Dell Fluid File System Overview A Dell Technology White Paper Dell Product Group January 2016
THIS TECHNOLOGY WHITE PAPER IS FOR INFORMATIONAL PURPOSES ONLY, AND MAY CONTAIN TYPOGRAPHICAL ERRORS AND TECHNICAL INACCURACIES. THE CONTENT IS PROVIDED AS IS, WITHOUT EXPRESS OR IMPLIED WARRANTIES OF ANY KIND. 2016 Dell Inc. All rights reserved. Reproduction of this material in any manner whatsoever without the express written permission of Dell Inc. is strictly forbidden. For more information, contact Dell. Dell, the DELL logo, and the DELL badge, EqualLogic, Compellent, PowerVault and NetVault are trademarks of Dell Inc. Microsoft and Windows are either trademarks or registered trademarks of Microsoft Corporation in the United States and/or other countries. Symantec, BackupExec, and NetBackup are trademarks of Symantec Corporation or its affiliates in the U.S. and other countries. CommVault and Simpana are registered trademarks of CommVault Systems, Inc. Other trademarks and trade names may be used in this document to refer to either the entities claiming the marks and names or their products. Dell Inc. disclaims any proprietary interest in trademarks and trade names other than its own. January 2016, version 5.0
Contents Abstract... 5 Dell Fluid File System Architecture... 6 Layered Approach to FluidFS Logical Architecture... 6 Network Load-Balancing Layer... 7 File Protocol Access... 7 NAS Volumes... 7 NAS Pool... 8 The Physical Building Blocks... 8 NAS Appliance... 9 Back-end SAN Storage... 9 Data Integrity... 10 Cache Mirroring and Metadata... 10 High availability... 10 Data Protection... 11 Snapshots... 11 Backups... 11 Replication... 12 Snapshot Retention Policies... 12 Data Efficiency... 13 Data Deduplication... 13 Compression... 13 Thin Provisioning... 14 Thin Clones... 14 Data Management... 15 NAS Volumes... 15 Quotas... 15 File Exclusion Filters... 15 Security... 16 Authentication and Authorization... 16 Network Isolation of data... 16 User Auditing... 16 Data Governance... 16 Management and Monitoring... 17 Summary... 18
Abstract Traditional approaches to handling file data growth have proven to be costly, hard to manage, and difficult to scale effectively and efficiently. Dell Fluid File System (FluidFS) is designed to go beyond the limitations of traditional file systems with a flexible architecture that enables organizations to scale out and scale up non-disruptively. Hence, it addresses organizational challenges by allowing them to gain control of their data, reduce complexity, and meet growing data demands over time at the same time reducing overall cost of deploying and managing storage infrastructure. The FluidFS architecture is an open-standards based scale-out network attached storage (NAS) file system that supports industry standard protocols such as SMB and NFS. It provides innovative features proving high availability, high performance, efficient data management, data integrity, and data protection. Combined with Dell SC series storage, the solution provides a single unified storage solution for both block and file storage needs. With ability to independently scale capacity and performance and all inclusive enterprise class features, FluidFS provide a great storage solution for a multitude of use cases such as: General purpose file sharing for projects, departments and users in enterprises Visual Effects Rendering, post production workflows, media asset management and archival Health care and life sciences workflows Virtual Machine storage for VMware environments Video Surveillance Video on demand applications Backup and archive target, etc. Continue reading to learn more about FluidFS architecture and features that make it a unique offering in the market and a potential fit to provide file storage for your applications. 5
Dell Fluid File System Architecture The relentless growth of unstructured file data is accelerating the need for network file storage systems. Organizations coping with data growth are confronted with several challenges: Data silos prevent easy access to vital business information. Data migration, backup, and disaster recovery are complex, consuming administrative time and resources. Meeting data growth by deploying more and more storage systems increases both the administrative burden and capital expenditure at a time when businesses need to run lean. Traditional file systems have scalability limitations that make them unwieldy for organizations with rapidly expanding file data. FluidFS is an enterprise-class, fully distributed file system that provides customers with the tools necessary to manage file data in an efficient and simple manner. The underlying software architecture leverages a symmetric clustering model with distributed metadata, native load balancing, advanced caching capabilities and a rich set of enterprise-class features. FluidFS removes the scalability limitations such as limited volume size associated with traditional file systems and supports high capacity, performance-intensive workloads via scaling up (adding capacity to the system) and by scaling out (adding nodes, or performance, to the system). A FluidFS storage solution can scale to 4PB of capacity in a single file system, delivering up to 12GB/s aggregate throughput. To grow namespace beyond maximum size of a single file system, multiple FluidFS systems can be combined in a single FluidFS Global Namespace that can present 10s of petabytes in a single SMB or NFS namespace. Any existing NAS which supports SMB 2 or higher or NFS v4 protocols can also be joined to a FluidFS Global Namespace to simplify data access within or across sites. FluidFS