical for low RPOs. An often-discussed point about direct SAN access for the Veeam proxy is the possibility that the Veeam proxy can mount VMware Virtual Machine File System (VMFS) LUNs and write a Windows signature onto the LUN. Doing so results in an inaccessible VMFS, which requires VMware support to fix the issue. Veeam Knowledge Base (KB) article #1446 provides instructions on how to configure SAN access for use with Veeam Backup & Replication. Although Microsoft Windows 2003 has automount enabled by default, the SAN policy in the Windows Server 2008 R2 Enterprise and Datacenter editions is set to Offline Shared by default. Even if an administrator changes the policy to Online All, Veeam reverts the policy back to Offline Shared. As a result, Windows does not mount and resignature VMFS LUNs unless an administrator changes this setting again or manually mounts VMFS LUNs while ignoring Windows warnings. NetApp recommends the direct SAN access transport mode for VMs whose disks are on shared VMFS SAN LUNs that are connected to ESX(i) hosts through FC or iscsi. In the direct SAN access transport mode, Veeam Backup & Replication leverages VMware vstorage APIs for Data Protection (formerly VADP) to transport VM data directly from and to FC and iscsi storage over the SAN. VM data travels over the SAN, bypassing ESX(i) hosts and the LAN. The direct SAN access transport method provides the fastest data transfer speed and produces no load on the production network. You can use the direct SAN access transport mode for all operations in which the backup proxy is engaged, including the following: Backup Replication Full VM restore VM disk restore Replica failback Quick migration Requirements for the Direct SAN Access Transport Mode To use the direct SAN access transport mode, you must meet the following requirements: The backup proxy that uses the direct SAN access transport mode must have direct access to the production storage through a hardware or software host bus adapter (HBA). If a direct SAN connection is not configured or is not available when a job or a task starts, the job or the task fails. For restore operations, the backup proxy must have write access to LUNs where the VM disks are located. Limitations for the Direct SAN Access Transport Mode The direct SAN access transport mode has the following limitations: The direct SAN access transport mode is not supported for VMs that reside on a virtual storage area network. You can use the virtual appliance or network transport mode to process such VMs. For details about virtual storage area network restrictions, see the VDDK 5.5 Release Notes. If at least one VM disk is on a VVol, you cannot use direct SAN access transport mode. 33 NetApp SolidFire Reference Architecture with

34 You can use the direct SAN access transport mode only for the initial run of the replication job. For subsequent replication job runs, Veeam Backup & Replication uses the virtual appliance or network transport mode. You can use only the direct SAN access transport mode to restore thick VM disks. Because of VMware limitations, you cannot use the direct SAN access transport mode for incremental restore. You must either disable changed block tracking (CBT) for VM virtual disks during the restore process or select another transport mode for incremental restore. For VMware vsphere 5.5 and later, IDE and SATA disks can be processed in the direct SAN access transport mode. For VMware vsphere 5.1 and earlier, consider the following issues: IDE disks can be backed up in the direct SAN access transport mode. However, restore of IDE disks in the direct SAN access transport mode is not supported. If a VM disk fails to be processed in the direct SAN access transport mode, Veeam Backup & Replication does not fail over to network mode. If a VM has some disks that cannot be processed in the direct SAN access transport mode, Veeam Backup & Replication uses network mode for VM disk processing. 8.2 Data Restore in Direct SAN Access Transport Mode Data restore in the SAN access transport mode includes the following steps (see Figure 14): 1. The backup proxy retrieves data blocks from the backup repository or a datastore in the target site. 2. The backup proxy sends a request to the ESX(i) host on the source site to restore data to a necessary datastore. 3. The ESX(i) host on the source site allocates space on the datastore. 4. Data blocks that are obtained from the backup proxy are written to the datastore. Figure 14) Data restore in direct SAN access transport mode. You can use only direct SAN access transport mode to restore VMs with thick disks. Before VM data is restored, the ESX(i) host must allocate space for the restored VM disk on the datastore: When thick disks are restored, the ESX(i) host allocates space on the disk before writing VM data. 34 NetApp SolidFire Reference Architecture with

