DISASTER RECOVERY
JOURNAL

Return to the Winter 2001
Index

P. O. Box 510110
St. Louis, MO 63151
(314) 894-0276
Fax: (314) 894-7474
Internet www.drj.com
E-mail drj@drj.com

PUBLISHER &
EDITOR-IN-CHIEF
Richard L. Arnold, CBCP
richard@drj.com

SENIOR EDITOR
Janette Ballman
janette@drj.com

EDITOR
Michelle Saab
michelle@drj.com

COPY EDITORS
Edward H. Pearce, CBCP
drj@drj.com
Richard Sandhofer
richards@drj.com

INTERNET /
ADVERTISING
Robert Arnold
bob@drj.com

_____________

Corporate

President/CEO
Richard L. Arnold, CBCP
richard@drj.com

Vice President
Robert Arnold
bob@drj.com

CONFERENCE COORDINATOR
Patti Fitzgerald, CBCP
patti@drj.com

CONFERENCE REGISTRAR
Merce Knese
mercedes@drj.com

CIRCULATION
Laura Baugh
laurab@drj.com

INTERNATIONAL
CONTACTS
England: Thom Hetherington
Business Continuity
Phone: 0161-237-1007
thomh@tempus.demon.co.uk

Australia: Anthony J. Harvey
Journal of Business Continuity
Phone: 0011-613-953-0055-8
fax: 0011-613-953-0528
sector@notability.com.au

Japan: Shinji Hosotsubo
Quake Japan Co., Ltd.
Phone: 03-3215-2880
fax: 03-3215-2881

Brazil: Jose Carlos Ferreira
Disaster Recovery Mercosul
Phone: 55 11 3666-9506
conc2000@uol.com.br
ww.drms.com.br

Click Here for a Printable Version

New SAN Architectures Benefit Business Continuity Re-Harvesting Storage Subsystems Investment for Business Continuity

- by David A. Cozzens

In today’s data intensive business world, companies need to ensure the integrity of their critical systems by developing a comprehensive business continuity plan. Business continuity refers to the restoration of critical business systems in the event of a disruption that precludes operation in the production site. Although there are many approaches to business continuity, very few of them totally address the operational and financial needs of an enterprise that constantly must have access to its business-critical information.
The traditional method for business continuity has been tape backup, which enables companies to restore their applications and data from tape onto storage devices, configured like the storage devices in their production environment, at an alternate site and to re-load and re-start the applications to the point of failure. However, over time, it has become evident that tape recovery for very large databases may not be feasible, since transferring data from tape to disk can be a very slow process.
The concept of remote mirroring for data disaster recovery was developed to provide the capability of using remote telecommunications circuits to copy data in real-time, or near real-time, to an alternate site. Mirroring has been a major advance in business continuity, providing round-the-clock (24/7/365) businesses with the ability to recover their data quickly, eliminating or minimizing the number of lost transactions. However, while this does address the speed issue, the ability to remotely mirror data is complicated by the need to equip the remote site with storage subsystems that are comparable to those found in the production facility. In addition, vendors offering mirroring solutions push companies to convert their existing heterogeneous storage systems found in the production environment to the vendor’s proprietary technology to ensure interoperability with the alternate backup site’s hardware.

When establishing an alternate site for mirroring, organizations hesitate to invest in the latest storage systems, since they test them only a few times each year and truly hope they will never need to use them. Duplicate systems in both locations can become cost-prohibitive for many companies. To ensure successful recovery from business interruptions, most companies are either forced to purchase costly duplicate systems or to rely on lesser means of business continuity, such as remote journaling, standby operating systems or, at the very least, tape backup.