operates across a symmetric cluster of purpose-built NAS controllers (housed in pairs within a 2U appliance), which interface over a fabric to shared back-end Dell Storage Center (SC) series arrays. The following sections provide descriptions of the logical and physical layers of FluidFS. Layered Approach to FluidFS Logical Architecture The FluidFS architecture approaches layering with performance in mind. The layered architecture presents a traditional file system to network clients while performing a range of special functions at the back end. The specific goal of this design is to utilize all available resources at the network, server and disk levels to support the fastest possible response times. The logical layers that comprise a FluidFS system and influence system performance are described below. Network Load-Balancing SMB Shares or NFS Exports NAS Volumes NAS Pool Block Device LUNs 6
Network Load-Balancing Layer FluidFS presents multiple network ports from multiple controllers to the client network. To simplify access and administration, FluidFS supports Virtual IP addresses and physical distribution of client sessions is managed by a built-in load balancing capability. The load is balanced between the interfaces within a controller using one of two industry standard algorithms Adaptive Load Balancing (ALB) or Link Aggregation Protocol (LACP). ALB is a MAC address-based balancing mechanism that does not require any configuration on the network switch. LACP, which is also known as 802.3ad, is another available option and is supported by most major manufacturers managed switches; some configuration is required on the switch. For client access within the same layer 2 network, FluidFS balances client connections across all available network interfaces within the cluster based on client MAC addresses. When load balancing across a routed network boundary, FluidFS utilizes DNS Round Robin in addition to the MAC-based address lookup. Each NAS controller can access and serve all data stored in the FluidFS system. Figure 1 shows load balancing in a Fluid File System cluster on a single layer 2 network. File Protocol Access Figure 1 - Load Balancing in a FluidFS Cluster FluidFS supports SMB v2, v2.1 and v3 1 and NFS v3, v4 and NFS v4.1 2 and FTP protocols. All shares and exports configured by the system administrator are available from all NAS controllers at all times in a manner that is completely transparent to the client. NAS Volumes NAS volumes are virtual entities that provide policy-based management of snapshots, replication, quotas, deduplication, backup and security style as appropriate for different workloads. NAS volumes can be created on the fly and can shrink or grow, non-disruptively, in capacity up to the physical limits of the back-end storage. Figure 2 illustrates NAS volumes provided and managed by FluidFS. Each NAS volume can have one or more SMB shares or NFS exports. The policy associated on a NAS volume, for example a snapshot policy, is applicable for all shares and exports within a volume. 1 Not all SMB 3 features are supported. 2 FluidFS does not support pnfs protocol. Refer to FluidFS Support Matrix for details on protocol level feature support. 7
Figure 2 Logical Constructs of a FluidFS Cluster NAS Pool At its core, FluidFS creates a single distributed file system which spans across all SAN block device LUNs that are presented to the system. This layer, represented as the NAS Pool, provides file system services to the NAS volumes defined by the administrator in an optimized manner. All NAS controllers actively participate in the FluidFS distributed file system and can write or read any data housed within the file system. FluidFS utilizes Dell Storage Area Network (SAN) to storage user data and system configuration. Volumes or LUNs on the SAN are automatically provisioned to FluidFS based on the NAS pool size to be configured. Automated configuration saves valuable administrator time spent on routine tasks such as provision and managing LUNs. Figure 2 illustrates how multiple LUNs are deployed and aggregated by FluidFS to present a single virtualized NAS pool for the administrator to create NAS volumes and shares. The Physical Building Blocks FluidFS interacts with two main hardware elements: Up to four FS8600 NAS gateway appliance(s) Up to eight Dell SC Series SAN storage fabric and array as back-end The minimum hardware configuration of a FluidFS system is a single FS8600 NAS appliance and a single SC Series SAN. This configuration can scale capacity by adding additional disks to the SAN and scale performance by adding additional NAS appliances or SAN storage controllers. The NAS system can scale capacity and performance independently and online, without disrupting system availability. 8
NAS Appliance Two active-active, hot-swappable NAS controllers are housed side by side within a 2U FluidFS NAS appliance. A 40Gb/s mid-plane between the two controllers enables cache mirroring and write-back operations. Each controller in a NAS appliance contains 24GB to 48GB of battery protected RAM, a large majority of which is used to cache reads and writes. Internal hard drives and backup batteries protect metadata and ensure data integrity. The appliance is connected to SC Series SAN via 8Gb Fibre Channel or 10GbE iscsi. The NAS cluster can scale out to up to 4 FS8600 NAS appliances or 8 controllers to match to client workload characteristics and business needs. To achieve data distribution and maintain high availability, each appliance in a cluster has access to all other appliances in the cluster through a dedicated and redundant interconnect network. Each FluidFS NAS controller can also read from and write to any file in any file system. Figure 3 represents the logical architecture of a FluidFS cluster with 2 NAS appliances. Figure 3 A Dual NAS Appliance Cluster The number of NAS appliances clustered together in a FluidFS system dictates the aggregate compute, cache and client network bandwidth available on the system. However, proper system sizing for performance must also consider the performance of the SAN back-end. Back-end SAN Storage FluidFS is integrated with Dell s SC series SAN offerings to provide a single unified storage solution. One hundred percent of the SAN capacity can be assigned for NAS-only access or the capacity can be split across NAS and application hosts requiring block storage access via iscsi, FC o FCoE protocols. 9
