“NetApp”的意思、由来-开放百科全书

NetApp, Inc. is a hybrid cloud data services and data management company headquartered in Sunnyvale, California. It has ranked in the Fortune 500 since 2012.^[3] Founded in 1992^[4] with an IPO in 1995,^[5] NetApp offers hybrid cloud data services for management of applications and data across cloud and on-premises environments.

History

NetApp was founded in 1992 by David Hitz, James Lau,^[6] and Michael Malcolm.^[4]^[7] At the time, its major competitor was Auspex Systems. In 1994, NetApp received venture capital funding from Sequoia Capital.^[8] It had its initial public offering in 1995. NetApp thrived in the internet bubble years of the mid 1990s to 2001, during which the company grew to $1 billion in annual revenue. After the bubble burst, NetApp's revenues quickly declined to $800 million in its fiscal year 2002. Since then, the company's revenue has steadily climbed.

In 2006, NetApp sold the NetCache product line to Blue Coat Systems^[9]. On June 1, 2015, Tom Georgens stepped down as CEO and was replaced by George Kurian.^[10]

In May 2018 NetApp announced its first End to End NVMe array called All Flash FAS A800 with release of ONTAP 9.4 software^[11]. NetApp claims over 1.3 million IOPS at 500 microseconds per high-availability pair; Read throughput of up to 300GBps (Giga Byte per second) per all-flash 24 node cluster and 50% higher IOPS and up to 34% lower latency by upgrading previous model A700 with ONTAP 9.4^[12].

Acquisitions

Competition

NetApp competes in the computer data storage hardware industry.^[17] In 2009, NetApp ranked second in market capitalization in its industry behind EMC Corporation, now Dell EMC, and ahead of Seagate Technology, Western Digital, Brocade, Imation, and Quantum.^[18] In total revenue of 2009, NetApp ranked behind EMC, Seagate, Western Digital, and ahead of Imation, Brocade, Xyratex, and Hutchinson Technology.^[19] According to a 2014 IDC report, NetApp ranked second in the network storage industry "Big 5's list", behind EMC(DELL), and ahead of IBM, HP and Hitachi.^[20]. According to Gartner’s 2018 Magic Quadrant for Solid-State Arrays NetApp named a leader, behind with only Pure Storage Systems.

Products

NetApp's OnCommand management software controls and automates data-storage.^[21]. ActiveIQ comes to NetApp with the acquisition of SolidFire. ActiveIQ is SaaS portal with built-in monitoring, prediction, recommendations for optimizing configurations and performance for NetApp storage systems based on machine-learning capabilities and artificial intelligence. Later ONTAP Analytics and Telemetry Service (OATS) product which can be installed in AWS cloud and on-premise was renamed to Active IQ Performance Analytics Services (ActiveIQ PAS).

NetApp FAS and AFF

NetApp's FAS (Fabric-Attached Storage) and AFF (All-Flash FAS) storage systems serve as the company's flagship products. Such a product is made up of a storage controller, and one or more enclosures of hard disks, known as shelves. In entry-level systems, the drives may be physically located in the storage controller itself.

In the early 1990s, NetApp's storage systems initially offered NFS and SMB protocols based on standard local area networks (LANs), whereas block storage consolidation required storage area networks (SANs) implemented with the Fibre Channel (FC) protocol.

In 2002, in an attempt to increase market share, NetApp added block-storage access as well, supporting the Fiber Channel and iSCSI protocols. {{As of | 2016}} NetApp systems support Fibre Channel, iSCSI, Fibre Channel over Ethernet (FCoE) and the FC-NVMe protocol.

ONTAP

NetApp AFF, FAS, ONTAP Select, Cloud Volumes ONTAP and Cloud Volume Services, Lenovo DM storage system series use NetApp's proprietary operating system called ONTAP which includes code from Berkeley Net/2 BSD Unix, Spinnaker Networks technology and other operating systems.^[22] There are three ONTAP platforms: FAS/AFF systems, software on commodity servers (ONTAP Select) as virtual machine or in the cloud (Cloud Volumes ONTAP). All ONTAP systems using WAFL file system which provide basis for snapshots and other snapshot-based and data protection technologies.

Cloud Backup

Previously known as Riverbed SteelStore after acquired by NetApp renamed to AltaVault and then to Cloud Backup. Cloud Backup was initially available in three forms: as a hardware appliance, virtual appliance, and cloud appliance. Later NetApp announced the end of sale for hardware and virtual appliances. Data placed on NAS share on Cloud Backup deduplicated, compressed, encrypted and transferred with Object Protocols to object storage systems like Amazon S3, Azure Blob Storage or StorageGRID; thus Cloud Backup appears as a transparent gateway for archiving data to a private or public cloud. There are two modes of Cloud Backup operation: with local caching or without. Cloud Backup systems could accept SnapMirror replication from ONTAP systems.

