celerra foundations srg

Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Celerra Foundations Student Resource Guide - 1

© 2007 EMC Corporation. All rights reserved.

Celerra FoundationsCelerra Foundations

Welcome to Celerra Foundations.

These materials may not be copied without EMC's written consent.

EMC believes the information in this publication is accurate as of its publication date. The information is subject to change without notice.

THE INFORMATION IN THIS PUBLICATION IS PROVIDED “AS IS.” EMC CORPORATION MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.

EMC is a registered trademark of EMC Corporation.

All other trademarks used herein are the property of their respective owners.



© 2007 EMC Corporation. All rights reserved. Celerra Foundations Student Resource Guide - 2

Revision History

CompleteMay, 20074.0

CompleteOctober, 20063.5

CompleteMarch, 20063.0

RevisionsCourse DateRev Number

Copyright © 2007 EMC Corporation. All rights reserved.

These materials may not be copied without EMC's written consent.

EMC believes the information in this publication is accurate as of its publication date. The information is subject to change without notice.

THE INFORMATION IN THIS PUBLICATION IS PROVIDED “AS IS.” EMC CORPORATION MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.

EMC and Celerra are trademarks of EMC Corporation.

All other trademarks used herein are the property of their respective owners.




Celerra Foundations

Upon completion of this module, you will be able to:

Identify the concepts, architecture, terminology, and environmental aspects of NAS using the Celerra

Describe the Celerra’s theory of operation and Data Mover configuration options

Identify and describe Celerra features, functions, and Management software offerings

Identify and describe the different Celerra Business Continuity options and Backup solutions

Identify EMC Celerra platforms and their differences

Identify the benefits provided by a Celerra solution

The objectives for this module are shown here. Please take a moment to read them.




Celerra Architecture

Upon completion of this lesson, you will be able to use the Celerra to identify the:

Concepts

Architecture

Terminology, and

environmental aspects of NAS

In this lesson, we look at the Celerra architecture and terminology.




The CelerraServes as a dedicated, high-performance, high-speed communication of data using both file and block level IOs

Sometimes called a file server

Uses one or more Network Interface Cards (NICs) to connect to the customer network

Uses proprietary optimized operating system; DART, Data Access in Real Time

Uses industry standard storage protocols to connect to storage resources Disk

Storage

Celerra

Network Drivers and Protocols

NFS CIFS

NAS Device OS (DART)

Storage Drivers and Protocols

IPNetwork

Client Application

The Celerra uses standard TCP/IP networking to enable application servers to access a centrally managed, protected, and highly available storage system. The Celerra is not considered a general-purpose computer, because it uses a significantly streamlined/tuned OS. It is sometimes called a file server because it focuses all of its processing power solely on file service and file storage. Common network interface cards (NICs) include gigabit Ethernet (1000 Mb/s) or Fast Ethernet (10Mb/s), ATM, and FDDI. Most Celerra models also support NDMP (Network Data Management Protocol) for backup, Novell Netware, FTP and HTTP protocols.

The NAS operating system for EMC Celerra is “DART” - Data Access in Real Time. These operating systems are tuned to perform file operations including open, close, read, write, etc. The Celerra generally uses a standard drive protocol, some form of SCSI, to manage data to and from the disk resources.




Architecture

• NFS and CIFS handle file requests to remote file system

• I/O is encapsulated by TCP/IP Stack to move over the network

• Celerra converts requests between file and block IO to read orwrite data to disk storage

Storage InterfaceStorage Interface

Storage ProtocolStorage Protocol

NAS Operating SystemNAS Operating System

NFS / CIFSNFS / CIFS

TCP/IP StackTCP/IP Stack

Network InterfaceNetwork Interface

ApplicationApplication

Operating SystemOperating System

I/O RedirectI/O Redirect

NFS / CIFSNFS / CIFS

TCP/IP StackTCP/IP Stack

Network InterfaceNetwork Interface

IP Network

Celerra

Block I/O to storage device

Client

The Network File System (NFS) protocol and Common Internet File System (CIFS) protocol handle file I/O requests to the remote file system, which is located in the NAS device storage. I/O requests are packaged by the initiator into the TCP/IP protocols to move across the IP network. The remote Celerra file system converts the request to block I/O and reads or writes the data to the disk storage. To return data to the requesting client application, the Celerra appliance software re-packages the data to move it back across the network.

Here we see an example of an I/O being directed to the remote Celerra and the different protocols that play a part in moving the request back and forth to the remote file system located on the Celerra.




Network Protocols

Network transport Protocols– Universal Datagram Protocol (UDP) for non-connection

oriented networks– Transmission Control Protocol (TCP) for connection

oriented networks

Network file system Protocols– NFS manages files in a networked Unix environment– CIFS manages files in a networked Windows environment

Network transport protocols are standards that allow computers to communicate. A protocol defines how computers identify one another on a network, the form that the data should take in transit, and how this information is processed once it reaches its final destination. TCP/IP is a common protocol used in sending information via the Internet. Protocols also define procedures for handling lost or damaged transmissions, or "packets”. UDP and TCP are examples of transport protocols.

In a non-connection oriented communication model (UDP), the data is sent out to a recipient using a best effort approach with no acknowledgement of the receipt of the data being sent back to the originator by the recipient. Error correction and resend must be controlled by a higher layer application to ensure data integrity.

In a connection-oriented model (TCP), all data packets sent by an originator are acknowledged by the recipient and transmission errors. Lost data packets are managed at the protocol layer.

Network file system protocols are used to manage how data requests are processed once it reaches its final destination. Both NFS and CIFS support UDP and TCP transport protocols.




Environment UNIX and Windows ClientsGroup Name = SALES

UNIX and Windows ClientsGroup Name = Accounting

Router

DNS Server

Storage Array

Celerra

DHCP Server

NIS Server

Switch

Switch

This slide shows a scaled-down example of what a customer environment might look like. It consists of switches, which can link multiple network connections together, and routers, which are responsible for the passing of traffic between networks. Routers also divide networks logically instead of physically. An IP router can divide a network into various subnets so that only traffic destined for particular IP addresses can pass between segments.Several things must happen in order for computers to be able to communicate data across the network. First, the computer must have a unique network address, referred to as the IP Address. An address can be assigned in one of two ways; dynamically or statically. A static address requires entering the IP address that the computer uses in a local file. However, if two computers on the same subnet are assigned the same IP address, they would not be able to communicate. Another approach is to set up a computer on the network to dynamically assign an IP address to a host when it joins the network. This is called the Dynamic Host Configuration Protocol (DHCP) Server. A second requirement for communication is to know the address of the recipient of the communication. The most common approach is to communicate by name, for example, the name you place on a letter. However, the network uses numerical addresses. A more efficient solution is the Domain Name Service (DNS). The DNS is a hierarchical database, which resolves host names to IP addresses.The Network Information Service (NIS) is used by UNIX clients for host name resolution. It has similar functions as the DNS server. If for any reason the NIS server is down, UNIX clients then contact its host file and even the DNS server depending on the order found in its host configuration file.




Celerra Software Operating SystemsEMC Linux

This is an industry-hardened and EMC-modified Operating System loaded on the Control Station to provide:

Secure NAS management environmentGrowing in popularity and corporate acceptance

DART – Data Access in Real TimeThis is a highly specialized Operating System designed to optimize network traffic Input/Output throughput and is loaded on the Data MoversIs multi-threaded to optimize load balancing capabilities of the multi-processor Data MoversAdvanced volume management - UxFS

Large file size and filesystem supportAbility to extend filesystems onlineMetadata logging for fast recoveryStriped volume support

Feature rich to support the varied specialized capabilities of the Celerra rangeData Mover FailoverNetworking functionality – Port Aggregation, FailSafe Network device, multi-protocol supportPoint in time file system copiesWindows environmental specialties

EMC Linux OS is installed on the Control Station. Control Station OS software is used to install, manage, and configure the Data Movers, monitor the environmental conditions and performance of all components, and implement the Call Home and dial-in support feature. Typical Administration functions include volume and file system management, configuration of network interfaces, creation and exporting of file systems to clients, performing file system consistency checks, and extending file systems.