35 When thin disks are restored, the ESX(i) host attempts to allocate space dynamically as requests for data block restores are received. As a result, restoring thin disks involves extra allocation overhead when compared with restoring thick disks, which results in decreased performance. To restore VMs with thin disks, you can use virtual appliance mode or network mode. If you plan to process a VM that has both thin and thick disks, select direct SAN access transport mode and choose to fail over to the network mode if the SAN becomes inaccessible. In that case, Veeam Backup & Replication uses direct SAN access transport mode to restore thick disks and network transport mode to restore thin disks. Alternatively, you can restore all VM disks as thick. 8.3 Off-Host Backup (Hyper-V) In off-host backup mode, backup processing is shifted from the source Hyper-V host to a dedicated machine, an off-host backup proxy. The off-host backup proxy acts as a data mover. The Veeam transport service that runs on it retrieves VM data from the source datastore, processes it, and transfers it to the destination. This type of backup does not impose load on the Hyper-V host. Production hosts remain unaffected while resource-intensive backup operations are performed on the off-host backup proxy. To perform off-host backup, Veeam Backup & Replication uses transportable shadow copies. Transportable shadow copy technology lets you create a snapshot of a data volume on one server and import, or mount, it onto another server within the same subsystem (SAN) for backup and other purposes. The transport process is accomplished in a few minutes, regardless of the amount of data. The process is performed at the SAN storage layer, so it does not affect host CPU usage or network performance. To perform off-host backup, you must meet the following requirements: You must configure an off-host backup proxy. You can assign the role of an off-host backup proxy to the following systems only: A physical Microsoft Windows 2008 Server R2 machine with the Hyper-V role enabled A Windows Server 2012 machine with the Hyper-V role enabled A Windows Server 2012 R2 machine with the Hyper-V role enabled A Windows Server 2016 machine with the Hyper-V role enabled The version of the Hyper-V host and off-host backup proxy must be the same. For example, if you use a Windows 2008 Server R2 machine with the Hyper-V role enabled as a Hyper-V host, deploy an off-host backup proxy running the same software. In the properties of a backup or replication job, you must select the off-host backup method. If necessary, you can point the job to a specific proxy. The source Hyper-V host and the off-host backup proxy must be connected through a SAN configuration to the shared storage. To create and manage volume shadow copies on the shared storage, you must install and properly configure a VSS hardware provider that supports transportable shadow copies on the off-host proxy and the Hyper-V host. Typically, when configuring a VSS hardware provider, you must specify a server that controls the LUN and disk array credentials to provide access to the array. The VSS hardware provider is usually distributed as part of the client components that are supplied by the storage vendor. Any VSS hardware provider that is certified by Microsoft is supported. Some storage vendors might require additional software and licensing to work with transportable shadow copies. If you back up VMs whose disks reside on a Windows Cluster Shared Volume (CSV) with data deduplication enabled, make sure that you perform the following tasks: Use a Windows 2012 R2 or a Windows 2016 machine as an off-host backup proxy. Enable data deduplication on this off-host backup proxy. 35 NetApp SolidFire Reference Architecture with

36 Otherwise, off-host backup fails. The off-host backup process (Figure 15) includes the following steps: 1. Veeam Backup & Replication triggers a snapshot of the necessary volume on the production Hyper-V host. 2. The created snapshot is split from the production Hyper-V server and is mounted to the off-host backup proxy. 3. The Veeam transport service that runs on a backup proxy uses the mounted volume snapshot to retrieve VM data. The VM data is processed on the proxy server and is copied to the destination. 