Data Integrity FluidFS incorporates a series of mechanisms to provide a high level of integrity and system resiliency for data at rest and in transit. This section details those mechanisms. Cache Mirroring and Metadata During normal operation, the write cache, which includes the data and the metadata, is mirrored between the controller pairs in the FluidFS cluster. Additionally, important metadata is journaled to back-end storage capacity by a journaling process that runs continuously. This ensures file system consistency, in the event of a failure. When a controller fails, cache mirroring is not possible, so the surviving controller immediately journals its data and metadata to the storage presented by the back-end storage subsystem. This ensures that the file system remains consistent and that all data is protected in case of additional failures. In the event of a power outage, the write cache is journaled to a temporary staging location within the controller. This ensures that all I/O in flight remains consistent and prevents data loss. In addition, this ensures data consistency is not compromised, irrespective of the duration of the outage, as seen by traditional battery backed systems, which will lose data if power is lost before the write completes. These mechanisms provide resiliency and redundancy for a multitude of failure scenarios to ensure that data is always intact and accessible. High availability In a FluidFS cluster, any single controller can fail without affecting data availability or causing data loss even if write operations were in flight. Cross-cluster reliability is achieved through a variety of mechanisms, including a high speed cluster interconnect, write cache mirroring, failsafe journaling, and data integrity checks to ensure data store consistency. FluidFS monitors the health of the NAS appliance, including temperature and power conditions, to ensure cluster reliability and maximize data availability in cases of hardware or software failures. If failures occur, hardware components in the storage subsystem are redundant and hot- swappable. Each controller receives its power from the power grid and a dedicated backup power supply (BPS), which is regularly monitored to ensure that the BPS maintains a minimum level of power for normal operation. The BPS has sufficient battery power to allow the controllers to execute their shutdown procedures and use the cache as NVRAM. The BPS also provides enough time to write all the data from the cache to disk in the event of a loss of power. 10
Data Protection FluidFS enables data protection within a single system, across systems, and to external NAS repositories. This section will discuss some of the features and benefits of using FluidFS to store and protect data long-term, and how customers can govern the use and contents of their FluidFS systems. Snapshots Snapshots provide the first level of data protection by providing the ability to recover data instantly and are an integral part of FluidFS. Each NAS volume has its own snapshot policy to allow flexibility and to optimize space management. The administrator has the ability to schedule point-in-time, automatic snapshots as frequently as every five minutes and can initiate manual snapshots at any time. FluidFS incorporates redirect-on-write snapshots, instead of the copy-on-write solutions typical of other file systems. Redirect-on-write requires only one I/O operation, thereby preventing performance degradation. Figure 4 shows the redirect-on-write mechanism and the different stages of the snapshot process. Figure 4 - Redirect-on-Write Snapshot Process Snapshots are available to users as a read-only copy of the file system, which allows them to restore their documents in a simple manner without administrator intervention. Administrators also can easily restore very large data sets (terabyte scale) as a whole to a particular point in time. This eliminates long file copies and the need for free space for the recovery process. Backups The Dell Fluid File System supports standard backup software using Network Data Management Protocol (NDMP) version 4 with no changes required to existing backup workflows. FluidFS supports both 3-way or remote NDMP and 2-way or direct NDMP over a Fibre Channel fabric. Dell has partnered with industry leaders to provide comprehensive backup solutions that integrate with Fluid File System. Currently supported backup software includes Dell Quest NetVault, CommVault Simpana, Symantec NetBackup and BackupExec and IBM Tivoli Storage Manager. Please refer to the product specification sheets for the latest information on backup software supported with FluidFS products. 11