HCI

NetApp Hyper-converged infrastructure (HCI) or sometimes referred by NetApp as Hybrid Cloud Infrastructure. NetApp HCI is based on commodity blade servers, NetApp Element software and VMware vSphere. NetApp HCI includes web-based GUI with installation wizard called NetApp Deployment Engine (NDE) for configuring vCenter, IP addresses, login & password, and storage nodes, install NetApp Element plugin in vCenter to manage storage and configure NetApp HCI in about 30 minutes according to NetApp.^[23] NetApp HCI different than conventional HCI designs because it has dedicated storage nodes, while other HCI systems like Nutanix or Dell EMC VxRail or vSAN do not have dedicated storage nodes and utilize disk drives installed in each server. Dedicated storage nodes allow the cluster to grow or decrease storage capacity and performance separately from compute nodes. Minimum NetApp HCI configuration requires two compute blade server nodes and additionally, Element software requires a minimum of 4 storage nodes but is available to customers as four physical storage nodes or two physical storage nodes and two nodes as a virtual machine playing witness role on compute nodes. Each storage node drive set consists of 6 SSD drives directly connected to a dedicated storage node and installed in front of the blade chassis, all drives in a set must be the same capacity. Each node either storage or compute server is one rack unit half-width, so minimum NetApp HCI configuration consumes four rack units with six used blade server slots and two empty slots for expansion. Each storage and compute blade nodes have 25 Gigabit Ethernet ports which could be used as 10Gbit/s ports as well as dedicated 1Gb ports for management purposes. There are three types of compute nodes differentiated from each other by CPU power and memory capacity. Though there are three types of storage nodes, differentiated from each other by SSD capacity. It is possible to intermix any compute and storage nodes in a NetApp HCI installation, though with storage nodes it is required to have a minimum of 2 with the same capacity. It is also possible to use NetApp HCI with software other than VMware vSphere. Network switches are not included in NetApp HCI and must be bought separately from third parties, while all other hardware components must be bought from NetApp. ONTAP Select available as SDS on NetApp HCI for customers interested in NAS protocols. Self-Service portal allows automating common provisioning and management tasks without involving the IT team. NetApp Kuberneties Service will support NetApp HCI with the acquisition of Stackpoint. NetApp SolidFire storage and NetApp HCI can be expanded & mixed in a single cluster. At the Insight 2018 conference in Las Vegas NetApp presented two new compute nodes: H410C and H610C, where H610C includes additional GPU cards which can be used in VDI environments. Starting with Element OS version 11, automatically detected and enabled by default with the upgrade, HCI have Protection Domains functionality to provide resiliency into HCI chassis. In the case of maintenance or chassis failure helix algorithm spans data blocks workload will automatically fail-over to another operational chassis.

SolidFire

SolidFire storage system uses OS called NetApp Element Software (formally SolidFire Element OS) based on Linux and designed for SSDs and scale-out architecture with the ability to expand up to 100 nodes and provide access to data through SAN protocols iSCSI natively and Fiber Channel with two gateway nodes. Element OS provides a REST-based API for storage automation, configuration, management, and consumption. SF node H610S has 12 2.5" NVMe SSD drives and can install only Element version 10.4, while previous models have 10 SSD drives. Element SW version 11, will not support FC. SolidFire uses iSCSI login redirection to distribute reads and writes across the cluster using helix algorithm.^[24] This architecture does not have disk shelves like traditional storage systems and expands with adding nodes to the cluster. Each node has pre-installed SSD drives. Each node can have only one type of SSD drives with the same capacity. Each SolidFire cluster can have a mix of different node models & generations. Element X uses the replication factor of 2, where blocks of data spread across the cluster which has no performance impact but require more space in contrary to Erasure Coding technology. Such architecture allows users to expand performance and capacity separately as needed. Also, SolidFire has the ability to set three types of QoS for its LUNs: minimum, maximum and burst. Burst is used as credits which were not used by the LUN while it was not received its maximums. Element X available as software-only on commodity servers. SolidFire systems using S3 protocol could backup data to an Object storage systems like StorageGRID. SolidFire could replicate data with SnapMirror protocol to ONTAP systems and starting with Element OS 11 to Cloud Volumes ONTAP. VEEAM backup & Replication 9.5 Update 4 will implement seamless integration with Netapps HCI and Solidfire provide application consistent storage snapshot capabilities, Instant VM Recovery, and Single Item Restore for some applications^[25]. CommVault Simpana also provides application-consistent storage snapshot capability for Netapps HCI and Solidfire. All HCI configurations require at list 4 10/25 Gbps ports for connections until Element OS 11, where two ports are enough.