4. When the backup process is complete, the snapshot is dismounted from the off-host backup proxy and is deleted from the SAN. Figure 15) Off-host backup process. Note: If you plan to perform off-host backup for a Hyper-V cluster with CSV, deploy an off-host backup proxy on a host that is not part of a Hyper-V cluster. When a volume snapshot is created, this snapshot has the same LUN signature as the original volume. The Microsoft Cluster Service does not support LUNs with duplicate signatures and partition layout. For this reason, volume snapshots must be transported to an off-host backup proxy outside the cluster. If the off-host backup proxy is deployed on a node of a Hyper-V cluster, a duplicate LUN signature is generated and the cluster fails during backup or replication. Off-Host Backup Proxy By default, when you perform backup, replication, or VM copy jobs in the Hyper-V environment, VM data is processed directly on the source Hyper-V host where the VMs reside. The data is then moved to the target, bypassing the Veeam backup server. VM data processing can produce unwanted overhead on the production Hyper-V host and can affect the performance of the VMs that run on this host. To take data processing off the production Hyper-V host, use off-host backup mode. The off-host mode shifts the backup and replication load to a dedicated machine, an off-host backup proxy. The off-host backup proxy functions as a data mover that retrieves VM data from the source datastore, processes it, and transfers it to the destination. The machine that performs the role of an off-host backup proxy must meet the following requirements: 36 NetApp SolidFire Reference Architecture with

37 The role can be assigned only to the following systems: A physical Microsoft Windows 2008 Server R2 machine with the Hyper-V role enabled A Windows Server 2012 machine with the Hyper-V role enabled A Windows Server 2012 R2 machine with the Hyper-V role enabled A Windows Server 2016 machine with the Hyper-V role enabled The off-host backup proxy must have access to the shared storage that hosts the VMs to be backed up, replicated, or copied. To create and manage volume shadow copies on the shared storage, you must install a VSS hardware provider that supports transportable shadow copies on the off-host proxy and the Hyper-V host. The VSS hardware provider is usually distributed as part of the client components that are supplied by the storage vendor. When you assign the role of off-host backup proxy to a machine, Veeam Backup & Replication automatically installs the lightweight components and services needed for a backup proxy. Unlike the Veeam backup server, backup proxies do not require a dedicated SQL Server database. All the settings are stored centrally within the SQL Server database that the Veeam backup server uses. To enable a Hyper-V host or a Windows machine to act as an off-host backup proxy, Veeam Backup & Replication installs the following services on it: Veeam Installer Service. An auxiliary service that is installed and started on any Windows (or Hyper- V) server. Installation occurs after the server is added to the list of managed servers in the Veeam Backup & Replication console. This service analyzes the system and installs and upgrades necessary components and services. Veeam transport. Responsible for deploying and coordinating executable modules that act as data movers and that perform the main job activities on behalf of Veeam Backup & Replication. These activities include performing data deduplication, compression, and so on. Veeam Hyper-V Integration Service. Responsible for communicating with the VSS framework during backup, replication, and other jobs and for performing recovery tasks. The service also deploys a driver that handles changed block tracking for Hyper-V. 8.4 Components Table 13 describes each component in. Table 13) components. Component Veeam backup server Description The Veeam backup server is a Windows-based physical machine or VM on which Veeam Backup & Replication is installed. It is the core component in the backup infrastructure that fills the role of the configuration and control center. The Veeam backup server performs all administrative activities: Coordinates backup, replication, recovery verification, and restore tasks Controls job scheduling and resource allocation Is used to set up and manage backup infrastructure components and to specify global settings for the backup infrastructure In addition to its primary functions, a newly deployed Veeam backup server also performs the roles of the default backup proxy and the backup repository. (It manages data handling and data storing tasks.) 37 NetApp SolidFire Reference Architecture with