The OS that the Data Movers run is EMC’s Data Access in Real Time (DART) embedded system software, which is optimized for file I/O, to move data from the EMC storage array to the network. DART supports standard network and file access protocols: NFS, CIFS, and FTP.




Control Station

A dedicated management, Intel processor-based computer running EMC Linux which provides:

Specialized software installation and upgrade portal

Management of high availability features– Fault monitoring– Fault recovery– Fault Reporting (CallHome)

Management of Data Mover configuration and storage for the system configuration database

Remote diagnosis and repair

The Control Station is a dedicated management Intel processor-based computer that monitors and sends commands to the Data Movers. The private network connects the Control Station to the Data Movers through the system management switch module.

Some Celerras support two Control Stations for redundancy and continued management of the system. However if a Control Station fails without a standby system, all operational Data Movers continues to function without an impact to their clients. Only configuration and management tasks are affected until the Control Station is repaired.

Like previous versions, the Control Station provides software installation and upgrade services, and high-availability features such as fault monitoring, fault recovery, fault reporting (CallHome), and remote diagnosing. Two Control Stations can be connected to a public or private network for remote administration. Each Control Station has a serial port that connects to an external modem so that the Control Station can call home to EMC or a service provider if a problem should arise.




Control Station (cont)

The control station provides an interface to control, manage, and configure the NAS solution

Control Station provides the controlling subsystem of the Celerra, as well as the management interface to all file server components. It provides a secure user interface as a single point of administration and management for the whole Celerra solution. Control Station administrative functions are accessible via the local console, Telnet (not recommended), or a web browser.

The Control Station is a single Intel processor based, with high memory capacity. Dependent on the model, the Control Stations may have internal storage. The local LAN switch provides the internal communications network for the Data Movers and the Control Station and should not be integrated into a client networking infrastructure.

Within the NSX model there are no serial interconnections between the Control Station and the Data Movers and the internal switch has been built into the Control Station functionality.




Data Mover

An independent, autonomous file server that transfers requested files to clients

Hot pluggable

Standby Data Movers offer redundancy

Unaffected by Control Station failure

Fibre I/O module GbE I/O moduleNSX/NS80G Blade (DM)

Each Data Mover is an independent, autonomous file server that transfers requested files to clients and are managed as a single entity. Data Movers are hot pluggable and can be configured with standbys to provide unaffected service should a problem arise with a primary Data Mover. A Data Mover (DM) connects to a LAN through FastEthernet, Gigabit Ethernet, or Fibre Channel ports.

The default name for a Data Mover is “server n”, where n was its original slot location in the first NAS frames. This has been continued into the new frames and the naming convention remains slot-related. For example, in the Golden Eagle/ Eagle frame, a Data Mover can be in slot location 2 through 15 (i.e. server_2 - server_15), therefore the first Data Mover is any frame remains server_2, the second server_3, etc.

There is no remote login capability on the Data Mover. They do not run any binaries and all access to the Data Mover for management and configuration must be performed via the Control Station.




Celerra Theory of Operations

Upon completion of this lesson, you will be able to:

Describe the Celerra’s theory of operation and Data Mover configuration options

In this lesson, we examine how the Celerra works.




NAS OperationsTraditional IO operations use file level IO protocols

File system is mounted remotely using a network file access protocol, such as:– Network File System (NFS)– Common Internet File System (CIFS)

IO is redirected to remote system

Utilizes mature data transport (e.g., TCP/IP) and media access protocols

Celerra assumes responsibility for organizing data (R/W) on disk and managing cache

Disk

IP Network

Application

Celerra

SANORDirect

Attach

One of the key differences of a NAS disk device, like the Celerra, compared to DAS or other networked storage solutions such as SAN, is that all traditional I/O operations use file level I/O protocols. File I/O is a high level type of request that, in essence, specifies only the file to be accessed, but does not directly address the storage device. The client file I/O is converted into block level I/O by the Celerra to retrieve the actual data. Once the data has been retrieved it is once again converted back to file level I/O and returned to the client.

A file I/O specifies the file, and also indicates an offset into the file. For instance, the I/O may specify “Go to byte ‘1000’ in the file (as if the file were a set of contiguous bytes), and read the next 256 bytes beginning at that position”.

Unlike block I/O, there is no awareness of a disk volume or disk sector in a file I/O request. Inside the Celerra, the operating system keeps tracks of where files are located on disk. The OS issues a block I/O request to the disks to fulfill the file I/O read and write requests it receives.

The disk resources can be directly attached to the Celerra or using a SAN, referred to as a gateway configuration.

When it comes to the storage array, the Celerra can be connected to the disks in two different ways. The Celerra can be directly attached to the array making it dedicated to that array only. Another way is to have the Celerra connected to the array through a switch, which is part of a Storage Area Network (SAN). This way, the Celerra can share its backend with more than one array. Direct attached Celerras are known as ‘Integrated’, and Celerras that are connected to a SAN are known as ‘Gateway’.




Client Access

File system creation– AVM – Automatic Volume Management– Manually – Slices, stipes, and meta volumes must first be

created

Create mountpoint for file system– AVM creates the mountpoint automatically

Export or Share mountpoint

Mount or Map– NFS clients use the nfsmount command to export the file

system locally– CIFS clients executes a “map network drive”

There are two different ways to create a file system. One is by manually selecting the storage disks and then creating slices from the selected disks. Next, a stripe is created from the slices, and finally a meta volume is created from several stripes. The file system can only be created once the meta volume exists. An easier way of creating file systems is using the Automatic Volume Management (AVM). AVM creates all of the needed volumes and also creates the mountpoint used in publishing the file system. A mountpoint must be created to make the file system available.

Now that a network file system has been created and mounted, there are two ways it can be accessed using the network. The first method is through the UNIX environment using NFS. This is accomplished by performing an Export. The Export publishes to UNIX clients who can mount (access) the remote file system. Access permissions are assigned when the export is published. The second method is through the Windows environment using CIFS. This is accomplished by publishing a share. The share publishes to Windows clients who map a drive to access the remote file system. Access permissions are assigned when the share is published.

To access the network file system, the client must now mount a directory or map a drive pointing to the remote file system. “Mount” is a UNIX command used to set a local directory pointer to the remote file system. The mount command uses NFS protocol to mount the export locally. For a Windows client to perform this task, it executes a ‘map network drive’. The sequence is my computer> tools>map network drive. Select the drive letter and provide the server name and share name in the Folder field.




Data Mover Configurations

Standby Data Mover Configuration Options– Each standby Data Mover, as a standby for a single primary Data

Mover– Each standby Data Mover, as a standby for a group of primary Data

Movers

Failover Operational Modes– Automatic – failover process occurs without trying a recovery first– Retry – Data Mover is first rebooted before failover starts– Manual – failover is a manual process and no recovery takes place

automatically

Data Mover redundancy is the mechanism by which the Celerra family reduces the network data outage in the event of a Data Mover failure. The ability to failover the Data Movers is achieved by the creation of a Data Mover configuration database on the Control Station system volumes and is managed via the Control Station. No Data Mover failover occurs if the Control Station is not available. These Standby Data Movers are powered and ready to assume the personality of their associated Primary Data Movers, in the event of a failure. If a Primary Data Mover fails, the Control Station detects the failure and initiate the failover process.

Once the failed Data Mover is repaired, the failback mechanism is always manually initiated. This process is the reverse of the failover process and restores the primary functionality to the repaired Primary Data Mover and returns the Standby Data Mover into its standby state in preparation for any future outage.

There are three operational modes of operation for Failover: Automatic, Retry, and Manual. • Automatic Mode: Control Station detects the failure of a Data Mover. The failover process occurs

without trying any recovery process first.• Retry Mode: Control Station detects the failure, an attempt to reboot the failed Data Mover is tried

first before the failover procedure is initiated.• Manual Mode: Control Station detects the failure and removes power from the failed Data Mover.