E-Series

Previously known as LSI Engenio after NetApp acquisition the product renamed to NetApp E-Series. It is a general purpose enterprise storage system with two controllers for SAN protocols such as Fibre Channel, iSCSI, SAS and InfiniBand (includes SRP, iSER, and NVMe over Fabrics protocol). NetApp E-Series platform uses proprietary OS SANtricity and proprietary RAID called Dynamic Disk Pool (DDP) alongside with traditional RAIDs like RAID 10, RAID 6, RAID 5 etc. In DDP pool each D-Stripe works similar to traditional RAID-4 and RAID-6 but on block level instead of entire disk level therefore have no dedicated parity drives. DDP compare to traditional RAID groups restores data from lost disk drive to multiple drives which provide few times faster reconstruction time^[26] while traditional RAIDs restores lost disk drive to a dedicated parity drive. Starting with SANtricity 11.50 E-Series systems EF570 and E5700 support NVMe over Ethernet (RoCEv2) with 100Gbps Ethernet ports and NVMe over InfiniBand. Sync and async mirroring are supported with SANtricity 11.50. SANtricity Unified Manager is a web-based manager that supports up to 500 EF/E-Series arrays and supports LDAP, RBAC, CA & SSL for authorization & authentication.

StorageGRID

StorageGRID is a software-defined storage system which provides access to data via object IP-based protocols like S3 and OpenStack Swift. StorageGRID is available both as an appliance based on NetApp's E-Series system and as software-only as a virtual machine or docker container package. A node in a StorageGRID cluster is an appliance, virtual machine or docker container. StorageGRID is a geo-dispersed namespace clustered storage system, also known as "the grid", with an ability to make and store multiple copies (replicas) of objects (also known as Replication Factor) or in Erasure Coding (EC) manner among cluster storage nodes with object granularity based on configured policies for data availability and durability purposes. StorageGRID cluster can contain a mix of a different type (virtual machines, containers & physical appliances), performance, capacity and media, and doesn't have to be perfectly symmetrical by design, and can span up to 16 geographically distributed data centers around the world. StorageGRID EC capable of streaming chunks of data or byte range reads, instead of delivering entire object also known as all or nothing. All the objects could be accessed through any storage node in a StorageGRID cluster regardless of where they may be physically located. StorageGRID provides integration with Amazon's S3 service which can be used as Archive Level for objects or as a party which accepting replica from StorageGRID. StorageGRID stores metadata separately from the objects and allows users to configure data Life Cycle Management (ILM) policies on a per-object level to automatically satisfy and confirm changes in the cluster once changes introduced to the cluster like the cost of network usage, storage media usage changes a node was added or removed, etc. For example, it is possible to configure the system to move each new object to the fastest media and store 3 replicas each on a separate node; in 90 days remove one replica (so only 2 copies will be stored), and as for objects which weren't accessed for few years to move them to cheapest storage media or if there will be added new storage node with the cheapest storage media to the cluster, those objects will be moved there automatically and then remove all the replicas of those objects and create Erasure Coding instead. It is also possible to configure the system to make copies or move objects automatically to those storage nodes where they are requested the most. In version 11 StorageGRID supports CloudMirror functionality to copy data from StorageGRID system to Amazon S3 storage. StorageGRID can issue Event Notifications for events occurring on premises to AWS SNS to build event-driven pipelines using on-premises and cloud resources. CloudMirror does not generate additional fees for Inbound Data Transfer with Amazon S3 storage, but only fees for storage space consumed in Amazon S3 storage. Since all the data already available on StorageGRID system does not support CloudMirror replication back, and there is no need for it to copy data back thus no Outbound Data Transfer fees from S3 storage would apply. Company Bycast developed StorageGRID and was acquired by NetApp in 2010. StorageGRID is also available as a software-only system on commodity servers. ONTAP, Cloud Backup, SANtricity, and Element X can replicate data to StorageGRID systems. Minimum StorageGRID cluster requirements: 1 admin node, 3 storage nodes and optionally gateway nodes (usually at least 2), so all of those nodes (virtual machines or containers) can be placed on at least 3 physical servers in SDS deployment; or 3 physical NetApp appliances for storage nodes and 2 servers for the rest of the nodes. In October 2018 at Insight conference in Las Vegas, NetApp announced new SG6060 appliance consisting of E-series system with two controllers and one server which can connect to the E-Series via two FC 16Gbps ports. E-Series have 58 HDD drives and 2 SSD drives used for system boot and support purposes. Where server can run SG storage node and, to run a cluster needed three sets of those SG6060 appliances. SG6060 requires minimum version 11.1.1 of StorageGRID software. SG6060 is optimized for high transactional throughput, MA, AI, and FabricPool. Starting with 11.2 StorageGRID supports Amazon S3 Glacier for archived data and according to NetApp performance improved up to 2-3 times.