38 Component Veeam Backup Service Veeam Backup Shell Veeam Backup Catalog/Guest Catalog Service Veeam Backup SQL Server database Veeam Backup PowerShell snap-in Backup proxy services Backup proxy Backup repository Description This Windows service coordinates all the operations that are performed by Veeam Backup & Replication, such as backup, replication, recovery verification, and restore tasks. Veeam Backup Service runs under the local system account or the account that has the local administrator permissions on the Veeam backup server. Veeam Backup Shell provides the application UI and enables user access to the application s functionality. This Windows service manages guest OS file system indexing for VMs and replicates system index data files to enable search through guest OS files. Index data is stored in the Veeam Backup Catalog, a folder on the Veeam backup server. The Veeam Backup Catalog/Guest Catalog Service running on the Veeam backup server works with search components that are installed on Veeam Backup Enterprise Manager and (optionally) a dedicated Microsoft Search Server. This database is used by Veeam Backup Service, Veeam Backup Shell, and Veeam Backup Catalog Service to store data about the backup infrastructure, jobs, sessions, and so on. The database instance can be on a SQL Server installed either locally (on the same machine where the Veeam backup server runs) or remotely. This snap-in is an extension for Windows PowerShell 2.0. Veeam Backup PowerShell adds a set of cmdlets to enable you to perform backup, replication, and recovery tasks through the PowerShell CLI or run custom scripts. In this way, you can fully automate Veeam Backup & Replication operation. In addition to dedicated services, the Veeam backup server runs a set of data-mover services. A backup proxy is an architecture component that sits between the data source and the target and is used to process jobs and deliver backup traffic. The backup proxy retrieves VM data from the production storage. It also compresses the retrieved data and sends it to the backup repository (for example, if you run a backup job) or to another backup proxy (for example, if you run a replication job). As the data handling task is assigned to the backup proxy, the Veeam backup server becomes the point of control for dispatching jobs to proxy servers. A backup repository is a location that Veeam Backup & Replication jobs use to store backup files, copies of VMs, and metadata for replicated VMs. Technically, a backup repository is a folder on the backup storage. By assigning different repositories to jobs and by limiting the number of parallel jobs for each one, you can balance the load across your backup infrastructure. 9 SolidFire Volume Configuration Guidelines 9.1 SolidFire Quality of Service NetApp SolidFire QoS guarantees an unprecedented level of control over performance through Min IOPS, Max IOPS, and Burst IOPS. QoS delivers specific IOPS and bandwidth. 38 NetApp SolidFire Reference Architecture with

39 Min IOPS. The minimum number of IOPS that an administrator grants to a volume. When an application requests an IOPS value of at least the Min IOPS, the SolidFire cluster designates resources to fill the request. The Min IOPS level is the guaranteed IOPS and is the focus of most SLA provisions. This control is imperative to enable performance-sensitive applications to run in multitenant cloud environments. Max IOPS. The maximum number of sustained IOPS that a volume is allowed to process over an extended period. With this control, noisy neighbors (applications that steal performance from other applications) are eliminated. Burst IOPS. The maximum number of IOPS that is allowed in a short-burst scenario. If a volume has been running below the Max IOPS, burst credits are accumulated. When performance levels become very high and are pushed to maximum levels, short bursts of IOPS are allowed on the volume. This functionality is particularly useful during VM reboots, database flushes, and large file transfers. 9.2 SolidFire Storage Configuration Guidelines for Backup & Replication 9.5 Backup Repositories SolidFire volumes make excellent Veeam backup repositories because of their inherent speed and control. For optimal performance, NetApp recommends that you follow these QoS guidelines: Run multiple backup jobs to each repository. Use multiple volumes per repository for maximum performance. For SolidFire volume QoS information for backup repositories, see Table 14. Table 14) SolidFire volume QoS for backup repositories. Parameter Volume QoS Multipath input/output (MPIO) Suggested Tune Low Min IOPS that is sufficient to cover throughput requirements during peak load times, and high Max IOPS and Burst IOPS. Tuning the volume in this manner permits large volumes that do not consume a lot of the usable IOPS of your SolidFire system but permits the volume to burst its usage as backups are executed. Example: (Min:Max:Burst 4,000:25,000:25,000) You should connect SolidFire volumes used as backup repositories to the backup server through MPIO by following the instructions in the NetApp SolidFire Storage for Microsoft Windows Configuration Guide. This configuration permits maximum throughput to the volume. 