However, no further Data Mover recovery action is taken until there is administrative intervention. Recovery after a Data Mover failover is always a manual process.




Celerra Features and Functions




In this lesson, we review the many features and functions offered by the Celerra.




Management - Command Line Interface (CLI)

The command line can be accessed on the Control Station via– An ssh interface tool (i.e. PuTTy)– Telnet

It’s primary function is for scripting of common repetitive tasks that may run on a predetermined schedule to ease administrative burden

It has approximately 80 UNIX syntax-like commands

Telnet access is disabled, by default, on the Control Station due to the possibility of unauthorized access if the Control Station is placed on a publicly accessible network. If this is the case, it is strongly recommended that this service is not enabled.

The preferred mechanism of accessing the Control Station is the SSH (Secure Shell) daemon via an SSH client such as PuTTy.




Management – Celerra Manager

GUI management has become consolidated into one product with two options; Celerra Native Manager Basic Edition and Celerra Management Advanced Edition.

The Basic Edition is installed, along with the DART OS, and provides a complete set of common management functionality for a single Celerra at a time. The Advanced Edition adds multiple Celerra supports, along with some advanced feature GUI management, and is licensed separately from the DART code.




Celerra Manager Wizards

Celerra Manager offers a number of configuration Wizards for various tasks to assist with new administrator ease of implementation.




Celerra Manager Tools

Celerra Manager offers a set of tools to integrate Celerra monitoring functionality and launch Navisphere Manager.

With the addition of the Navisphere Manager Launch capability, the SAN/NAS administrator has a more consolidated management environment.




EMC ControlCenter V6.x.x NAS Support

Discovery and Monitoring– Data Movers– Devices and volumes– Network adapters and IP

interfaces– Mount points– Exports– File systems (including

snapshots and checkpoints)

The EMC flagship management product, EMC ControlCenter, has the capability of an assisted discovery of both Celerra and third party NAS products, namely NetApps filers.

Currently, management of the Celerra family is deferred to the specific product management products due to the highly specialized nature of the NAS environment. This product functionality is focused mainly around discovery, monitoring, and product management software launch capability.

The ControlCenter V6.x.x has enhanced device management support for the Celerra family. The ControlCenter Celerra Agent runs on Windows and has enhanced discovery and monitoring capabilities. You can now view properties information on Celerra Data Movers, devices, network adapters and interfaces, mount points, exports, file systems, and volumes from the ControlCenter Console. You can also view alerting information for the Celerra family as well.




VLANCreate logical LAN segment– Divide a single

LAN into logical segments

– Join multiple separate segments into one logical LAN

VLAN Tagging– 802.1q

Simplified Management– No network

reconfiguration required for member relocation

Hub Hub

Hub Hub

Bridge

or

Switch

Bridge

or

Switch

Hub Hub

Router

Workstation VLAN B

VLAN B

VLAN A

VLAN A

Collision DomainLAN Segment

Collision DomainLAN Segment

Collision Domain

LAN Segment

Broadcast Domain

LAN

Broadcast Domain LAN

Network domains are categorized into Collisions or Broadcasts. A collision is a a LAN segment within which data collisions are contained. A Broadcast, refers to the portions of the network through which broadcast and multicast traffic is propagated. Collision domains are determined by hardware components and how they are connected together. The components are usually client computers, hubs, and repeaters. A network switch or a router that generally does not forward broadcast traffic separates a Collision domain from a Broadcast domain. VLANs allow multiple, distinct, possibly geographically separate network segments to be connected into one logical segment. This can be done by subnetting or using VLAN tags (802.1q.), which is an address added to network packets to identify the VLANs to which the packet belongs. This could allow servers that wereconnected to physically separate networks to communicate more efficiently and it could prevent servers that were attached to the same physical network from impeding one another.

By using VLANs to logically segment the Broadcast Domains, the equipment contained within this logical environment does not need to be physically located together. This means that if a mobile client moves location, an administrator does not need to do any physical network or software configuration for the relocation as bridging technology would now be used. A router would only be needed to communicate between VLANS.

VLAN Tagging allows a single Gigabit Data Mover port to service multiple logical LANs (Virtual LANs). This allows data network nodes to be configured (added and moved as well as other changes) quickly and conveniently from the management console, rather than in the wiring closet. VLAN also allows a customer to limit traffic to specific elements of a corporate network and protect against broadcasts (such as denial of service) affecting whole networks. Standard router-based security mechanisms can be used with VLANs to restrict access and improve security.




File system Control - User Quota Restrictions

There are three main types of quotas used in data space control:– Soft Quota

Amount of data space or number of files used under normal working conditions

– Hard QuotaTotal space or number of files a user/group can use or create on a file system

– Tree QuotaTotal space or number of files that a user/group can use or create on a data directory tree. They are used as a logical mechanism to segment large file systems into smaller administrative portions that do not affect each other’s operation

One of the most common concerns in a distributed data environment is that users tend to save many copies of the same information. When working in a collaborative distributed environment, the amount of data space required by each user expands rapidly and, in some cases, uncontrollably. To minimize data space outages, the user space can be controlled by imposing Quotas on users, or groups of users, to limit the number of blocks of disk space they can use or the number of files they can create.

The Soft Quota is a logical limit placed on a user that can be exceeded without the need for any administrative intervention. Once the soft quota limit has been exceeded, the user has a grace period to use the extra space defined by the hard quota limit. However, the user/group cannot exceed the hard limit.

The grace period is a time limit during which the user, or group, can continue to increase the amount of disk space used or number of files created. If the grace period expires, the user/group must reduce the amount of space used, or the number of files, to below the soft limit. This must be done before any new space or files can be created.

The Celerra family supports all of these Quota methodologies, assisting administrators accustomed to these management tools, with a seamless transition into an EMC Celerra environment.




Usermapper Windows and UNIX integration

Usermapper is the methodology by which Windows SIDsare equated with UNIX UIDs on the Celerra

Choosing the Usermapper methodology is determined by the client environment– Windows only

Internal Usermapper in Windows-only environments is recommended. Celerra Network Server installations after version 5.2 use this by default

– Mixed protocol UNIX and Windows In multiprotocol environments, file systems can be accessed by UNIX and Windows users. Some of the methodologies that enable this to be achieved are:

Active Directory (using Microsoft Management Console snap-ins)A Data Mover’s local user and group filesNetwork Information Service (NIS)

EMC Data Mover operating system, DART, utilizes a very specialized UNIX-like file system and has the same security structures. To support disparate clients, NFS and CIFS, the various environmental security structures need to be equated to the Data Mover structures. In the NFS environment, no translation needs to be performed. However in the Microsoft environment, the Security Identifiers (SID), need to be equated to the security structures on the file system. Usermapper is the mechanism that is used by the Celerra to achieve this mapping.

In multiprotocol environments, file systems can be accessed by UNIX and Windows users. File access is determined by the permissions on the file or directory, the UNIX permissions, Windows access control lists (ACLs), or both permissions and ACLs. Therefore, if a user has a UNIX and Windows user accounts, you should choose a mapping method that allows you to indicate that the two accounts represent the same user. Some of the methodologies that enable this to be achieved are:

−Active Directory (using Microsoft Management Console snap-ins)−A Data Mover’s local user and group files−Network Information Service (NIS)

If a user in a multiprotocol environment only uses a single logon ( through Windows or UNIX), then Usermapper is ideal. If a user has only one account, mapping to an equivalent identity in the other environment, then Usermapper is not necessary.




