Storage Expo 2007.

Features from exhibitors at Storage Expo 2007 the UK's largest event dedicated to data storage. Now in its 7th year, the show features a comprehensive FREE education programme and over 90 exhibitors at the National Hall, Olympia, London from 17th-18th October 2007

Clustered storage will break the NAS and SAN stranglehold 

Philip Crocker, Director of EMEA Marketing, Isilon Systems 

The rise of large unstructured data files is unstoppable. From high definition video on demand, to clinical data for medical research, many organisations are increasingly forced to work with data capacities that are reaching into the realm of Petabytes.

Both SAN and NAS technologies are widely established but were never designed to deal with terabyte sized files or high throughput access requirements that are emerging in the unstructured data market. The comparatively simple storage requirement of enterprise applications such as Microsoft Exchange, SAP and Oracle are much more the preserve of these technologies.

Clustered storage in contrast was designed from the ground up to deal with large files and capacities that scale into the Petabyte range and beyond. However, size is not the only factor, the overriding issue is the same as it has always been, namely cost.

Under the NAS and SAN ideology, the bigger it gets the more costly it is to implement, manage and fix in the event of the inevitable problems and there will always be problems! Hard disks fail, switches falter, data gets corrupted, irrespective of vendor, these issues happen and incur substantial costs.

Clustered storage has evolved from the clustered server space where instead of a massive mainframe processing all the application traffic, hundreds and in the case of advocates like Google--thousands of servers share the workload of requests, processes and delivery. Failures still happen but with clustering, the rest of the working servers just keep on ticking.

In a clustered storage environment, each node is independent and intelligent. No one node stores all the file data and each cluster has copies or parity bits to allow it to reassemble any file in the event of the loss of any individual disk, node or for ultra secure systems--several nodes. Nodes connect together like individual bricks in a wall, scaling from tens to thousands of terabytes.

However, lets not get carried away, clustered storage is not the magic bullet for everything. For database access, email servers and ERP systems, traditional SAN or NAS storage is a better option. For small data capacities under 2 terabytes, the upfront cost of clustered storage is probably slightly more expensive, not reaching its full potential until you get into the 10-15 TB region. These truisms may change over the next few years as a number of hybrid solutions currently only at the drawing board reach the market--but these are still someway off.

Clustered storage is also new technology and although it has some heavy weight customer such as Kodak EasyShare, Pratt & Witney and MySpace.com, like any 'new kid on the block', it has to have been around a while for certain more conservative customers to take the plunge ... but the water is looking increasingly inviting.

It's still early days for clustered storage but the take up has so far been phenomenal and the need within the market is growing even faster, for many industry watchers, it's not a case of if clustered storage will eclipse NAS and SAN--its simply a case of when!

Isilon Systems UK Ltd is exhibiting at Storage Expo 2007.

Are you covering your back end? Tips on securing your storage network

Andrew Wilson, Hitachi Data Systems 

The storage network is now of the utmost importance to any enterprise; not only does it hold essential data, the lifeblood of any organisation, it also provides the capacity for running business-critical applications and services. Unfortunately this makes the storage network a potential target for malicious attacks from outside the organisation and also renders it vulnerable to accidental damage from within the company.

Storage security has never been more important than at the present time, when business and regulatory compliance demand confidential data security. Yet while 72% of EMEA companies acknowledge that regulatory issues are an important driver in their storage investment, the issues that result from security breaches, such as financial loss and brand damage, can be just as serious as receiving a fine or legal action for non-compliance.

A carefully planned and well-executed security strategy throughout the enterprise is essential, so what should companies take into consideration to reduce the risk of data corruption and loss?

Don't assume your data is secure

It is a common belief amongst business and technical personnel alike that because the storage network exists far from the many entry points and is not on an Internet Protocol (IP) network, that it doesn't need additional security. This assumption is often what makes the storage network the weak link in the security chain. While your average employee may have little idea of how the storage environment operates or how to access key data, a malicious attacker will often take advantage of this attitude to the storage environment. Securing, hardening and frequently monitoring the storage system is crucial to prevent unauthorised individuals obtaining and potentially misusing valuable data.