Replication FluidFS allows fast and reliable snapshot-based replication of any number of volumes to a partner. After the initial synchronization, only incremental changes are replicated, which improves network bandwidth utilization. This replication is native to FluidFS and does not require any additional hardware. The data is always consistent on the partner site and available as read-only. A read/write thin clone of the replicated volume can be created on partner site to test data availability at partner site or to provide read/write access to a point in time copy of the data from primary site. In addition to user data, NAS configuration (such as configured volumes, shares, exports, etc.) are replicated. This reduces administrative burden and enables continuous access to data in the event of a disaster or site failure to assure business continuity. Replication is bi-directional, meaning that the same system can host both source and destination volumes. In addition, the direction can be reversed without requiring a full resynchronization. FluidFS also supports one-to-many and many-to-one replication between NAS systems using unique volumes. Figure 5 shows the supported replication options. Figure 5 - FluidFS Replication Snapshot Retention Policies When a NAS volume is replicated to a partner site, snapshots of the NAS volume can be replicated as well. FluidFS provides a set of flexible policies that govern replication and retention of snapshots. Following policies are available: 1. No History: if chosen for a NAS volume replication, the partner site only keeps the last snapshot that is required for the replication. All previous snapshots of the NAS volume that may existing on the primary site are not stored on the partner site. This policy is helpful if there are capacity constraints on the partner site or the goal is to just keep the latest consistent copy of the data. 2. Exact copy: as the name suggests, if selected for a NAS volume replication, the snapshots available on primary site are the same as ones available on the partner site. In this case, the partner site is an exact point in time copy of the primary site. 3. Archive: the archive policy enables administrators to archive snapshots of the NAS volume on partner site. Using this policy, a snapshot can be retained on the partner site for a longer period of time than on the primary site. The primary site can have fewer number of snapshots for instant local recovery, however older snapshots can be retained on the partner site should a need arise to recover data from an archive. 12
Data Efficiency With a constant pressure to reduce costs, while still meeting business needs, it s imperative to maximum resource utilization of the storage infrastructure, hence maximizing the return on investment. FluidFS includes a number of features such as Data Reduction (Deduplication and Compression) and Thin Provisioning to efficiently store data on Dell storage. Data Deduplication FluidFS deduplication utilizes a post-write process architecture. Files are written to the FluidFS appliance in their native form and then deduplicated once they meet a pre-specified set of reduction criteria. The administrator can set deduplication policies for each NAS volume based on the last time a file was accessed or modified, and can also schedule deduplication activity for a certain time window. For instance, a data set that will be rarely or never read or modified again, such as backup, can be set to be immediately reduced, and to do so after hours. Hot data that will typically experience many writes, such as active files in user home shares, are not reduced and hence suffer no performance impact. When a deduped file is modified, FluidFS will only rehydrate the portion of the file with write activity, minimizing I/O on the storage system. Fluid Data Reduction uses a variable-block sliding-window strategy. It crawls each file, looking for matching chunks of data sized between 64KB and 192KB with a general goal of 128KB. The reduction engine then eliminates redundancies, creates an object block map and stores that information back on the NAS volume. Under normal circumstances, FluidFS will not optimize data that has not been accessed or modified in fewer than 5 days. However, there will be times where data will rarely be accessed after it is written to the file system. This is the case with backup/archive data or large image stores as one might find in a PACS system. For these instances, NAS volumes can be set in archive mode that allows the NAS administrator to set a policy that enable data deduplication immediately, without waiting for the minimum active data policy to run. Compression Compression is the second sub feature of the core data reduction capabilities of FluidFS. Compression of files in FluidFS is accomplished using the LZPS algorithm. This was developed by Dell and uses less CPU overhead than traditional compression algorithms. This cuts down on the overhead associated with reading/writing small files. Any NAS volume can be configured for only deduplication or deduplication plus compression. The NAS administrator can schedule when to run data reduction and on which NAS volumes. As more NAS appliances or controller pairs are added to a NAS cluster and as the new controllers own more data, not only does the NAS cluster scale in file performance, but also scales in the speed at which data can be reduced. 13