StorageGRID on NetApp HCI

In November 2018 NetApp issued technical report TR-4734 with Solution Deployment of StorageGRID on NetApp HCI which can be deployed in three forms: Fully contained; High performance and scale; NetApp HCI and StorageGRID appliance. Fully contained deployment used for running admin & Storage Nodes as virtual machines stored on HCI with SSD media. High performance and scale deployment used for running Admin & Storage Nodes as virtual machines on HCI but Admin Node stored on HCI SSD media while Storage Nodes placed on a NetApp E-Series system for object data. NetApp HCI and StorageGRID appliance deployment used for running Admin Node & Load Balancers on HCI with SSD media, while storage nodes running as physical StorageGRID appliances.

Converged Infrastructure

FlexPod, nFlex and ONTAP AI are commercial names for Converged Infrastructure (CI). Converged Infrastructures are joint products of a few vendors and consists from 3 main hardware components: computing servers, switches (in some cases switches are not necessary) and NetApp storage systems:

Converged Infrastructures have tested and validated design configurations from vendors available to end users and typically include popular infrastructure software like Docker Enterprise Edition (EE), Red Hat OpenStack Platform, VMware vSphere, Microsoft Servers and Hyper-V, SQL, Exchange, Oracle VM and Oracle DB, Citrix Xen, KVM, OpenStack, SAP HANA etc. and might include self-service portals PaaS or IaaS like Cisco UCS Director (UCSD) or others. FlexPod, nFlex and ONTAP AI allow an end user to modify validated design and add or remove some of the components of the Converged Infrastructure while not all of the other Converged Infrastructures from competitors allows modification.

FlexPod

There are few FlexPod types: FlexPod Datacenter, FlexPod Select, FlexPod Express (Small, Medium, Large and UCS-managed), FlexPod SF. FlexPod Datacenter usually using Nexus switches like 5000, 7000 & 9000; Cisco UCS Blade Servers; Mid-Range or High-End NetApp FAS or AFF systems. FlexPod Select often used with BigData framework software like Hortonworks & Cloudera, the architecture using Cisco UCS Rack Servers with direct attached NetApp E-Series and in some configurations in addition to that include Low-End FAS systems & Cisco Switches. FlexPod Express usually have Low-End NetApp FAS/AFF systems; Small, Medium, Large using Cisco UCS rack servers with Nexus 3000 switches while UCS-managed FlexPod Express using Cisco Blade servers, may add rack servers and might include switches from Cisco or other vendors in the architecture. FlexPod SF has in its architecture Nexus 9000 switches, Cisco UCS Blade servers and NetApp SolidFier storage based on Cisco UCS rack servers. Cisco UCS Director used as the orchestrator for FlexPod for a self-service portal, workflow automation and billing platform to build PaaS & IaaS. FlexPod systems supported under the cooperative center of competence. NetApp Converged Systems Advisor (CSA) is a software-as-a-service (SaaS) platform that consists of an on-premises agent and a cloud-based portal. Converged Systems Advisor validates the deployment of FlexPod infrastructure and provides continuous monitoring and notifications to ensure business continuity. CSA validates configs with an automated review for best-practice rules, monitoring FlexPod remotely for compliance with best practices, notifying administrators about recommendations for hardware and firmware compliance. Multi-Pod is a FlexPod Datacenter solution with a FAS or AFF system leveraging MetroCluster technology for stretching storage system between two sites. NetApp and Cisco looking to incorporate NetApp MAX Data product into FlexPod solutions once persistent memory technology will be available in UCS servers.

FlexPod Datacenter has the biggest variety of designed and validated by Cisco and NetApp architectures and applications including:

nFlex

Is Converged infrastructure architecture with next key components: NetApp FAS/AFF systems, Extreme Networks data center switches and Fujitsu Primergy servers. nFlex available with FUJITSU Software Enterprise Service Catalog Manager which provides a self-service portal for enterprises and service providers to automate the delivery of their software services, infrastructure services, or platform services to their employees and customers.