Virtual Data Movers

Virtual Data Movers on Single Physical Data Movers– Ability to create multiple virtual CIFS servers on each logical Data

Mover– Consolidation of multiple server file serving functionality onto single

Data Movers as each virtual Data Mover can maintain isolated CIFS servers with their own root file system environment

– Allows whole Virtual Data Mover environments to be loaded, unloaded, or even replicated between physical Data Movers for ease of Windows environmental management

CIFS servers are not logically isolated and although they are very useful in consolidating multiple servers into one Data Mover, they do not provide the isolation between servers as needed in some environments such as data from disjoint departments hosted on the same data mover.

Virtual Data Movers (VDM) support separate isolated CIFS servers, allowing you to place one or multiple CIFS servers into a VDM, along with their file systems. The servers residing in a VDM store their dynamic configuration information (such as local groups, shares, security credentials, and audit logs, etc.) in a configuration file system. A VDM can then be loaded and unloaded, moved from physical Data Mover to Data Mover, or even replicated to a remote Data Mover as an autonomous unit. The servers, their file systems, and all of the configuration data that allows clients to access the file systems are available in one virtual container. Having the file systems and the configuration information contained in a VDM does the following:

• Enables administrators to separate CIFS servers and give them access to specified shares.• Allows replication of the CIFS environment from primary to secondary without impacting server

access.• Enables administrators to easily move CIFS servers from one physical Data Mover to another.

A VDM can contain one or more CIFS servers. The only requirement is that you have at least one interface available for each CIFS server you create. The CIFS servers in each VDM have access only to the file systems mounted to that VDM, and therefore can only create shares on those mounted file systems. This allows a user to administratively partition or group their file systems and CIFS servers.




High Availability - Network FailSafe Device

Network outages, due to environmental failure, are more common than Data Mover failures

Network FailSafe Device – DART OS mechanism to minimize data access disruption due to

network failures – Logical device is created using physical ports or other logical ports

combined together to create redundant groups of ports– Logically grouped Data Mover network ports monitor network traffic

on the ports– Active FailSafe Device port senses traffic disruption– Standby (non-active) port assumes the IP Address and Media

Access Control address in a very short space of time, thus reducing data access disruption

To minimize data access disruption due to network outages caused by environmental failures, the DART OS has a mechanism that is environment agnostic—the Network FailSafe Device. This is a mechanism by which the Network ports of a Data Mover may be logically grouped together into a partnership that monitors network traffic on the ports. If the currently active port senses a disruption of traffic, the standby (non-active) port assumes the active role in a very short space of time, thus reducing data access disruption.

The way this works is a logical device is created, using physical ports or other logical ports, combined together to create redundant groups of ports. In normal operation, the active port carries all network traffic. The standby (non-active port) remains passive until a failure is detected. Once a failure has been detected by the FailSafe Device, this port assumes the network identity of the active port, including IP Address and Media Access Control address.

Having assumed the failed port identity, the standby port now continues the network traffic. Network disruption due to this change is minimal and may only be noticed in a high transaction oriented Celerra implementation or in CIFS environments due to the connection-oriented nature of the protocol.

Network FailSafe device configuration is handled transparently to client access and the ports that make up the FailSafe device need not be of the same type. This high availability solution can be combined with logical Aggregated Port devices to provide even higher levels of redundancy.

Care must be taken to ensure that once failover has occurred, that client expected response times remain relatively the same and data access paths are maintained.




High Availability - Link AggregationCombining of two or more data channels into a single data channel for high availability; if one link fails, the others take overTwo Methods: CISCO FastEtherChannel & IEEE 802.3ad LACP

Having discussed the network FailSafe device, the next methodologies we look at are the two Link Aggregation methodologies. They are IEEE 802.3ad Link Aggregation Control Protocol and CISCO FastEtherChannel using Port Aggregation Protocol (PAgP). Link aggregation is the combining of two or more data channels into a single data channel.

The purpose for combining data channels in the EMC implementation is to achieve redundancy and fault tolerance of network connectivity. It is commonly assumed that link aggregation provides a single client with a data channel bandwidth equal to the sum of the bandwidth of individual member channels. This is not, in fact, the case due to the methodology of channel utilization and, it may only be achieved with very special considerations to the client environment. The overall channel bandwidth is increased, but the client only receives, under normal working conditions, the bandwidth equal to one of the component channels.

Ethernet Trunking (Ether Channel) increases availability. It provides statistical load sharing by connecting different clients to different ports. It does not increase single-client throughput. Different clients get allocated to different ports. With only one client, the client accesses Celerra via the same port for every access. This DART OS feature interoperates EtherChannel capable Cisco switches. EtherChannel is Cisco proprietary.

To implement Link Aggregation, the network switches must support the IEEE 802.3ad standard. It is a technique for combining several links together to enhance availability of network access and applies to a single Data Mover and not across Data Movers. The current implementation focuses on availability, therefore check the Celerra support matrix. Only full duplex operation is currently supported.




Disk-Based Replication and Recovery Solutions

SynchronousSynchronousDisasterDisasterRecoveryRecovery

SRDF/SSRDF/S

Seconds

FileFileRestorationRestoration

Celerra SnapSure Celerra SnapSure

Hours

FileFile--basedbasedReplicationReplicationTimeFinder/FSTimeFinder/FS

Celerra ReplicatorCelerra ReplicatorSRDF/ASRDF/A

Minutes

High-end environments require non-stop access to the information pool. From a practical perspective, not all data carries the same value. The following illustrates that EMC Celerra provides a range of disk-based replication tools for each recovery time requirement.File restoration is used for the information archived to disk and typically saved to tape. Here we measure recovery in hours. Celerra SnapSure enables local point-in-time replication for restoring deleted files, and backups. File-based replication is used for information that is recoverable in time frames measured in minutes. Information is mirrored to disk by TimeFinder and the copy is made accessible with TimeFinder/FS. The Celerra Replicator creates replicas of production file systems either locally or at a remote site. Recovery time from the secondary site depends on the bandwidth of the IP connection between the two sites. SRDF/A provides an extended-distance replication facility.The Replicator feature supports “data recovery” for both CIFS and NFS by allowing the secondary file system (SFS) to be manually switched to read/write mode after the Replicator session has been stopped (either manually or due to a destructive event). Note that there is no re-synch or failback capability.Synchronous disaster recovery is used for the information requiring disaster recovery with no loss of transactions. This strategy allows customers to have data recovery in seconds. SRDF, in synchronous mode, facilitates real-time remote mirroring in campus environments (up to 60 km).File restoration and file-based replication (Celerra Replicator, EMC OnCourse) are available with Celerra /CLARiiON. The entire suite of file restoration, file-based replication, and synchronous disaster recovery is available with Celerra /Symmetrix.




Celerra SRDF Disaster Recovery

Increases data availability by combining the high availability of the Celerra family with the Symmetrix Remote Data Facility

– Allows an administrator to configure remote standby Data Movers waiting to assume primary roles in the event of a disaster occurring at the primary data site

– SRDF allows administrator to achieve a remote synchronous copy of production file systems at a remote location

– Real-time, logically synchronized and consistent copies of selected volumes– Uni-directional and bi-directional support– Resilient against drive, link, and server failures– No lost I/Os in the event of a disaster– Independent of CPU, operating system, application, or database– Simplifies disaster recovery failover and failback

CelerraCelerraUni or bi-directional

Campus (60 km) distance

Network

In combining the high availability of the Celerra family with the Symmetrix Remote Data Facility, data availability increases exponentially. What the SRDF feature allows an administrator to achieve is a remote synchronous copy of production file systems at a remote location. However, this entails the creation of Symmetrix specific R1 and R2 data volumes and this functionality is currently restricted to Celerra / Symmetrix implementations. This feature allows an administrator to configure remote standby Data Movers waiting to assume primary roles in the event of a disaster occurring at the primary data site. Due to data latency issues, this solution is restricted to a campus distance of separation between the two data sites.The SRDF solution for Celerra can leverage an existing SRDF transport infrastructure to support the full range of supported SAN (storage area network) and DAS (direct-attached storage) connected general purpose server platforms. The Celerra disaster recovery solution maintains continuously available file systems, even with an unavailable or non-functioning Celerra or Symmetrix. SRDF connects a local and remote Celerra over a distance of up to 37 miles (60 km) via an ESCON or Fiber Channel SRDF connection. The Celerra systems communicate over the network to ensure the primary and secondary Data Movers are synchronized with respect to metadata, while the physical data is transported over the SRDF link. In order to ensure an up-to-date and consistent copy of the file systems on the remote Celerra, the synchronous mode of SRDF operation is currently the only supported SRDF operational mode. SRDF has two modes of operation: active-passive and active-active. Active-passive (Uni-directional) SRDF support means that one Celerra provides active Data Mover access while a second (remote) Celerra provides all Data Movers as failover. Active-active (Bi-directional) SRDF support means that one Celerra can serve local needs while reserving some of its Data Movers for recovery of a remote Celerra, which reserve some of its Data Movers for recovery of the first Celerra . The mode of operation with SRDF/S is Active-Active.