9.3 SolidFire Volumes as VMware Datastores SolidFire volumes leveraged as VMware (or other hypervisor) datastores function exactly as any other datastore volumes backed up by Veeam. However, some tuning can be performed on both the SolidFire volume as well as the systems used for Veeam. For SolidFire volume QoS for VMware datastores, see Table 15. Table 15) SolidFire volume QoS for VMware datastores. Parameter Volume QoS Suggested Tune Full backups. Increase the Max IOPS and Burst IOPS to 100,000 during the backup. You should return Max IOPS and Burst IOPS to their normal levels after the backup completes. Incremental backups. Increase the Max IOPS and Burst IOPS by a multiple of four (for example: 1,000:5,000:5,000 becomes 1,000:20,000:20,000). You should return Max IOPS and Burst IOPS to their normal levels after the backup completes. 39 NetApp SolidFire Reference Architecture with

40 These changes can be performed automatically by scripts at the beginning and end of the backup job. 9.4 Directly Connected SolidFire Volume (File System Backups) This section includes client-side tunes for Windows systems with directly connected (through iscsi) SolidFire volumes. For SolidFire volume QoS for directly connected SolidFire volumes, see Table 16. Table 16) SolidFire volume QoS for directly connected SolidFire volumes. Parameter Volume QoS Suggested Tune Full backups. Increase the Max IOPS and Burst IOPS to 100,000 during the backup. You should return Max IOPS and Burst IOPS to their normal levels after the backup completes. Incremental backups. Increase the Max IOPS and Burst IOPS by a multiple of four (for example: 1,000:5,000:5,000 becomes 1,000:20,000:20,000). You should return Max IOPS and Burst IOPS to their normal levels after the backup completes. These changes can be performed automatically by scripts at the beginning and end of the backup job. 10 SolidFire Host Configuration Guidelines 10.1 Host Connectivity iscsi All hosts connect to a NetApp SolidFire cluster through the storage virtual cluster IP address. After the connection is established, a redirector service passes the connection to the node that contains the primary metadata copy for the volume the host is requesting access to. During the node removal process, great care is taken to prevent a host from losing connection to its volume. If there is an active connection from the host to a node that is being removed, the secondary copy is promoted to primary and a new secondary copy is created. A seamless iscsi redirect then occurs, and the application remains connected to its volume. A similar process occurs when you add a node to a cluster. The newly added node takes on its share of volumes to help evenly distribute load and capacity on the system. As volumes are rebalanced to provide optimal placement, a seamless iscsi redirect occurs if the active connection is moved to one of the new nodes added to the system. Fibre Channel Nodes SolidFire FC nodes provide connectivity to an FC switch that you can connect to FC clients. FC nodes act as a protocol converter between the FC and iscsi protocols. This allows you to add FC connectivity to any new or existing SolidFire cluster. FC nodes have the following characteristics: FC switches manage the state of the fabric, providing optimized interconnections. The traffic between two ports flows only through the switches; it is not transmitted to any other port. Failure of a port is isolated and does not affect the operation of other ports. Multiple pairs of ports can communicate simultaneously in a fabric. 40 NetApp SolidFire Reference Architecture with

41 11 SolidFire Plug-In for VMware vcenter The NetApp SolidFire Plug-In for VMware vcenter is a VMware vcenter Server plug-in that provides integrated management of SolidFire storage arrays from within a VMware vsphere web client. The vsphere web client is a single management interface that you can use to manage the VMware infrastructure and all your day-to-day storage needs. By managing SolidFire within vcenter, you can: Discover and manage multiple SolidFire clusters. Perform vsphere datastore create, extend, clone, share, and delete operations. Perform SolidFire account create, edit, and delete operations. Perform SolidFire volume create, edit, clone, delete, and access group add and remove operations. Perform SolidFire access group create, edit, and delete operations. For the latest download and details, see the SolidFire FTP site. 12 SolidFire VSS Hardware Provider 12.1 SolidFire VSS Hardware Provider Introduction The NetApp SolidFire VSS Hardware Provider integrates VSS shadow copies with SolidFire snapshots and clones. The provider runs on Windows 2008 R2 and 2012 R2 editions and supports shadow copies created using DiskShadow and other VSS requesters. A GUI-based configuration utility is provided for adding, modifying, and removing cluster information used by the SolidFire VSS