ONTAP AI

Converged infrastructure solution based on Cisco Nexus 3000 switches with 100Gbps ports, NetApp AFF storage systems, Nvidia DGX supercomputer servers interconnected by RDMA over RoCE, and developed for Deep Learning based on Docker containers with NetApp Docker Plugin Trident. With SnapMirror ONTAP AI solution can deliver data between edge computing, on-prem & the cloud as part of Data Fabric vision. ONTAP AI tested & validated for use with NFS & FlexGroup technologies. Combined technical support provided to the customers to all the architecture components.

OnCommand Insight

OnCommand Insight (OCI) is data center management software, capacity management, infrastructure analytics, centralized view into historical trends to forecast performance and capacity requirements and workload placement. OCI works with all NetApp storage systems and with competitor storage systems and in public cloud. Licensed server-based software.

Memory Accelerated Data

NetApp MAX Data for short. NetApp officially announced MAX Data product availability at NetApp Insight 2018 in October and currently supported with RHEL Server 7.5, CentOS 7.5; and done some vSphere 6.7 and KVM testing while containers are in the long-term roadmap. MAX Data also can be run in the cloud. MAX Data came from acquisition of Plexistor company in May of 2017 and positioned for Artificial Intelligence (Deep learning & Machine Learning, Fraud Prevention/Banking), Real-Time Analytics/Trading Platforms, Data Wherehouse, IoT workloads and applications like MongoDB, Cassandra DB, Couchbase, Oracle DB, JDE type databases, Spark, etc. SAP HANA is currently qualifying whether MAX Data is suitable for the application. MAX Data consists of two tiers: Tier 1 and Tier 2, where cold data destaged to Tier 2 from Tier 1 or promoted from Tier 2 to Tier 1 when accessed, by MAX Data tiering algorithm, transparently to the applications. Currently, NetApp has recommended ratio for MAX Data as 1 to 25 for Tier 1 and Tier 2 respectively. Plexistor was founded in 2013. MAX Data according to NetApp^[27] will have two modes: to use MAX Data as a POSIX-compatible (the internal name is M1FS) file system or as API memory extension. MAX Data installed on hosts to utilize ultra-low latency with persistent memory such as the Optane, NVDIMM or DRAM (when persistence not needed) memory for Tier 1 and a storage system like NetApp AFF for Tier 2. Optane is the Intel brand name of products that use 3D XPoint technology. MAX FS is Persistent Memory based Filesystem (PM-based FS) which doesn't require application modification but also can be Direct Access enabled File system (DAX-enabled FS) for applications with optimization for Persistent Memory using SPDK. DAX is the mechanism that enables direct access to files stored in persistent memory arrays without the need to copy the data through the page cache^[28]. MAX FS based on ZUFS (Zero-copy User-mode File System) interface for the user-space file system and do not need a dedicated kernel module, in comparison to filesystems like NTFS based on FUSE, or XFS & EXT4 directly connected to vfs, and therefore each FS require its own kernel module. ZUFS operating in Linux userspace and consisting out of two modules called zuf (Zu Feeder) & zus (Zu Server) available as OpenSource; zuf is a kernel to userspace bridge. MAX Data used as Memory Extention is NUMA aware, byte addressable near-memory latency (3-10 microseconds), optimized for random I/O, persistent memory and as FUSE-based filesystems do not have page caching. MAX Data is integrated with NetApp ONTAP as part of Data Fabric vision. Currently, MAX Data works over SAN (FC & iSCSI) with AFF systems on the back end, but NetApp considering to add FC-NVMe protocol, and NAS as well and been tested MAX Data with SDS instances of ONTAP, i.e., ONTAP Select and NetApp Cloud Volumes ONTAP according to Tech ONTAP podcast 154 in SoundCloud. NetApp is also looking into integration MAX Data with ONTAP FlexCache functionality for NFS caching in the long run. NetApp also tested MAX Data with Element OS but currently needs some server-side modifications. MAX Data provides persistent memory capabilities and include data protection capability like snapshots (MAX Snap) and SnapMirror, Clusterization for High Availability (Synchronous mirror), cloning, and data tiering found as a key functionality by enterprise companies running In-memory processing applications. For Cluster connection, NetApp recommends using a dedicated RDMA over Ethernet (RoCEv2) 100Gbps network (also referred to as Persistent Memory over Fabrics, PMoF) which takes additionally 2-3 microseconds compare to single node configurations. MAX Snap is a snapshot on MAX Data (Tier 1) level which can call another MAX Data service called Sync Snap to synchronize snapshot on Tier 2 so snapshot on storage system will capture and merge data into coherent, consistent snapshot on the storage system which later can be replicated with SnapMirror. MAX Data have APIs for application consistent snapshots and plans to add functionality to integrate it with SnapCenter, but currently, doesn't have data protection software integrations. MAX Data have 3-month release cadence. MAX Data Memory API allows new applications to integrate with MAX Data for application-based and potentially more intellectual memory management, therefore needs application modification while MAX Data Tiering algorithm provided by default doesn't need any application modifications. MAX Data have Pin functionality which can pin a file or a certain range within a file so it will not be going to destaged into the second tier. According to the Yahoo, Cloud Serving Benchmark showed on Tech Field Day on 24 Oct 2018, MAX Data reaching 3.7 more IOPS and 4-5 times less latency, compared to Linux XFS on NVMe flash in AWS EC2 instance and 64GB memory (with 32GB cache size for MongoDB) with 50% read modify write workload. MAX data have a per-server license and does not depends on CPU, Memory or storage capacity. MAX Data has two different licensing tiers: Basic and Advanced.