File-based Replication - Celerra TimeFinder/FSPoint-in-time copy of file systemProvides an independent mirror copy of data for “out-of-band” business processes and support functionsProvides read and write functionality independent of the original dataRequires Symmetrix storageCelerra controlled features

– Point-in-time copies– Dynamic mirroring– Multiple BCVs– Spans volumes– Entire file system

Applications– Backup and restore– Data warehouses– Live test data– Batch jobs

BCV = Business Continuance Volume

Point-in-timecopy

CelerraSymmetrix

FSAFSAPFS PFS

Copy

A data replication method that provides high availability and rapid recovery is the TimeFinder/FS feature. In the Symmetrix Storage array, TimeFinder is used to produce a stand-alone copy of a production volume. It uses a specially defined volume, called a Business Continuance Volume (BCV), to facilitate this functionality. As the Symmetrix Array is the only array currently able to define a BCV, TimeFinder/FS on the Celerra Family is currently restricted to implementations with Symmetrix only. The TimeFinder/FS implementation is different from a standard TimeFinder implementation. It allows users to copy file systems into Business Continuance Volumes (BCVs) for a wide variety of purposes.

TimeFinder/FS creates a point-in-time copy, or a dynamic mirror, of a file system. The TimeFinder/FS option, which is integrated in the Control Station, allows users to create file system copies with only a brief suspension of access to the original file system. These copies permit independent read/write copies of data, useful for non-disruptive file backups, “live copy” test beds for new applications, and mirror copies of files for redundancy and business continuity. It facilitates backup and restore of older versions of a specific file, directory, or complete file system. It can also function in mirroring and continuous updates mode for an active file system. File system copies require that the configuration of the Symmetrix system attached to the Celerra include BCVs. A BCV, which attaches to a standard volume on which a file system resides, provides the foundation for the file system copy. File systems can share BCVs, although the BCV remains dedicated to a volume.

After issuing the single TimeFinder split command, the Celerra initiates the synchronization of the primary and BCV file system volumes. When the synchronization is almost complete (typically within two seconds), access to the primary file system is halted and Data Mover memory is flushed to ensure a consistent copy of the data is split off. The BCV becomes available for use only when synchronization is complete. The BCV is mounted to a second Data Mover for backup or for use in secondary business processes.




TimeFinder/FS NearCopy

Synchronous data replication and Disaster Recovery solution– Requires Symmetrix storage– R1 / R2 data is synchronized

for disaster recovery– Read-only R2 data accessible

via BCV for backup and testing purposes

– Synchronous SRDF as base technology

– ESCON / Fibre Channel

SRDF

Celerra Symmetrix Symmetrix

R1

BCV

R2

Celerra

Windows & UNIX

DataNetwork

Local Site Remote Site

Combining the TimeFinder/FS product with Symmetrix Remote Data Facility (SRDF), has enabled the TimeFinder/FS concept to be used for Disaster Recovery because the stand-alone copy of the data can now be synchronously updated at a remote site. It allows this solution to be utilized as a Data Replication and Disaster Recovery solution. This is known as TimeFinder/FS Near Copy because the supported SRDF network distance between the two sites is 60 km (campus) due to the synchronous nature of the R2 volume updates.

The Remote TimeFinder/FS Near Copy solution applies to environments that have a requirement for real-time, synchronous, disk-based recovery. Synchronous SRDF is used to maintain the R1/R2 pair. TimeFinder BCVs can be generated from the R2 and made available (read-only) to independent Data Movers in the remote Celerra. The Celerra at the remote site can make the content available for secondary business processes such as testing or backup. This solution works for environments with SRDF active-active mode, where R1s and R2s exist in both sites. It also works for environments with active-passive mode, where all the R1s are located in one site with SRDF to a passive R2- only Symmetrix. Synchronous SRDF operates over ESCON/Fibre Channel. The BCV at the R2 site is read-only and restore must be done manually.




TimeFinder/FS FarCopy

Asynchronous data replication solution – Replicated point-in time

copy of primary site file system

– Data replication time does not impact production file system performance

– Requires Symmetrix storage

– Uses SRDF Adaptive Copy as base technology

– Sites can be geographically distant

SRDFAdaptive

Copy

Symmetrix CelerraCelerra SymmetrixR1/ BCV

STD

R2/ BCV

STD

Windows & UNIX Windows & UNIX

R1 SiteNetwork

R2 SiteNetwork

Local Site Remote Site

All remote copies of data are not designated for Disaster Recovery, but could be used for data replication. To facilitate these kinds of solutions, where time taken to replicate the data does not impact the performance of the production file system, TimeFinder/FS Far Copy can be utilized.

The Remote TimeFinder/FS FarCopy solution applies to environments that have a requirement for remote point-in-time copies of the file systems beyond the typical distances associated with synchronous SRDF—that is, greater than 60 km. Adaptive SRDF is used to replicate the information over geographical distances. The read-only copy at the remote site can be made available for secondary business processes such as testing or backup.

Implementation of this solution allows data to be replicated asynchronously over a very wide area to where it is needed. It does not affect the Production file system, PFS, because a BCV copy of the PFS is first made and then the BCV is copied to the remote location, while the Production Filesystem continues serving data to the clients uninterruptedly.

As this solution is dependent on the TimeFinder/FS, it is only supported with the Celerra/Symmetrix configuration.

The process for performing this action is:1. Create a R1/BCV of STD.2. Sync R1/BCV with R2/BCV over SRDF link.3. Restore R2/BCV to Local STD (read-only if the relationship between the R2BCV needs to be maintained).4. Import File System on R2 Celerra.




Celerra SRDF/A

SRDF/A: Asynchronous File-based Replication– Delta Set-synchronized copies of volumes– Extended distances

NS Series/NSX Gateway

SRDF/A

NS Series/NSXGateway

Network

Celerra also integrates with the Symmetrix SRDF/A, asynchronous, product for extended distance support between the data centers. However data replication in this case is asynchronous and therefore is not considered a Disaster Recovery solution. After a manual failover and recovery process, users gain access to file systems in the event that the local Celerra and/or the Symmetrix becomes unavailable. The mode of operation supported with this solution is Active-Passive.

The Delta Sets for data transfer between sites are as follows:−Capture (Source, aka the N cycle)− Transmit (Source, aka the N-1cycle)−Receive (Target, aka the N-1cycle)−Apply (Target, aka the N-2 cycle)




Celerra ReplicatorPoint-in-time read-only file system copy over IP

Production file system available during replication

Only sends changed data over IP (after initial synchronization)

One-way link for single target content distribution

Asynchronous Data Recovery

Data recovery for CIFS data

Requires Symmetrix storage

Windows & UNIXWindows & UNIX

Celerra/Symmetrix Celerra NS

R1 SiteNetwork

R2 SiteNetwork

Production Filesystem

Primary SavVol

Secondary SavVol

IPNetwork

Secondary Filesystem

Log

Replication Process

Celerra Replicator is an IP-based replication solution. Replication between a primary and a secondary file system can be on the same or a remote Celerra system. The data flow is described in the following steps:

1. Manually synchronize production file system and secondary file system - Initial Copy. Before starting Celerra Replicator, a full copy of the Production File System, PFS, has to be made on the Secondary File System, SFS.

