• Home
• Articles Archive
• Features List

• Media Pack
• Contact Us
• Subscribe

Have you got the right protection?

From STORAGE Magazine Vol 7, Issue 4 - June 2007

As enterprises increasingly seek to include remote global locations in day-to-day business operations, the need to centralise and consolidate data and applications becomes paramount, writes EDITOR Brian Wall

As enterprises continue to expand, the challenge of gathering and centralising data generated at widely distributed locations - and making that data consistently and reliably available to everyone across the enterprise - grows along with it.

This pursuit of global data availability and business continuity planning - not to mention regulatory compliance - is driving the need for more efficient data backup and replication strategies. And with that also comes the imperative to ensure that your data is always safe throughout the replication process. But how do you ensure that, too? What are the risks? And how do you minimise them?

So, in response to all of these challenges, what exactly is a business looking for when it asks for data replication these days? According to Stephen Watson, HP Storage Works Division product marketing manager, that request may not be activated by a technical enquiry, but rather one based on business drivers, which will often include the need for better customer satisfaction or increased revenue.

"Also, an IT manager may be tasked with increasing efficiency and reducing IT costs to maximise margins. Nowadays, translating IT solutions into tangible business outcome, or vice versa, can prove to be a demanding and hard task. Essentially, there must be a review of systems in place to allow for an assessment of what will be needed, and then analysed for what must be added or replaced to achieve the desired business goal. This has to be done for each business. The solution is, after all, bespoke - one that must be moulded to the needs of a business and not a copy of what others have done or think is necessary."

The data contained, or which will be contained on the systems, must also be examined, he argues. A series of questions must be answered before the appropriate system can be introduced. For example: will the data be crucial to the operation? Or can it be stored in more cost effective ways that will allow for its restoration, but perhaps not as a priority? Or will it be an 'instant-on' system that does not allow for downtime, which will start to cost the business money and orders immediately?

"Once the answers are generated, only then can the architecture be built around these information requirements," adds Watson. "Those data, files and information not critical to business may be placed on to tape-based backup systems. However, the information critical to the ongoing operation of the business may be replicated across systems, sites or geographies many kilometres or miles apart.

"Replication can reduce or remove risks, depending on how it is implemented. If local replication is performed, will the original reason for an outage also affect the secondary data set? In such an instance, it may be reasoned that a second site needs to be farther away to mitigate such risk. Such decisions will change the cost structure of a solution, so ultimately commercial restrictions may be placed on such a design, and impact the overall business and project outcomes."
One of the major concerns when replicating data is corruption. The bad news, of course, is that, if data is corrupt at one site when replicating, the corrupt data will be transferred across to another site.

"This will occur, regardless of whether deliberate action is taken to damage the data or not and the type of replication process used," says Alan Laing, CA technology specialist, storage. "It is therefore essential to use data technology that also has the ability to rewind, back to the point prior to corruption. Continuous Data Protection (CDP) is arguably the most effective way to reduce this risk. By using asynchronous replication, CDP makes it possible to roll back to the transaction level or back to a given point in time prior to when the corrupt data was transferred.

"Providing real-time replication of all data, using CDP technology, helps to accelerate recovery time and minimise data loss. This is the new approach that businesses should be taking to ensure safe and effective replication, yet the majority of companies still rely on traditional backup processes, which put companies at greater risk from a recovery management perspective by creating a significant time lag between Recovery Point Objectives [RPO] and Recovery Time Objectives [RTO]."
The ability to conduct a complete disaster recovery test plan, whilst providing full business service, should be an integral component of any replication solution.

Assured recovery
or automatic disaster recovery testing temporarily stops the replication process between different sites. This enables businesses to test all data automatically on a chosen remote site, without disrupting business function.

"The primary site will continue to process data," adds Laing, "and, when the process restarts, continuous data protection will be re-established. This helps businesses to maintain best practice processes and to meet the growing number of compliance regulations."

The real-time, high-availability capabilities of CDP are an important differentiator when replicating data, he stresses. "Once established, companies no longer face the prospect of losing half a day's data by relying on overnight backup; by rolling back with CDP, companies should only lose a couple of seconds of data.

However, most businesses still focus on building the disaster recovery infrastructure, rather than how to get business applications back up and running following disaster recovery."

CA is still one of the few vendors that include CDP technology in all software modules with replication capabilities. However, businesses are starting to change their approach with the development of traditional disaster recovery plans to include recovery point and recovery time objectives, and wider adoption has been further encouraged by the evolution of CDP technology.

Companies distributing large mission-critical data sets to multiple shared storage systems and to teams in multiple locations clearly need intelligent and reliable storage replication solutions that also drive down costs through greater simplicity and ease of use. And the solution must be able to move mission-critical data quickly and efficiently throughout a globally distributed network to ensure the highest levels of availability.