Figure 1

New Approaches to Storage Networks
As the amount of enterprise information grows at an accelerating pace - with analysts predicting that data within a production environment will average 100 percent growth each year through 2004 - new, more economical and efficient approaches to storage management and business continuity are required. Figure 1 shows the projected demand for network storage solutions in response to the increasing amounts of data.
Within the enterprise, there has been a recent shift away from direct-attached storage approaches - where a single host is attached to a storage device through the device chassis or externally through an I/O bus (usually SCSI) extension - toward storage networking solutions, like Network Attached Storage (NAS) or Storage Area Networks (SAN). NAS architectures attach storage devices through a network in order to provide file access services to interconnected computer systems, while SAN architectures build a separate network, whose primary purpose is the transfer of data directly between computer systems and storage. This shift in network architecture has created new possibilities for business continuity. These storage networks are emerging as the solutions of choice to meet the rapidly growing enterprise storage requirements for several reasons: reliability, availability, scalability, cost and performance.
Both NAS and SAN architectures enable increased storage capacity and improve flexibility, while allowing organizations to sidestep the cost and complexity associated with scaling expensive, direct-attached server infrastructures. NAS and SAN also enable storage managers to dynamically monitor and re-direct storage to servers that require additional storage. In today’s SAN/NAS world, where specific servers will only work with a homogeneous storage infrastructure, it remains difficult to mix the best and most cost-effective storage devices on the back-end and still have the flexibility to group all heterogeneous storage devices into a single, easily managed pool.

Leveraging the Storage Network for Business Continuity
A network architecture that has emerged to address this situation places an intelligent network storage appliance in the center of the storage network. (Figure 2) These SAN/NAS appliances come with standard host bus adapter (HBA) connectivity and can be quickly (1 to 2 days) dropped into any existing distributed systems environment to allow for heterogeneous, any-to-any connectivity. Now, any host can communicate with any vendor’s storage device. In addition, this appliance can be dropped into an existing infrastructure, regardless of current and future network protocols, such as TCP/IP, Fibre Channel, SCSI, InfiniBand and iSCSI.
Companies can now choose the most cost-effective storage systems for their production environment and move the previous storage into a business continuity role at an alternate site. Instead of having to write off the expense associated with discarding the storage subsystems that were in the production site as part of the effort to support and replicate a mirrored environment, companies can now leverage these prior investments. This model ultimately allows them to save money by extending the useful life of their equipment.
A further step to leverage the prior investment in storage systems would be to use the previous generation of storage hardware for application development, in addition to business continuity.

Figure 2

Network Storage Appliance Feature: Virtualization
The virtualization feature of network storage appliances also allows a company to plan, organize, direct and monitor all of their stored information from a single user-friendly console or GUI interface. Virtualization enables all of the corporate storage pools to be tied together, monitored and dynamically altered. As the production system’s storage pools are manipulated, users can simultaneously manipulate the storage pools at the remote site, all but eliminating the need to travel to the remote site to perform manual re-configuration.

Network Storage Appliance Feature: Security
A network storage appliance also delivers enhanced data integrity within the heterogeneous storage environment. Through the intelligence in the appliance, unique hosts have access to the complete storage pool through Logical Unit Name (LUN) Mapping of all the storage devices, depending how they are apportioned.
To illustrate this, imagine mixing Windows NT and Sun servers in a storage environment. One technical disadvantage of this is that NT identifies any attached storage as if it is its own. For example, when a company is operating both NT and Sun environments, they cannot mix the two storage subsystems together, because the NT system will grab the Sun storage, which can result in the possible corruption of data. To add additional storage, devices must physically be added to each server. Through the use of an in-band network storage appliance, however, storage managers can utilize internal LUN Mapping, and its associated security features, to show the NT system only the appropriate devices. This makes it possible to dynamically add or subtract storage as required by each operating environment, without running the risk of the NT environment corrupting any of the other subsystems. If the Sun storage pool is only 50 percent utilized, the storage manager can dynamically re-assign used space to the NT storage pool, which is currently operating at say 90 percent capacity, without having to physically disconnect and re-connect the storage devices, or install new hardware to satisfy the requirement. Storage requirements for heterogeneous environments can now be combined in the same SAN, even on the same physical volume, if desired.
It is important to note that hosts can see and utilize only data that is logically attached or mapped for its use. When a server communicates with the storage device, it executes a certain command set. The storage device understands this command set, executes against it and returns an expected result back to the server. The intelligence in the network storage appliance acts as a command-set translator. It recognizes the server commands, relays it to whatever type of storage device is on the back-end, executes the command, receives the results back from the storage device and presents a response in the form that the server is expecting. It is this feature that allows for the highly sought-after any-to-any connectivity.