Once the data reduction job starts, all files meeting the data reduction criteria are processed for data reduction. Since only the files that are not actively used are reduced, FluidFS does not incur any performance impact for active data. For read access to data that has been reduced, the blocks requested by the client are rehydrated or changed to native form on the fly by FluidFS and passed to the client. Due to rehydration, read performance can be impacted. Reads do not alter how the blocks are stored on the disk- reduced blocks remain in reduced state. Any blocks within reduced data that are modified are stored in native form. When such files with modified blocks meet the data reduction policies, the modified blocks will be again reduced and stored in reduced form. Thin Provisioning Storage administrators often have to balance the end user capacity demand, usage growth pattern which often starts slow and grows over time and the budget limitations. When buying a new storage system, often, administrators buy more capacity than is actually needed, resulting in low return on the initial investment. FluidFS takes advantage of thin provisioning so that administrators can over provision the NAS capacity visible to end users, and add physical capacity only when needed. The NAS pool is thin provisioned by default. In addition, the NAS volumes are also, by default, thin provisioned providing another layer of flexibility to manage capacity growth. Capacity utilization alerts can be set up to warn administrators of consumed and/or free physical capacity so that when needed, with no disruption to end users, additional disk capacity can be added to the NAS pool. Thin Clones For use cases such as test and development, creating demo or lab environments or deploying an application from a template data set, there is need to create a copy of the data. Creating a full copy of the data consumes unnecessary capacity on the storage system as well as takes time if data set size if huge. To help making data set copies more efficient, FluidFS include an ability to create thin clones of NAS volumes or individual files. A thin clone of NAS volume is created using an underlying snapshot of a parent volume. A thin clone is just like another NAS volume, however with little capacity overhead. Only when new data is written to a clone NAS volume, additional capacity is consumed for the new data. This gives administrators a powerful tool to instantaneously create a copy of the data without worrying about the size of data set. Similarly, a clone of a file can be created to create a thin copy of the file. Thin clones are extremely helpful in environments such as NFS datastores for VMware, deploying multiple VMs from a single golden copy or creating a R/W data set of replicated data to test data availability at disaster recovery site. 14
Data Management FluidFS incorporates multiple design features and functionality to make managing data simpler on Dell storage. This section describes some of key data management features of FluidFS NAS Volumes As explained earlier, NAS volumes are flexible virtual entities that help segregate user data under various policy domains that govern data protection, data optimization, quota limits settings among others. In a scenario where multiple departments or user groups have varying needs or service levels for example how often to create snapshots, if and how often to replicate, whether to enable dedupe and compression, etc., a NAS volume can be created per department. Since the NAS volumes grow or shrink in size nondisruptively, these provide an extremely flexible method for administrators to carve data for varying user needs. Quotas With growing need of storing unstructured data, storage administrators also have to control the usage per user, group or department, so that a single user or group does not consume more capacity that needed and hence fair capacity is available to all shared users. To control the capacity usage, FluidFS includes setting quotas policies. Following quota policies can be created: User and Group Quotas control maximum capacity a user or a Group can consume on a NAS volume. Such quota policies are defined for each NAS volume and apply to all data written to the shares or exports in the NAS volume. Directory Quotas are defined for a directory within a NAS volume and apply to all data written to the directory tree starting at the specified directory. Directory quotas are very helpful in controlling departmental shares and monitoring capacity usage over time. File Exclusion Filters Often, administrators are faced with a challenge of controlling what kind of data gets written to the file shares. It may be the case that no video files should be stored on a document share, but it has not been easy to control what files gets stored on a shared folder. With FluidFS, flex extension filters can be created for an SMB share that disallows users to store files with specified extensions. Administrators can create policies that do not allow users to store files with say.mp3 or.exe extensions. In such case, users get an error when they try to store files that match an extension specified in the exclusion filter. 15