Cloud Business

Cloud Central is web-based GUI interface which provides multi-cloud application orchestration layer and single pane of glass based on Qstack for NetApp's cloud products like Cloud Volumes Service, Cloud Sync, Cloud Insights, Cloud Volumes ONTAP, SaaS Backup in multiple public cloud providers. NetApp offers some of its original or modified products as part of its Cloud portfolio alongside with new cloud-native products. For example, Cloud Volumes ONTAP, NPS are re-purposed for use in the cloud, while others are cloud-native. Netapp Cloud business, offerings, and products available on http://cloud.netapp.com portal.

Cloud Volumes ONTAP

Formally ONTAP Cloud. Cloud Volumes ONTAP is software defined (SDS) version of ONTAP available in some public cloud providers like AWS, Azure, and IBM Cloud and named Cloud Volumes ONTAP. Cloud Volumes ONTAP is a virtual machine which is using commodity equipment and running ONTAP software as a service.

Cloud Volumes Service

Is service in Amazon AWS & Google GCP public cloud provider based on NetApp All-Flash FAS systems and ONTAP software, therefore will be able to synchronize data between cloud and on-premise NetApp systems. NetApp claims they will provide similar to AFF systems persistent performance SLA and also claims data could be moved in a way that will not generate huge ingress and egress charges using SnapMirror technology. Cloud Volumes Service will provide high availability for data with next protocols which can be consumed by containers: NFS v3, NFS v4, and SMB. Cloud Volumes Service support NetApp's Snapshots. Current maximum for Cloud Volumes Service is 100TB. In the Azure cloud currently only NFS available and called Azure NetApp Files. Cloud Volumes Service available as service in corresponding public cloud providers directly through their Marketplaces. NetApp claims them as reliable enterprise NAS services as in on-premises enterprise-grade data storage with predictable performance, management, security, and protection. NAS integrates with some other services in the public cloud providers. RESTful APIs will be available to automate NAS storage services such as provisioning, snapshots, and SnapMirror. As of January 2018 services for both public AWS and Azure in the preview stage.

NetApp Private Storage

NetApp Private Storage (NPS) is based on Equinix partner provided colocation service in its data centers for NetApp Storage Systems with 10 Gbit/s direct connection to public cloud providers like Azure and AWS etc. Some of Equinix data centers located in the same building with public cloud providers thus network connectivity to dedicated storage system is the same as with a storage service in a public cloud provider. Such configuration provide better performance compare to storage service in public cloud provider based on sharable commodity hardware and help to fit some companies with regulatory compliances which require strict data placement, security, availability and disaster recovery which public cloud provider could not provide. NPS storage could be connected to few cloud providers or on-premise infrastructure, thus in case of switching between clouds does not require data migration between them.

Kubernetes Service

NetApp Kubernetes Service (NKS) came from an acquisition of stackpoint.io it is a cloud-based SaaS application to deploy and manage Kubernetes clusters across cloud providers like AWS, Azure, Google Cloud, and NetApp HCI. Kubernetes Service used for: creating and managing kubernetes clusters, managing access control, setting up and managing kubernetes clusters across clouds, creating Helm charts and deploying them from a GitHub repository. Kubernetes Service has features like: load balancing, applications migration across kubernetes clusters, load spreading across various clusters, Istio Traffic Management, Persistent Volume with NetApp Trident plugin for Docker & Kubernetes upgrades and HTTP API for automation. NetApp allows upgrading to newest Kubernetes with about 5-day delay after release.

SaaS Backup

NetApp SaaS Backup (Previously Cloud Control) is back up and recovery service for SaaS Microsoft Office 365 and Salesforce which provide extended, granular and custom retention capabilities of backup and recovery process compare to native cloud backup. NetApp planning to extend SaaS Backup and recovery service for Google Apps, Slack and ServiceNow.