2. Any changes to the production file systems are recorded in the log - Log Changes. On the primary, where PFS is mounted read/write, all CIFS block addresses of changed blocks are logged in a main memory log. Periodically, copy changes to SavVol. Transfer is triggered by special events. The trigger can be controlled by a user-defined policy or via an explicit request on the Control Station.

3. Remote replication copies log changes to primary SavVol, and begins movement to secondary SavVol. The transfer step comprises creating a local copy of all of the changes made to the log file and copying them to a local “delta set”called SavVol. The copy process is designed to copy a checkpoint of the blocks, allowing the PFS to be modified during the copy.

IP Transfer: A transfer is set up with the remote replica for the newly updated set of SavVol blocks. While the transfer is in process, read/write activity on the PFS is not halted. Instead, a new log area is set up to track subsequent changes. Concurrently with the copy process, the newly created delta set is transferred to the secondary over IP. In local replication (when the secondary Data Mover is in the same cabinet as the primary), no transfer is required. The delta set is accessible as a shared volume.

4. Remote SavVol changes are incorporated into secondary file systems. Playback on Secondary: The delta set is made available to the secondary in its SavVol. When the delta set arrives at the SavVol of the secondary and has been flagged as “valid”, the secondary starts to replay the blocks from its local SavVol (a local copy of the replicated delta set) and applies the delta set to the SFS. This operation occurs transparently and with almost no interruption to the SFS access.

The Replicator feature is able to support “data recovery” for both CIFS and NFS by allowing the secondary file system (SFS) to be manually switched to read/write mode after the Replicator session has been stopped, manually or due to a destructive event.




Celerra Replicator for iSCSIProvides an application-consistent point-in-time replication solution for iSCSI– Replicates application-consistent iSCSI

snaps (via VSS)– Sends only changed data over the IP– Managed by EMC Replication

Manager/SE

Production LUN available during replication; remote copy available for backup or repurposingCost-effective, long-distance data recovery– In the event of failure, the LUN copy

can be made available as read/write– Changes to the LUN copy can be

reapplied to the production LUN on failback

Production LUN

NSX

IPnetwork

Windows + Replication Manager/SE

Secondarynetwork

Primarynetwork

4

NS80

LUN Copy

iSCSISnaps

iSCSISnaps

As iSCSI is deployed within the infrastructure, protection of files that are accessed over iSCSI is critical. EMC Celerra Replicator supports application-consistent iSCSI replication. Managed by EMC Replication Manager/SE, only the changed data is sent to the replica, improving network efficiencies.

Celerra Replicator for iSCSI provides data protection and integrity. If data on the production LUN is deleted or corrupted, it can be rebuilt (up to the point of the latest replicated snapshot) from the replica LUN. The destination LUN is protected (read-only), preventing the iSCSI initiator from writing to it. The copy can be made read/write in the event of a failure of the primary. The replica of the production LUN can be used for content distribution, data mining, testing, or other applications that offload activity from the production LUN.

Celerra Replicator for iSCSI provides failover capability so if the production LUN becomes unavailable, the replica on the destination becomes writable and serves as the production LUN. If the original production LUN returns to service, a manual process can update it with the latest data modifications made at the failover site. It can then fail back to reestablish the replication relationship. Celerra Replicator for iSCSI support is currently limited to Windows and the failback process is manual.




File Restoration - Celerra SnapSureEnables speedy recovery

– Low volume activity, read-only applications

– Simple file undelete– Incremental backup

Logical point-in-time view of Celerra data

– Works for all Celerra Implementations– Saves disk space– Maintains pointers to track changes to the

primary file system – Not a mirror; creation of specialized

volumes (R1/R2, BCVs) not required– Multiple Checkpoints for recovery of

different point-in-time imagesGUI Checkpoint schedule manipulationCheckpoint out of order delete Automatic mounting

Productionfilesystem

Checkpoint

Celerra CLARiiON orSymmetrix

The SnapSure feature of the Celerra family provides high data availability and speedy recovery. This methodology uses a logical point-in-time view to facilitate Incremental backup views of a Production File System, PFS, individual file recovery, and roll back of an entire file system to a previous point-in-time image. SnapSure maintains pointers to changes to the primary file system and reads data from the primary file system or a copy area. The copy area is defined as a meta-volume (SavVol).One of the obvious benefits of this solution is that it is storage array agnostic (it works for all Celerra implementations). This also means that there are no specialized volumes that need to be configured for this feature to function. Some other replication methodologies, such as SRDF and TimeFinder/FS, are dependent on the creation of Symmetrix Remote Data Facility and Business Continuity Volumes in the Symmetrix. SnapSure works with any back-end storage array, CLARiiON, or Symmetrix.Multiple Checkpoints can be done on the PFS and thereby facilitate the ability to recover different point-in-time images of files or file systems. Without using any other similar replication methodologies, such as CelerraReplicator, the currently supported maximum of Checkpoints per file system is 32.For ease of management, Checkpoints can be manipulated with the GUI management interfaces, along with the ability to schedule the frequency of the Checkpoints.Most Checkpoint technology is chronologically linked. However, the current version of Celerra DART supports out of order deletion of checkpoints, while maintaining SnapSure integrity. This feature allows customers to delete any Checkpoint instead of being constrained to having to delete Checkpoints from the oldest to maintain integrity.A customer may also delete an individual scheduled checkpoint instead of the entire schedule, and may refresh any checkpoint instead of only the oldest. Checkpoints are automatically mounted upon creation and maintenance of a hidden checkpoint directory in any subdirectory. This new hidden directory now also allows changing the default name (yyy_dd_hh_mm_ss_GMT) into something more administratively friendly.




Celerra FileMover for Data Migration

FileMover migrates less-used files to secondary storage for higher user service levels

Migration process is transparent to user

FileMover is an API, when used in conjunction with a policy engine, migration software and a secondary storage infrastructure

Data supports concept of Information Lifecycle Management when combined with DiskXtender

ATATape/Optic

Centera

SecondaryStorage

Celerra

User

1. Search

2. ReadData

Access

Policy andMigrationSoftware

3. Write

4. Replace file w/ Stub

File

Celerra FileMover solutions deliver dynamic file mobility—the ability to automate the movement of files in an EMC Celerra NAS environment across a hierarchy of storage platforms.

Celerra FileMover is an open, flexible solution that enables a choice of secondary storage based on business, application, and information requirements. Active files requiring fast access can be placed on higher performance platforms while older, less frequently used files can be automatically moved to second-tier platforms including ATA, EMC Centera, tape, and optical. This results in higher service levels to users and a lower TCO for storage assets.

A core component of a Celerra FileMover solution is the policy and migration software. EMC, as well as third-party software providers, have integrated to the Celerra FileMover API. This enables EMC, along with our partners, to offer solutions that manage the optimization and placement of files from Celerra to secondary storage. EMC DiskXtender Data Manager for Celerra is a powerful, highly scalable solution that delivers policy-based, file system data management that enables automated and transparent data movement between Celerra and secondary storage platforms. DiskXtender Data Manager allows you to more effectively manage your data between the primary Celerra system and the secondary storage platforms.




Celerra Network BackupThird-party backup products– CIFS access through shares– NFS access through mounts Network

CLARiiONor

Symmetrix

Celerra

Tape Library

Data FlowControl Information Flow

Network backups utilize a backup server that mounts the remote file systems and transfers them to tape. The data path includes the Celerra Data Mover, the network, and the backup server. Most third-party backup products support CIFS network backups through remote file shares and NFS network backups using remote file mounts. Third-party backups should not be used for multi-protocol files. For example, if they are backed up over a CIFS share, they lose any NFS attributes and permissions upon restoration.