Hardware Provider. Using VSS snapshot capabilities with the SolidFire VSS Hardware Provider creates snapshots that are application consistent with business applications that use SolidFire volumes on a system. A coordinated effort between VSS components provides this functionality. SolidFire snapshots and clones minimize recovery time without affecting stability and performance. Snapshots also provide instant volume recovery SolidFire VSS Hardware Provider Overview There are three major components of the Windows VSS: VSS providers, VSS writers, and VSS requesters. A VSS provider is a storage-level component that offers functionality to create a shadow copy of one or more volumes. A VSS writer is application-specific software that makes sure that application data is ready for shadow copy creation. The application that initiates the creation of a shadow copy is a VSS requester. The high-level architecture diagram below shows the interaction between VSS components and the SolidFire system. The backup application DiskShadow is a VSS requester for creating shadow copies of volumes. When DiskShadow executes a backup, VSS orchestrates the interaction between the VSS requester, VSS writers, and VSS providers. VSS maintains application consistency when creating a volume shadow copy. There are several VSS writers installed on a system, such as the Microsoft SQL Server VSS writer. The SolidFire VSS Hardware Provider manages the shadow copies created by a VSS requester at the hardware level by using SolidFire snapshots and clones. The SolidFire VSS Hardware Provider uses snapshots and clones to make shadow copies available to the VSS requester. Shadow copy volumes permit immediate read-only access to the data from the Windows host. The Windows host must have access to both storage and the management network so that the VSS provider can work correctly. Administrators must add one or more clusters to the SolidFire VSS Hardware Provider configuration. This allows the provider to connect to SolidFire clusters and interact with the SolidFire API to create snapshots and clones during the backup process. 41 NetApp SolidFire Reference Architecture with

42 12.3 SolidFire VSS Hardware Provider Requirements Hardware Prerequisites Table 17 lists the SolidFire VSS Hardware Provider hardware prerequisites. Table 17) Hardware prerequisites. Component System processor System memory Free disk space required for installation Networking SolidFire cluster Description 1.5GHz (dual core recommended), Intel compatible 64-bit Minimum 4GB recommended Minimum 5MB required; 100MB recommended Ethernet ports for cluster communication; Ethernet ports for iscsi or FC A cluster of SolidFire nodes Software Prerequisites Table 18 lists the SolidFire VSS Hardware Provider software prerequisites. Table 18) Software prerequisites. Component Application Description Operating system Microsoft Windows Server Windows Server 2008 R2 64-bit Windows Server 2012 R2 64-bit Microsoft Hyper-V.NET framework 4.5 SolidFire cluster Supported Configurations Windows Server 2012 R2 64-bit Element OS version 7.0 and later Table 19 lists the supported configurations. VSS Hardware Provider configurations leverage directly connected SolidFire iscsi or FC volumes. Only the configurations described in the preceding tables are supported. Table 19) Supported configurations. Supported Version Operating systems Windows Server 2008 R2 Windows Server 2012 R2 Hyper-V Server 2012 R2 Windows Server 2008 R2 Standalone Shared Storage Cluster AlwaysOn Cluster SQL Server 2008 R2 SP3 SQL Server 2012 SP2 SQL NetApp SolidFire Reference Architecture with

43 Windows Server 2008 R2 Standalone Shared Storage Cluster AlwaysOn Cluster SQL Server 2008 R2 SP3 SQL Server 2012 SP2 SQL 2014 For the latest download and details, see the SolidFire FTP download site. 13 Active IQ NetApp Active IQ is a web-based tool that provides continually updated historical views of clusterwide statistics. Data for volumes, nodes, drives, and entire clusters is stored for five years and is updated at regular intervals depending on the data being monitored. You can set up notifications about specified events, thresholds, or metrics on a cluster so that you can address them when they arise. The Active IQ tool makes monitoring capacity and performance, as well as being informed about cluster health, easy and accessible from anywhere. For more information, go to the NetApp SolidFire Active IQ site. 14 Summary The combination of Veeam Backup & Replication and NetApp SolidFire brings simplicity to operating and protecting highly virtualized, multitenant environments, while delivering powerful and cutting-edge capabilities. SolidFire arrays offer the performance that you need when recovering your applications and give you confidence that the data that you backed