"Until now, traditional storage systems and host-based replication tools have been used to address these challenges at great cost, complexity and with inadequate feature sets," asserts Phil Crocker, director of marketing, EMEA, Isilon Systems. "One-size-fits-all block level replication technology, long the industry norm in storage systems, fails to provide customers with the ability to replicate files by file type, file size and geographic location, or to enable customers to easily manage bandwidth utilisation and cost."

An alternative approach, he suggests, is replication software that provides file-based, asynchronous data replication across clusters of storage devices. "Businesses can perform replication jobs either on the LAN [internal to a data centre] or over the WAN [external data centres]. Only files that have changed are replicated to the target clusters, so transfer times are improved. Users can also limit the amount of network bandwidth for any given replication job, which allows the scheduling of different bandwidth levels during defined windows of time, according to network availability, utilisation and cost. Replication jobs can be tuned to utilise as much or as little of the total storage cluster resources as needed, balancing the speed of replication with the performance of the applications and the demands of users accessing the cluster."

By adopting the replication software model, businesses can quickly and reliably replicate their large stores of valuable unstructured data to always-on, disk-based disaster recovery or backup sites and to distributed clusters, while administrators gain a powerful set of tools to optimise remote replication that can save time, increase storage capacity, and minimise bandwidth utilisation and associated costs.

"A specific example of this solution is Isilon's SyncIQ replication software," adds Crocker. "SyncIQ is the only replication tool that leverages Isilon's OneFS operating system software and clustered storage architecture, enabling replication jobs to be parallelised and evenly distributed across all sending and all receiving nodes in the Isilon IQ storage clusters. SyncIQ replication software provides data-intensive businesses with a powerful solution for file-based replication over the WAN and LAN."

As organisations evolve across national borders, reinvent themselves in the remote office and travel the globe on the next generation of mobile PDA phone, we have seen an increased reliance on technology, particularly email, to capture and consolidate critical business information. With business critical information and intellectual property tied up in these dispersed environments, the need to centralise and consolidate data and applications has become an important business priority across the globe.

"In order to support today's 24x7 organisation," says Guy Bunker, chief scientist, Symantec, "the process of protecting data is becoming more complex. Effective data availability and protection is critical for businesses to survive. The good news is, advancing technologies, especially around data replication, are here to help.

“Organisations need to start taking an increasingly measured approach to ensure they get the right data protection solution for their specific business needs. It could be argued the most important piece of advice for an organisation to remember is that 'not all data is equal'. Before selecting a solution, it's vital that business decision makers sit down with the IT department and consider, 'what is the value of this data and how much is it worth spending on protecting it?'"

Forward-thinking organisations are taking a methodical approach to data replication as part of their data availability solution, he says. This involves mapping out business processes, identifying where data is stored and evaluating how important it is. "As the office evolves beyond the desktop to the remote office, the laptop and the PDA phone, it is imperative that any solution addresses data usage in all these locations - the 'where' component is a key part of this equation."

Ensuring compliance regulations are met is more complex. "With enterprises generating, storing and replicating data across geographic borders, it is important that the regulations of individual countries are understood - thereby guaranteeing that regulations are not breached and specific compliance needs are met. With all the requirements in mind, an organisation can gain an insight into what solution or solutions are right to achieve the business objectives."

There are various types of replication and continuous backup solutions in the market, all of which have different costs and benefits associated with them. "Storage vendors, including Symantec, often provide specialist consultants to assist organisations in this process," says Bunker. "Once the need has been identified, organisations can choose from a solution to meet these needs, which can range from a simple backup, through to a comprehensive synchronised data replication and clustering solution.”

Essentially, storage-based replication combines the best aspects of host-based and appliance-based solutions. The application servers have little or no knowledge of the replication; there is no overhead on the application servers. Management is centralised and any host supported by the storage system can use the replication functions of the storage device.

"Unlike appliance-based solutions, there are no extra SAN switch ports needed to implement storage-based replication," states Jim McKinstry, senior systems engineer at Engenio Information Technologies.

"Since the replication is native to the storage controllers, the impact is minimal to the application servers utilising the storage. The only drawback to storage-based replication is that replication can only take place between homogeneous storage systems. In the past, this could prove to be costly; but today, many storage devices allow replication from more expensive Fibre-Channel drives to less expensive SATA drives and also support remote replication from a higher-end model to a lower end model. For example, a StorageTek FLX280, with a mix of FC drive and SATA, can perform snapshots of the data on the FC drives and store them on the SATA drives, and also remotely replicate the data on the FC drives to a StorageTek FLX240 that may have only SATA drives."

As he acknowledges, implementing replication can be a costly proposal, but can usually address an important business issue or multiple issues that make it is easy to justify the purchase. ST