Security Authentication and Authorization Providing access to only authorized users and limiting access based on a user s permissions is one of the key facets of ensuring a secure file storage solution. FluidFS fully integrates with Microsoft Active Directory, LDAP and NIS user repositories to authorize users who can access resources on FluidFS storage. If no external directory service is available, local users can be created to share data. File permissions on FluidFS are stored in their native format and both Windows ACLs, NFS v3 POSIX permissions and NFS v4 ACLs are supported. For environments where data needs to be shared across SMB and NFS protocols, user mapping can be set automatically or manually to ensure secure sharing of data in a multi-protocol environment. With support for NFS v4 Kerberos authentication, administrators can ensure data is shared securely among NFS users by ensuring identify of the user as well as option use of encryption to prevent data access by an unwanted entity. Network Isolation of data In some environments, it is desirable or required to ensure that access to a network share is only limited to users with a network segment or a subnet. In an organization, clients in different departments or offices could be on different network segments, hence controlling access at the network layer is another easier method to control access to data. FluidFS includes the capability to create rules that limit visibility and access to shares and exports within a NAS volume to a given set of network subnets. Clients trying to access from an unauthorized network subnet do not see the protected shares nor can they access any data. User Auditing In case of unauthorized access, modification or deletion of data, it is critical to find out the user who had unauthorized access and the operations performed on the data in question. This is especially true in environments where data stored needs to meet legal compliance requirements such as from HIPPA, etc. To audit user actions, FluidFS includes support for SMB System Access Lists (SACL). When enabled and configured, an audit log for configured files and/or folders is stored on the FluidFS appliance. User actions such as successful or failed access, file operations such as reads, writes, deletions, etc. are logged in the SACL audit log which can be viewed and exported via Enterprise Manager and FluidFS CLI. Data Governance FluidFS also supports external governance tools, such as Dell Quest Change Auditor. These tools provide a wide range of security management components, and file-access is a major part in the overall security policy. File and folder access requests from SMB clients will generate a notification message that is forwarded to Change Auditor. There, the message is logged, and can generate further activities based on policies, including forwarding email, SNMP, etc. As a result of this integration, administrators can have a single pane of glass to track and audit file access throughout their organization. In future, Dell will support more external auditing tools including support for auditing NFS exports. 16
Management and Monitoring Dell Storage Enterprise Manager simplifies administration of Dell unified storage environments by providing comprehensive monitoring and management of all local and remote storage systems. Dayto-day storage management time can be minimized using a single interface that provides a complete view of both SC Series SAN and FS8600 NAS storage environment, streamlining administration and reducing operational costs. Enterprise Manager simplifies deployment of NAS pool by automatically creating volumes needed to deploy the NAS pool of configured capacity. As the NAS pool is expanded, Dell Storage Manager expands existing volumes or creates new volumes to support NAS capacity expansion. Routine NAS tasks such as managing NAS volumes, SMB shares and exports as well as setting share level permissions for SMB shares can be accomplished using Enterprise Manager console. System firmware upgrades can also be performed online using the management application. With Enterprise Manager, one can reduce costs by maximizing utilization and purchase storage more efficiently using accurate capacity and performance monitoring. Administrators can view current and historical NAS cluster and controller level performance including protocol level performance data such as NFS and SMB throughput and file operations to latency and throughput on disks that make up the NAS pool. Data protection tasks such as creating automated snapshots and replication schedules can be configured in a few clicks. Quotas and home shares can also be managed with just a few clicks. All system events for both SAN and NAS can be viewed in a single interface. SACL Audit events can also be viewed and exported from Enterprise Manager console. Additionally, NAS administrators can also view open files and current sessions and clients connected to SMB shares and NFS exports. If needed, sessions can be terminated by the administrator or a clients can be moved to specific NAS controllers. Information such as list of shares or exports, system events, audit events, connected clients, open sessions, etc. can be exported to reports in various formats such as PDF, HTML, XML, etc. Figure 6: Dell Storage Enterprise Manager To integrate management operations with automation tools for the entire IT environment, REST API and PowerShell extensions are available. These allow one to automate provisioning operations when needed. With one single tool to manage both SAN and NAS, intuitive interface and workflows, storage administrators will find it simple to manage small to large storage deployments in a single or multiple sites. 17
Summary Dell s unified storage solutions utilizing FluidFS to provide scale-out NAS capability enable customers to deploy a single solution to meet both SAN and NAS storage needs. Without complex and expensive licensing, customers can benefit from all enterprise class features such as snapshots, replication, thin provisioning, dedupe and compression, auditing, etc., Without locking customers into a scale-up or a scale-out only solution, FluidFS can scale-out or scale-up, thus allowing independent scale of performance and/or capacity. This flexibility enables customers to architect a NAS solution to meet a workload s performance and capacity requirements, hence reduces overall storage cost and solution complexity. 18