Network backups entail simply mounting the file systems across the network and backing up to the backup server. EMC Data Manager, and most backup utilities, support this option. EDM also preserves the bilingual (CIFS and NFS) information with network backups.




NDMP BackupEMC Networker for NDMPThird-party NDMP– VERITAS NetBackup– HP OmniBack

Celerra backs up data to directly attached tape library unit (TLU)Backup is performed by client running NDMP-compliant ISV softwareNo LAN performance impact: only control data goes via the LANMulti-protocol support: both CIFS and NFS file system attributes

Data FlowControl Information Flow

Client with NDMP backup software

ProductionNetwork

TapeLibrary Unit

Celerra

Data MoverNDMP Server

CLARiiONor

Symmetrix

NDMP (Network Data Management Protocol) is an industry standard LAN-less backup for NAS devices. NDMP backups only use the LAN for control information and the data is transferred to the local backup device. EMC Networker can be used in NDMP backups. VERITAS NetBackup and HP OmniBack, among others, also support Celerra NDMP backups.

Backup activity can be localized to a single “backup” Data Mover, requiring only one Data Mover be physically attached to the TLU (tape library unit). This option is implemented through TimeFinder/FS. File systems are split off and mounted to the backup Data Mover and backed up with no impact to the primary file system.

Tape library units are connected to a Data Mover via a SCSI interface. Backup traffic is offloaded from the network and allows for dual accessed file systems to be backed up, preserving both permission structures on the file system. NDMP backups preserve the bilingual file information.




NDMP2D – Network Data Management Protocol to Disk

A software-based tape target that resides on a Data Mover that emulates SCSI Tape Library Units (TLU) and tape drives

Emulation allows NDMP to work with disk storage, rather than tape storage.

NDMP2D can be configured and operated from NDMP Client S/W as if it were a physical TLU attached to a Data Mover

Network

Celerra

NetWorker

PFS

NDMP2D

Meta Data

NDMP Control

Network Data Management Protocol to Disk (NDMP2D) is a file system-based storage sub-system that emulates SCSI Tape Library Units (TLUs) and tape drives. Applications communicating with NDMP2D get the impression of communicating with a standard commercial TLU. A software-based tape target that resides on a Data Mover that emulates SCSI Tape Library Units (TLU) and tape drives. This feature is supported in DART v5.4 and later versions. It allows for the creation of a tape target on a Celerra using either Fibre Channel or ATA drives with a 4+1 or 8+1 RAID 3 disk configuration.




Celerra Database Solution

Metrics that determine if Celerra is a fit:

Transaction Rates or IOPS

Bandwidth or Throughput

IOPS 30k

NFS/iSCSI Capability

DSS Workloads

OLTP Workloads

CommercialAppsB

W –

GB

/s2

GB

/s

When it comes to backing up a database, backup/recovery processes are designed to run on the primary site during normal operations. EMC Snapsure technology is used in the backup process of data files and log files. The reason in using Snapsure is to create a read-only image of the entire production file system. The idea here is to balance the performance and availability impact of backup processing with application processing.

OLTP (Online Transaction Processing) workloads have lower throughput and a higher IOPS (IO’s per second) need. They have random reads/writes.

DSS (Decision Support Systems) or Data Warehouse workloads have higher throughput and lower IOPS needs. They have sequential IO’s. NAS (NFS+iSCSI) IOPS and throughput can meet most of the commercial application needs.

Here are some changes that made the Celerra feasible for Database backups:

Network Enhancement:Adaptation of Gig EthernetNetwork Data Separation

Increases in Host Processing Capabilities:Increases in Mhz and Processing PowerDecreases in cost of host RAM

Industry Knowledge of Managing Database Objects:Technology Knowledge Sharing: Web Training, Metalink, etc.DBAs are getting better at managing all database objects




Celerra Product Family



At the end of this lesson, you will be able to identify EMC Celerra platforms and their differences. Let’s examine the current NAS products offered by EMC.




NS Series - Integrated

6 Copper and 2 Optical GigE ports p/Blade


4 Copper GigEports p/DM

Network Access

60 TB FC/ATA32 TB FC/ATA10 TB FC/ATAMax Storage

1 - 2 11Control Station

2 - 4 Blades1 - 2 Blades1 - 2 DMsData Movers

4 GB4 GB4 GBRAMDual 3.4 GHz IntelDual 2.8 GHz IntelDual 1.6 GHz IntelCPU

NS80NS40NS350

Back-end storage is directly connected to the Celerra and not shared with any other servers

The NS Series/Integrated solution includes advanced functionality at the mid-tier’s best price/performance along with an easy upgrade path. It also provides simple implementation and management without sacrificing configurability options. This series is perfect for distributed file server environments with specific performance, availability, and scalability requirements in a multi-protocol environment where SAN infrastructure is not a requirement.

With EMC Celerra NS Series/Integrated, non-stop file access and data protection your business needs is a reality. All systems deliver advanced clustering—so all applications run at the same high performance levels, even if there’s been a failure. Keep availability high with transparent, dynamic failover to a hot standby X-Blade, hardware RAID protection, non-disruptive component replacement, storage processor battery backup, and advanced volume management.

Choose the entry-level NS350—which includes an integrated EMC CLARiiON back end—for a low-cost path to consolidation in a distributed environment, and single or dual Data Mover configurations.

Move up to the NS40 for higher performance and capacity, and single or dual X-Blade configurations.

For even higher performance and capacity, select the NS80, which also includes the powerful EMC CLARiiON backend and 2 to 4 X-Blades.




NS Series - Gateway



Network Access

60 TB FC/ATA32 TB FC/ATAMax Storage

1 - 2 1Control Station

2 - 4 Blades1 - 2 BladesBlades

4 GB4 GBRAMDual 3.4 or 3.6 GHz IntelDual 2.8 GHz IntelCPU

NS80GNS40G

Backend storage is shared between different application servers (SAN environment).

With the NS Series/Gateway, you get exceptional price and performance, including file access up to four times faster than standard NAS. You also get flexible configuration options and exceptional ease-of-use. Both models support Fibre Channel and ATA drives and provide extensive iSCSI support and iSCSI wizards. This solution is great for environments with existing SAN infrastructure, growing deployments of file servers, and specific multi-protocol performance, availability, and scalability requirements.

NS Series/Gateways ensure non-stop access to vital data via integrated advanced clustering. With advanced clustering, the hot spare takes over the full workload in the event of a failure, running at the same performance and service. Hardware-based RAID controllers ensure maximum performance even during rebuilds.

NS40G entry-level gateway delivers high performance, high availability, and simple management—at an exceptional price.

NS80G IP storage gateway offers advanced clustering, with 2 to 4 X-Blades for exceptional flexibility. All at a level of performance formerly available only on high-end NAS systems.




NSX Celerra

Dual 3.4 or 3.6 GHz Intel CPUs

4 GB RAM

4-8 X-Blades

2 Control Stations

Up to 168 TB usable FC and ATA drives

6 Copper and 2 Optical Gigabit Ethernet ports per X-Blade

EMC Celerra NSX is the industry’s most powerful, scalable IP storage gateway and the right choice when there’s a need to lower costs, simplify operation, and manage growth. The NSX is perfect for data center-class server environments that require the highest levels of performance, scalability, and availability to support high-performance file-serving applications.

With Celerra NSX, you can create an extremely efficient, scalable gateway-to-SAN solution connected to Symmetrix DMX and/or the Clariion networked storage systems. You can consolidate storage from multiple applications and hundreds of file servers—all on one platform that’s easy to configure, manage, and use. And at the right price to fit your needs.

Achieve new levels of scalability and performance with an innovative architecture based on X-Bladeserver technology, an EMC-designed and built, Intel-based blade server platform. Scale from approximately 135,000 to more than 300,000 NFS operations per second. Choose from four to eight clustered X-Blades, increasing your usable capacity to up to 168 terabytes. Expand by adding additional X-Blades—without operational delays or disruption. And access files up to 4 times fasterthan standard NAS with EMC-patented Multi-path File System (MPFSi)—without the need to recode applications.