Ensure your technical team understands the storage network

Storage networks are often looked upon as simple systems that merely provide data storage. This can lead to misconfiguration of the system making it vulnerable to accidental security breaches. Those in charge of the storage network need to understand how to secure all parts of the environment in order to prevent this. In a small or mid-sized business a lack of technical knowledge or training can result in perfectly adequate equipment operating without proper protection because one element has been left unsecured. In larger enterprises it is also often the case that one or two experts administer storage security but a number of other technical personnel have access to the storage environment to carry out other tasks. There are cases where otherwise well-secured storage networks have been compromised by a technical team member 'borrowing' a cable from part of the storage network, having assumed it was an insignificant test environment, taking down part of the service by accident and leaving the entire system vulnerable. Implementing a company-wide labelling system of cables and other vital equipment will enable the relevant employees to see what they are being used for and whether or not they are safe to remove.

Secure the management network

The management network can serve as the easiest point of attack within a storage system as this is what allows control of the storage network environment. Often it is a simple box that acts as a bridge between the storage network and the company IP Local Area Network (LAN) and it is frequently improperly secured at the IP end. The management network should operate at the same security level as other entry points, as well as utilising controlled access management and authentication procedures to make unauthorised use as difficult as possible. This will prevent an opportunistic attacker exploiting this common weakness.

Segregate security domains properly

It is now standard practice to separate and firewall the organisation's network into appropriate security domains, ensuring that data can only be seen by authorised personnel. Unfortunately in many cases the storage system is connected in an unprotected way to multiple networks throughout the enterprise. This means that a single attack on the storage system puts all networks at risk.

The solution is to install different servers and applications with different data sets and ensure that the storage volumes at the back end are protected from rogue applications and servers. If this is done incorrectly, a new box plugged into the network without the correct security installed may try to take ownership of the disks around it. This can potentially cause problems with overwriting and loss of valuable data.

Encrypt moving data

Encryption on disk storage is often a useful tool but many organisations are wary of the prospect of losing encryption keys and rendering their data useless. In fact, 43% of companies in EMEA admit they do not have a data encryption policy at all. Yet the risks of this are far outweighed by the benefits when data is transferred to a portable device or tape. The potential for portable data to be lost by the carrier and get into the wrong hands is not such a scary prospect if it is impossible for that data to be viewed or used. Encrypting moving data should be an essential component of any organisation's security strategy.

Align storage security with enterprise-wide security strategy

An effective security strategy will cover people and processes in addition to technology. Clear policies and procedures that are regularly enforced will encourage employees to realise that data security is their responsibility also. Only 53% of firms in EMEA report having an internal policy for the security of data stored on mobile devices. This is a surprising statistic given recent reports of lost laptops, disks and USB keys. Companies need to apply and enforce security strategies which cover data-at-rest as well as data stored on mobile devices.

Advocate skills and knowledge sharing

Often storage administrators and the security team exist as separate divisions within an enterprise. This can mean the administrators have little knowledge of security best-practice and the security personnel do not have a sufficiently in-depth knowledge of storage to be able to see the weaknesses in the network. This can be the result of a lack of training, a territorial attitude or simply a lack of contact with one another. Cross-pollination of skills and knowledge is essential to prevent storage being the weak link in the security chain and it is important for companies to promote these practices.

Electronic and logical security can be extremely effective in preventing malicious or accidental attacks on storage networks but this is only part of the story. Physically securing equipment is a frequently neglected part of storage security and some organisations have paid the price for this. Regularly reviewing storage security practices as part of the company's overall strategy is fundamental to preventing attacks. For a large enterprise, include a review of how many people have keys to the data centre and how secure the room is; for a small or mid-sized enterprise, check that the keys to the storage rack haven't been left in the lock. Taking time to remind all employees that an attacker doesn't need to get through layers of electronic security to get hold of data if he or she can simply walk in and take it will pay dividends.

Following these tips will help you protect one of your most valuable assets: the data your company relies on to complete its mission-critical activities. Bear in mind that implementing a set of policies that help all employees in keeping company data secure is just as important as ensuring your technical team has taken the necessary steps to secure the data electronically.

Hitachi Data Systems is exhibiting at Storage Expo 2007.