up is protected and is available when you need it. Where to Find Additional Information NetApp SolidFire Documentation SolidFire Element OS support NetApp Support Portal Software Documentation Veeam documentation, guides, and datasheets Appendix A: QoS Tuning Scripts Included are three sample Windows PowerShell scripts that can be used as a basis for creating QoS tuning scripts. libsf.ps1. This library script includes helper functions that are called by the following two scripts. It should be saved in the same place as the other two scripts. pre-backup.ps1. This script increases the Max IOPS and Burst IOPS QoS settings on a volume during the backup by the multiplier given. For example, a multiplier of four increases a volume s QoS settings from 1,000:4,000:4,000 to 1,000:16,000:16,000 (Min:Max:Burst). The original QoS settings are saved in temporary file storage ($env:temp\<mvip>-<volumeid>.qos) to restore the original 43 NetApp SolidFire Reference Architecture with

44 settings after backup. If the multiplier is excluded, it is set to 20, which sets the Max IOPS and Burst IOPS values to at or near the system maximum of 100,000. post-backup.ps1. This script reads the original QoS settings from the temporary file location and resets QoS settings back to the prebackup levels. libsf.ps1 Function sflog() { Param( [parameter(position=1)] [System.String]$Message, $Color, [switch] $Deb, [switch] $Info, [switch] $Warn, [switch] $Error ) if (-not $Message) { Write-Host return } if ($Deb) { $fg = "Gray" } elseif ($Info) { $fg = "White" } elseif ($Warn) { $fg = "Yellow" } elseif ($Error) { $fg = "Red" } else { $fg = "White" } if ($Color) { $fg = $Color } $datestamp = Get-Date -Format "yyyy-mm-dd HH:mm:ss" $Message = $Message.Replace("`r`n", "`n") foreach ($line in $Message.Split("`n")) { Write-Host -ForegroundColor $fg "$datestamp $line" } } Function Ssh-Command() { Param( $Hostname, $RemoteCommand, $Username = "root", $Password = "solidfire" ) [ reflection.assembly]::loadfrom((resolve-path " Renci.SshNet.dll")) Out-Null { Param( $VmHost, $Rescan = $true ) sflog -Info "Rescanning iscsi on $VmHost" $iscsi_hba = Get-VMHostHba -Type iscsi -VMHost $VmHost Get-VMHostStorage -VMHost $VmHost -RescanAllHba -RescanVmfs Out-Null # Get a list of iscsi volumes sflog -Info "Looking for new iscsi volumes (this may take a few minutes)..." $luns = Get-ScsiLun -Hba $iscsi_hba $iqn2volume $volume2lun foreach ($iscsi_lun in $luns) { $path_info = Get-ScsiLunPath -ScsiLun $iscsi_lun $iqn = $path_info.sanid $pieces = $iqn.split(".") $volume_name = ($pieces[4] + "." + $pieces[5]) $iqn2volume[$iqn] = $volume_name 44 NetApp SolidFire Reference Architecture with

45 if ($volume2lun.containskey($volume_name)) { sflog -Error "Duplicate volume name detected. This script cannot handle that case." exit } $ volume2lun[$volume_name] = $ iscsi_lun.canonicalname } # Find the datastore for each LUN $host_obj = Get-VMHost -Name $VmHost Get-View $all_luns = $host_obj.config.storagedevice.scsilun?{$_.vendor -match "SolidFir"} $all_datastores = $host_obj.config.filesystemvolume.mountinfo $lun2ds foreach ($ds in $all_datastores?{$_.volume.extent.count -gt 0}) { $name = $ds.volume.extent[0].diskname foreach ($lun in $all_luns) { if ($lun.canonicalname - eq $ name) { $lun2ds[$name] = $ds.volume.name } } } # Create datastores for LUNs $new_ds = $false foreach ($iqn in $iqn2volume.keys Sort-Object) { $volume = $iqn2volume[$iqn] $lun = $volume2lun[$volume] if ($lun2ds.containskey($lun)) { $ds = $lun2ds[$lun] #sflog -Debug " $iqn -> $volume -> $lun -> $ds" } else { sflog -Info " Creating new datastore '$volume' with iscsi volume $iqn..." New-Datastore -VMHost $VmHost -Name $volume -Path $lun -Vmfs Out-Null $new_ds = $true } } return $new_ds } pre-backup.ps1 Param( $Mvip = "A.B.C.D", $SfUsername = "SFQoSadmin", $SfPassword = "SFPassword", $VolumeID = "VolumeID", $IOPSMultiplier = 20 ) #### Usage #### ## pre-backup.ps1 -Mvip <SF MVIP> -SfUsername <SF User Name> -SfPassword <SF Password> -VolumeID <Volume ID> [-IOPSMultiplier <Multiplier>] # Example: pre-backup.ps1 -Mvip SfUsername qosadmin -SfPassword qospassword - VolumeID 3 -IOPSMultiplier 4 ## Use in ISE debugger or standard PS shell (not in PowerCLI shell) #### Helper Functions #### $workingdir = Split-Path $MyInvocation.MyCommand.Path -Parent. "$workingdir\libsf.ps1" #### Main Script #### $ErrorActionPreference = "Stop" $WarningPreference = "SilentlyContinue" $ProgressPreference = "SilentlyContinue" 45 NetApp SolidFire Reference Architecture with