NSX Control Station

Celerra NSX – Front view

LEDs and Switches

Power Switch

NMI Switch

Reset Switch ID Switch

Power Boot Sequence LED

Status LED

HDD Act LED

HDD Fault LED Gb # 1 and Gb # 2 LED

USB Connectors 2 and 3

ID LED

Serial Port COM2

The Control Station is a dedicated management Intel processor-based computer that monitors and sends commands to the blades. The private network connects the two Control Stations (always shipped on NSX systems) to the blades through the system management switch modules. Like previous versions, it provides software installation and upgrade services. It also provides high-availability features such as fault monitoring, fault recovery, fault reporting (CallHome), and remote diagnosing. Two Control Stations can be connected to a public or private network for remote administration. Each Control Station has a serial port that connects to an external modem so that the Control Station can call home to EMC or a service provider if a problem should arise.




NSX Control Station (cont)

Celerra NSX – Rear view

eth3 - Public LAN Port

COM1 - To serial modem (for Call-Home)

eth0 – Internal Network (To Mgmt. Switch-A in

Enclosure 0)

Video Port

Gb2 – Internal Network(To Mgmt. Switch-B in

Enclosure 0)

Gb1 – IPMI (To eth1 of the other

Control Station)

This slide displays the rear view of the Next Generation Control Station. Note the lack of a 25-pin quad serial port and spider cable.




Data Mover HardwareSingle or Dual Intel Processors

PCI or PCI-X based

High memory capacity

Multi-port Network cards

Fibre Channel connectivity to storage arrays

No internal storage devices

Redundancy mechanism

Fibre I/O module GbE I/O moduleNSX / NS80G Blade

NS40 Data Mover

Take a moment to review the different Data Movers displayed in the slide above.




Celerra Benefits and Requirements


Identify the benefits provided by the Celerra

Describe the requirements in a Celerra solution

The objectives for this lesson are shown here.

Let’s take a look at the benefits and requirements that a Celerra solution brings.




Celerra BenefitsHighest availabilityScalabilityAvoids file replication

Centralizes StorageIncreases flexibilitySimplifies ManagementImproves securityReduces Costs

Firewall

Web Servers

Celerra

Internet

Data CenterSn

S2....

S1

InternalNetwork

Disk Storage

Users

Through the advent of NAS applications that use file system level access, the data can now be shared to large numbers of users that may be geographically dispersed, simultaneously. Therefore many users can now take advantage of the availability and scalability of networked storage. Centralizing file storage can reduce system complexity and system administration costs, along with simplifying backup, restore, and disaster recovery solutions.

Although the Celerra trades some performance for manageability and simplicity, it is by no means a lazy technology. Gigabit Ethernet allows the Celerra to scale to high performance and low latency, making it possible to support a myriad of clients through a single interface. The Celerra can support multiple interfaces and can support multiple networks at the same time.




Celerra Consolidation Example

Solution

Internet/Intranet

UNIX Windows

Celerra

General purpose OS serving files viaFTP, CIFS, NFS, HTTP. . .

Current Environment

Internet/Intranet

UNIXNT

W2K

UNIX Windows

In many companies the need for two differing environments maintains the separation of two technologies using the same infrastructure for the same purpose, but in different ways. Access to networked files for UNIX (NFS) and Microsoft (CIFS) are traditionally housed on separate server infrastructures. However by implementation of the Celerra, these same file structures can be housed together, while still maintaining their integrity. Within Celerra deployments, the same file system can be accessed by the same user via different technologies, either NFS or CIFS, and still maintain the integrity of the data and security structures (as long as the applications used for both methodologies understand the data structures presented).

Benefits of this solution:Provides continuous availability to filesHeterogeneous file sharingReduces cost for additional OS dependent serversAdds storage capacity non-disruptivelyConsolidates storage managementLowers Total Cost of Ownership




Celerra Scalability with SANConsolidated storage infrastructure for all applications

NAS front end scales independently of SAN back end

Storage allocated to Celerra and servers as needed

Centralized management for SAN and NAS

iSCSI gateway to SANWindows

UNIX

CLARiiONCX Family

ConnectrixSAN

SymmetrixDMX Family

Celerra NSxG Family

Celerra NSX

One of the reasons that the Celerra scales impressively is due to the gateway architecture that separates the NAS front end (Data Movers) from the SAN backend (Symmetrix or CLARiiON).

This allows the frontend and bac end to grow independently. Customers can merely add Data Moversto the Celerra to scale the front-end performance to handle more clients. As the amount of data increases, you can add more disks, or the Celerra can access multiple Symmetrix or CLARiiON. This flexibility leads to improved disk utilization.

Celerra supports simultaneous SAN and NAS access to the CLARiiON and Symmetrix. and can be added to an existing SAN, with general purpose servers now able to access non-NAS back-end capacity. This extends the improved utilization, centralized management, and TCO benefits of SAN plus NAS consolidation to Celerra, Symmetrix, and CLARiiON.

The configuration can also be reconfigured via software. Since all Data Movers can “see” the entire file space, it is easy to reassign file systems to balance the load. In addition, file systems can be extended online as they fill.

Even though the architecture splits the front end among multiple Data Movers and a separate SAN backend, the entire NAS solution can be managed as a single entity.




Celerra RequirementsPerformance– Physical device performance– CIFS/NFS performance

Discovery– Ability to discover the Celerra along with its hardware configuration

Space Management– Control over space usage

Backup/RecoveryAsset Management

The following requirements need to be considered when implementing a Celerra solution:Performance includes physical device performance (CPU utilization, memory utilization, number of I/O requests received, cache performance, etc.) as well as CIFS/NFS performance. Management software should not only collect this data, but also filter it so intelligent events can be initiated when thresholds are breached.Discovery is the ability to discover the Celerra, along with its hardware configuration and logical storage attributes, as they change, without manual entry.Space Management - Administrators need to know who is using it, what they are using, and how much of it.Backup/Recovery - Backup and recovery for the Celerra is similar to other file servers. − Logical backups at the file and file system level need to be considered along with more database-centric

backup and recovery solutions involving data associated with a relational database management system (RDBMS)

− Many tools are available that cover the full range of backup and recovery optionsAsset Management -There are two aspects to asset management: − Physical attributes of a Celerra and its associated configuration− Logical or data aspects, both need to be inventoried and, in most cases, have some level of cost recovery

The ever-changing capacity information of a Celerra needs to be collected in order to correlate that data to utilization. This can be either at the user or application level, with the data being fed into the fixed asset repository to create cost recovery reports.




Celerra Challenges

Speed– Network latency and congestion– Protocol stack inefficiency– Application response requirements

Reliability– Distance between network segments– Centralized storage silos may be single points of failure

Connectivity

Scalability

Some of the Celerra Challenges are listed on this slide:

SpeedNetwork latency and congestionProtocol stack inefficiency − encapsulation, possessor overhead, and relatively small payload

Application response requirements

ReliabilityDue to the large geographical coverage of enterprise networks there are inherent possibilities for network failures, but with redundancy planning these issues can be minimized.Centralized storage silos may become single points of failure without remote mirroring or backup facilities.

ConnectivityWithout newly emerging technologies, iSCSI, FCIP & iFCP, many applications required block level access therefore excluding the Celerra as a solution for businesses

ScalabilityAlthough Celerra can scale to terabytes of storage capacity, once the capacity is exhausted the only way to expand is to add additional devices. This can cause additional problems when data center real estate is at a premiumOnce a Celerra is fully populated, including external storage enclosures, the only remaining scaling option is to buy another system. When data center real estate is at a premium this can be seen as a major limitation.




Module Summary

Key points covered in this module:

Identify the concepts, architecture, terminology, and environmental aspects of NAS using the Celerra

Describe the Celerra’s theory of operation




Identify the benefits provided by a Celerra solution

These are the key points covered in this module. Please take a moment to review them.

celerra foundations srg

Documents