Cloud Sync

Cloud Sync is service for synchronizing any NAS storage system with another NAS storage, an Object Storage like Amazon S3 or NetApp Storage GRID using an object protocol.

Cloud Insights

SaaS application for monitoring infrastructure application stack for customers consuming cloud resources and also build for the dynamic nature of microservices and web-scale infrastructures. Cloud Insights uses similar to OnCommand Insight front-end API but different technology on the back-end. Cloud Insights available as a preview and will have three editions: Free, Standart and Pro. The free edition will be able to monitor NetApp storages & services using NetApp Active IQ with no cost.

NDAS

NetApp Data Availability Services with data protection in the cloud GUI. Cloud service currently located only in AWS but architecturally is cloud-agnostic and ready to be copied to other clouds like Google Cloud & Azure. NDAS is a target for backup, data protection and disaster recovery purposes from ONTAP storage. ONTAP systems starting with ONTAP 9.5 have a built-in proxy application which converting NetApp snapshots with WAFL metadata into the S3 format unlike FabricPool technology and can. Once data converted to S3 and delivered to the cloud, NDAS will discover newly transferred snapshots and create a searchable catalog of files which can be restored from the NDAS back to original or an alternate location. Currently, NDAS have only one working backup schema: primary ONTAP system with ONTAP 9.x transferring data to secondary ONTAP 9.5 with SnapMirror; the secondary system has NDAS proxy which converting snapshots and transferring them to S3 into the cloud. NetApp is planning to add NDAS to primary storage as well later on with a newer ONTAP version. NDAS is one of the Data Fabric manifestations.

Cloud Secure

Cloud Secure is a SaaS security tool that identifies malicious data access and compromised users. It provides centralized visibility and control of all corporate data access across hybrid cloud environments for security and compliance goals. Cloud Secure using data collectors to identify security breaches not solvable by file access control where the user has legitimate access to files and require post login detection to understand what is appropriate and what is bad behavior. Cloud Secure uses machine learning algorithms to identify the working community, abnormal user behavior and unusual patterns compare to other users. Cloud Secure can identify users been infected with ransomware generating abnormal change rate and prevent them from encrypting the files. Cloud Secure can take customizable actions to users identified with abnormal behavior, like quarantining the user.

Currently supported data repositories: NetApp Cloud Volumes, NetApp ONTAP, NetApp StorageGRID, OneDrive, AWS, Google Suite, HPE, DELLEMC Isilon, Dropbox, Box, @workspace and Office 365. Currently available as a preview only.

NetApp Cloud Orchestrator

Fabric Flows

MongoDB

Data Fabric

Often referred as "Data Fabric Story." Variety of integrations between NetApp's products and ways of data mobility considered by NetApp as Data Fabric vision for the future of data management. Data Fabric defines the NetApp technology architecture for hybrid cloud and provides the next features: SnapMirror replication from SolidFire to ONTAP; SnapMirror replication from ONTAP to Cloud Backup; FabricPool tiering feature for de-staging cold data from ONTAP to StorageGRID, Amazon S3 or Azure Blob; Volume Encryption with FabricPool provide secure data storage and secure over the wire transfer of enterprise data in a cloud provider; SnapMirror between FAS, AFF, ONTAP Select and Cloud Volumes ONTAP; Archiving and DR to public cloud; CloudMirror feature in StorageGRID replicates from on-premise object storage to Amazon S3 storage and triggers some actions in AWS Cloud; SolidFire backup to StorageGRID or Amazon S3; Cloud Backup archiving to variety of object storage systems (including StorageGRID) or many cloud providers; CloudSync is replication of NAS data to object format and back; replication to Cloud Volumes Service; Data backup to on-premise storage from SaaS Backup; SANtricity Cloud Connector for block-based backup, copy, and restore of E-Series volumes to an S3, NetApp Data Availability Services for data protection from ONTAP to cloud S3 storage with backup, DR & data mining capabilities, etc. Other features as part of Data Fabric vision approach committed by NetApp to be implemented in future, like SnapMirror between ONTAP and E-Series; SnapCenter capabilities to expand to include SolidFire storage endpoints, NetApp MAX Data host filesystem tiering to and from ONTAP storage systems connected with NVMeoF, etc.

Software Integrations

NetApp products could be integrated with a variety of software products which gives some additional flexibility, features and built-in provisioning and self-service storage capabilities. Most of the integrations are done for ONTAP systems.

Automation

NetApp provides a variety of automation services directly to its products with HTTP protocol or through middle-ware software.