Network Storage Appliance Feature: Mirroring
As noted earlier, in today’s business world, only very homogeneous, proprietary, single vendor technologies allow for remote mirroring. Not any more.
By inserting the network storage appliance, in-band, between the server and the storage device on the production side and then using a remote IP connection or direct Fibre Channel link to connect to another network storage appliance at the alternate site, companies can remotely mirror their data across heterogeneous storage devices. For example, companies can now mirror a high-performance storage array in their production environment with a JBOD disk configuration at the alternate site.
This model works because in-band network storage appliances intercept I/Os from the host and sends them out to not only the production storage subsystem, but also to the remote mirrored subsystem. The network storage appliance can mirror in either synchronous or asynchronous mode. It’s almost as if a dedicated I/O sub-processor has been placed in the data path within the distributed systems environment. The network storage appliance moves mirroring control, which can reside on the server (depending on the application) or the storage subsystem (depending on the storage vendor) and move it to an intermediate level, so that there is a discrete I/O processing capability in the data path. By offloading processing from the server and/or the storage, additional capabilities (system resources and cycles) on both the server and storage can be freed up. From a mirroring perspective, I/Os are completed to both the production device and the remote device.
From a technological standpoint, this mirroring is comparable to other mirroring systems available today, with the major difference being that the server and storage systems can be heterogeneous, enabling mirroring to be accomplished using a lower-cost storage pool in the alternate site.

Network Storage Appliance Feature: Point-in-Time Imaging
Due to the high costs of storage and the current need for a homogeneous backup environment, most companies currently test their remote site capabilities by using existing mirrored data. If an organization is testing on this existing production mirror and their actual production system goes down, that company is in trouble because the backup data is now in an unknown state, having been manipulated by application test scripts. Point-in-Time Imaging enables companies to run disaster recovery tests on consistent data without putting the status of current systems and resources at risk. By using this technique to create a virtual copy of the data, a company can keep the production mirror running, while testing on another copy of the data. This not only enables companies to test more frequently, but also enables them to sufficiently demonstrate their round-the-clock operations commitment to their customers.

Impact of Network Storage Appliances on Corporate Operations
As information continues to grow, organizations require additional people to manage all the disparate storage systems within their production environment. It seems obvious that organizations will want to reduce the amount of time and number of people needed to manage their information growth. By using the singular focused, user friendly console that comes with the network storage appliance to manage these disparate storage pools, the overall complexity of planning, organizing, directing and managing a company’s storage environment is reduced.
From a business continuity perspective, if an organization requires different people to manage the disparate storage pools within the production environment, they will need to accommodate these same roles in the recovery environment. Considering that the company hopes it never uses the recovery environment, that is a high cost to pay. Network storage appliances offer the promise of standardized skills associated with managing a company’s any-to-any storage pool. This also potentially reduces the complexity of managing the alternate recovery site, which can be accomplished remotely. This ultimately frees up staff time to perform other business-critical activities, which is vital today in the current job market when trained technical people are in such great demand.
The flexibility delivered by network storage appliances introduces many benefits into an organization’s storage environment and business continuity plans. The ability to extend the useful life of existing storage infrastructures and the associated cost savings are key. Organizations that are considering developing alternate sites, or upgrading their existing ones, need to study the economic benefits and the breakthrough features delivered by this innovative storage network architecture which moves intelligence to the center of the information infrastructure and reduces the complexities of planning and managing both production and alternate site storage environments.

David (Dave) A. Cozzens is currently the manager of applications and business continuity marketing at StorageApps Inc. He has more than 20 years experience in the information technology industry, specializing in the areas of acquisition/integration management, business continuity services, strategic planning and technology partnership management. Dave has extensive experience in the assessment, development, implementation and maintenance of technology and business continuity planning programs.