46 $api_params = New-Object psobject $api_params Add-Member -Name "startvolumeid" -Value $VolumeID -MemberType NoteProperty Out- Null $api_params Add-Member -Name "limit" -Value 1 -MemberType NoteProperty Out-Null $vol_details = Call-ApiMethod -Mvip $Mvip -Username $SfUsername -Password $SfPassword -MethodName ListActiveVolumes -MethodParams $api_params $vol_details Export-Clixml $env:temp\$mvip-$volumeid.qos $api_params = New-Object psobject $api_params Add-Member -Name "volumeid" -Value $VolumeID -MemberType NoteProperty Out-Null $qos_params = New-Object psobject $qos_params Add-Member -Name "miniops" -Value $vol_details.volumes[0].qos.miniops -MemberType NoteProperty Out-Null IF (($vol_details.volumes[0].qos.maxiops * $IOPSMultiplier) -gt ) { $qos_params Add-Member -Name "maxiops" -Value MemberType NoteProperty Out-Null } ELSE { $qos_params Add-Member -Name "maxiops" -Value ($vol_details.volumes[0].qos.maxiops * $IOPSMultiplier) -MemberType NoteProperty Out-Null } IF (($vol_details.volumes[0].qos.burstiops * $IOPSMultiplier) -gt ) { $qos_params Add-Member -Name "burstiops" -Value MemberType NoteProperty Out-Null } ELSE { $ qos_params Add-Member - Name " burstiops" - Value ($vol_details.volumes[0].qos.burstiops * $IOPSMultiplier) -MemberType NoteProperty Out-Null } $api_params Add-Member -Name "qos" -Value $qos_params -MemberType NoteProperty Out-Null Call-ApiMethod -Mvip $Mvip -Username $SfUsername -Password $SfPassword -MethodName ModifyVolume - MethodParams $api_params Out-Null post-backup.ps1 Param( $Mvip = "A.B.C.D", $SfUsername = "SFQoSadmin", $SfPassword = "SFPassword", $VolumeID = "VolumeID" ) #### Usage #### ## post-backup.ps1 -Mvip <SF MVIP> -SfUsername <SF User Name> -SfPassword <SF Password> -VolumeID <Volume ID> # Example: pre-backup.ps1 -Mvip SfUsername qosadmin -SfPassword qospassword - VolumeID 3# # Use in ISE debugger or standard PS shell (not in PowerCLI shell) #### Helper Functions #### $workingdir = Split-Path $MyInvocation.MyCommand.Path -Parent. "$workingdir\libsf.ps1" #### Main Script #### $ErrorActionPreference = "Stop" $WarningPreference = "SilentlyContinue" $ProgressPreference = "SilentlyContinue" $vol_details = Import-Clixml $env:temp\$mvip-$volumeid.qos $api_params = New-Object psobject $api_params Add-Member -Name "volumeid" -Value $VolumeID -MemberType NoteProperty Out-Null $qos_params = New-Object psobject $qos_params Add-Member -Name "miniops" -Value $vol_details.volumes[0].qos.miniops -MemberType NoteProperty Out-Null $qos_params Add-Member -Name "maxiops" -Value $vol_details.volumes[0].qos.maxiops -MemberType NoteProperty Out-Null $qos_params Add-Member -Name "burstiops" -Value $vol_details.volumes[0].qos.burstiops - MemberType NoteProperty Out-Null $api_params Add-Member -Name "qos" -Value $qos_params -MemberType NoteProperty Out-Null Call-ApiMethod -Mvip $Mvip -Username $SfUsername -Password $SfPassword -MethodName ModifyVolume - MethodParams $api_params Out-Null 46 NetApp SolidFire Reference Architecture with

47 Refer to the Interoperability Matrix Tool (IMT) on the NetApp Support site to validate that the exact product and feature versions described in this document are supported for your specific environment. The NetApp IMT defines the product components and versions that can be used to construct configurations that are supported by NetApp. Specific results depend on each customer s installation in accordance with published specifications. Copyright Information Copyright NetApp, Inc. All rights reserved. Printed in the U.S. No part of this document covered by copyright may be reproduced in any form or by any means graphic, electronic, or mechanical, including photocopying, recording, taping, or storage in an electronic retrieval system without prior written permission of the copyright owner. Software derived from copyrighted NetApp material is subject to the following license and disclaimer: THIS SOFTWARE IS PROVIDED BY NETAPP AS IS AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, WHICH ARE HEREBY DISCLAIMED. IN NO EVENT SHALL NETAPP BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. NetApp reserves the right to change any products described herein at any time, and without notice. NetApp assumes no responsibility or liability arising from the use of products described herein, except as expressly agreed to in writing by NetApp. The use or purchase of this product does not convey a license under any patent rights, trademark rights, or any other intellectual property rights of NetApp. The product described in this manual may be protected by one or more U.S. patents, foreign patents, or pending applications. RESTRICTED RIGHTS LEGEND: Use, duplication, or disclosure by the government is subject to restrictions as set forth in subparagraph (c)(1)(ii) of the Rights in Technical Data and Computer Software clause at DFARS (October 1988) and FAR (June 1987). Trademark Information NETAPP, the NETAPP logo, and the marks listed at are trademarks of NetApp, Inc. Other company and product names may be trademarks of their respective owners. 47 NetApp SolidFire Reference Architecture with