Docker

NetApp Trident software provides a persistent volume plugin for Docker containers with both orchestrators Kubernetes and Swarm and supports ONTAP, SolidFire, E-Series, and in Cloud NetApp Kubernetes Service. Also, NetApp with Cisco sells CI architectures which incorporate Trident plugin: FlexPod Datacenter with Docker Enterprise Edition and ONTAP AI.

CI/CD

NetApp Jenkins Framework provides integration with ONTAP storage for DevOps, accelerating development with automation operations like provisioning and data-set cloning for test and development and leverage ONTAP for version control, create and delete checkpoints etc. Jenkins also integrate with NetApp Service Level Manager software which provides RESTful API for guarantee level of storage performance. Apprenda and CloudBees integrate and accelerate DevOps through Docker persistent volume plugin and Jenkins Framework integration. Apprenda could be integrated with OpenStack running on top of FlexPod.

Backup and Recovery

CommVault, Veeam and Veritas have integrations with ONTAP, SolidFire, Cloud Backup and E-Series leveraging storage capabilities like snapshots and cloning capabilities for testing backup copies and SnapMirror for Backup and Recovery (B&R), Disaster Recovery (DR) and Data Archiving for improving restore time and number of recovery points (see RPO/RTO). Cloud Backup integrates with nearly all B&R products for archiving capabilities since it is represented as ordinary NAS share for B&R software. Backup and recovery software from competitor vendors like IBM Spectrum Protect, EMC NetWorker, HP Data Protector, Dell vRanger, Acronis Backup etc. also have some level of integrations with NetApp storage systems.

Enterprise Applications

NetApp systems can integrate with enterprise applications for backup purposes, cloning, provisioning, and other self-service storage features. Oracle DB can be connected using Direct NFS (DNFS) client build inside database app which will provide network performance, resiliency, load balancing for NFS protocol with ONTAP systems. Oracle DB, Microsoft SQL, IBM DB2, MySQL, Mongo DB, SAP HANA, MS Exchange, VMware vSphere, Citrix Xen, KVM integrate with NetApp systems for provisioning, cloning and additional backup and recovery build in capabilities like SnapShots, SnapVault and SnapMirror with variety of B&R software including NetApp's SnapCenter and SnapCreator. Integration with such applications provides the ability to the instant cloning of data sets for test and development and self-service restore capabilities for application administrators and for end users. NetApp ONTAP systems could be integrated with VMware vSphere using NFS plugin for ESXi hosts which offloads some of the storage operations on the storage system. With VASA protocol NetApp systems can integrate for VMware VVol technology. Similarly to NFS plugin for ESXi, ONTAP systems support ODX functionality with Windows systems with file and block-based protocols for offloading data copying or moving processes to the storage system.

OpenStack

NetApp systems have integration with such open source projects as OpenStack Cinder for Block storage (SolidFire, ONTAP, E-Series, OnCommand Insight, Cloud Backup), OpenStack Manila for Shared file system (ONTAP, OnCommand Insight), Docker persistent volumes through Trident plugin (SolidFire, ONTAP, E-Series) and others.

OEM

IBM used to OEM NetApp FAS systems under its own brand known as IBM N-series and this partnership ended 29 May 2014. Dell OEM NetApp E-Series under its own name PowerVault MD. September 13, 2018, Lenovo and NetApp announced its technology partnership, so Lenovo OEM Netapp products under its own name: Lenovo ThinkSystem DE (using NetApp's EF & E-Series array technology), and ThinkSystem DM uses ONTAP software with Lenovo servers and supports FC-NVMe (analog for NetApp FAS & AFF systems).

Reception

Controversy

Syrian surveillance

In November 2011, during the 2011 Syrian uprising, NetApp was named as one of several companies whose products were being used in the Syrian government crackdown. The equipment was allegedly sold to the Syrians by an authorized NetApp reseller.^[29]

On April 7, 2014, NetApp was notified by the US Department of Commerce "that it had completed its review of this matter and determined that NetApp had not violated the U.S. export laws", and that the file on the matter had been closed.^[30]

Legal dispute with Sun Microsystems

In September 2007, NetApp started proceedings against Sun Microsystems, claiming that the ZFS File System developed by Sun infringed its patents.^[31] The following month, Sun announced plans to countersue based on alleged misuse by NetApp of Sun's own patented technology.^[32] Several of NetApp's patent claims were rejected on the basis of prior art after re-examination by the United States Patent and Trademark Office.^[33] On September 9, 2010, NetApp announced an agreement with Oracle Corporation (the new owner of Sun Microsystems) to dismiss the suits.^[34]

Accolades

NetApp was listed amongst Silicon Valley Top 25 Corporate Philanthropists in 2013.^[35]