Dealing with disasters

Despite the political instability experienced by the Middle East in the past few years, disaster recovery occupies a lowly position on most IT managers’ agendas. However, for those yet to deploy such solutions, there appears to be a new option available to them. Rather than building mirror centres themselves, end users can now outsource their requirements to a third party, thereby cutting costs while gaining the protection they need.

  • E-Mail
By  Patrick Phelvin Published  April 1, 2004

|~|amin_M.jpg|~|Local organisations can lower their costs by outsourcing their disaster recovery requirements, says EMC’s Mohammed Amin.|~|A growing number of enterprises from around the Middle East are beginning to invest in the additional hardware, software and services required to create full disaster recovery solutions that provide them with the redundancy and back up they require to see them through minor mishaps such as power outages and full scale emergencies. A number of factors have encouraged this trend, from the ongoing political instability of the region through to simple market maturity. “The appreciation of what can go wrong and how this impacts on the business is driving disaster recovery adoption. For example, in Saudi Arabia, companies have been operating on a comfort level for the past 15-20 years but now there are more dependencies on the outside world and they can see how the internal situation is developing, they can see how a business can be totally wiped out due to a single incident and they are investing in disaster recovery accordingly,” says Noman Qadir, director of services & education, Computer Associates Middle East (CA-ME). “Over the last 12 months all the major government departments [in Saudi] have been investing large sums in disaster recovery policies and procedures and they have spent the major part the budget on back up storage systems and security,” adds Vernon Fryer, head of information security at Information Management Technologies (IMT). The Kingdom is not alone in ramping up its disaster recover activities and both the UAE and Oman have played host to large implementations in the past few months. Two of these have been carried out by Gulf Business Machines (GBM), which completed a full disaster recovery project at Oman Arab Bank (OAB) at the start of the year and has since finished creating a completely redundant IT infrastructure at Al Futtaim Group. The OAB solution not only protects the finance house against unplanned downtime but also acts as its core-banking network over which it can run both its existing and future applications. Comprising a customised local area network (LAN) and wide area network (WAN) stacked with Cisco products and a number of IBM servers, the solution has removed any single point of failure for the finance house and ensured OAB’s 500-plus users can access the data they require to service customers more effectively. Although the Central Bank of Oman’s mandate that all banks in the Sultanate should investigate some sort of disaster recovery solution spurred OAB towards its implementation, the bank’s desire to create a sound infrastructure for its applications and branches was the key driver behind the project. “The Central Bank of Oman has asked all Omani banks to think about disaster recovery. However, that was not the major force for us. The main reason is that OAB is the leader in corporate services and we have to have business continuity to make our systems available all of the time,” explains Hassan Abdel-Ali, IT consultant to the CEO at OAB. Although OAB has yet to experience a disaster, it knows its infrastructure is capable of accommodating any unplanned downtime as it has already carried out two live disaster recovery executions with GBM. “We have done two tests so far and it [the solution] worked on both occasions. We even carried out one test at a peak time and the switch over for the application, the servers and so on, was just 15 seconds,” says Abdel-Ali. At Al Futtaim Group, GBM built both a production and disaster recovery data centre. The former contains a highly available database cluster made up of IBM eServer pSeries 690 and 670 servers, Enterprise Storage Server 800 servers, six eServer pSeries 630 application servers and Tivoli Storage Manager, plus a 3584 Tape Library for backup. The disaster recovery site also comprises an IBM eServer pSeries 670 server and six IBM eServer pSeries 630 servers, along with a 3584 Tape Library and Tivoli Storage Manager for backup. Each site also has a high-speed Cisco switched network based on 6509 switches and the two data centres are connected using high-speed 7200 Cisco WAN routers and leased lines. The entire infrastructure has been designed to be highly available with no single point of failure, as in the case of a site outage the infrastructure will be switched over to run from the disaster recovery site. “This solution offers Al Futtaim a robust and highly available 24 x 7 environment to run its businesses and its critical trading business operations, such as logistics, will become operational around the clock,” says Ronald Franckaert, CIO, Al Futtaim Group. “This will improve customer service levels for the group’s end customers,” he adds. ||**||Defining a need|~|vernon_M.jpg|~|Top management buy-in is often missing when it comes to disaster recovery, says IMT’s Vernon Fryer.|~|Despite the apparent interest in disaster recovery in the Kingdom and a couple of marquee implementations in Oman and the UAE, many feel that the majority of organisations in the region are still stalling when it comes to evolving data back up procedures into full scale disaster recovery projects. The reasons for this are varied, but one of the most surprising is that some market commentators feel there is actually little need for full disaster recovery solutions in the Middle East. “Most companies do not need complete disaster recovery because they can deal with it in different ways through redundancy in machines and networks,” says Trevor de la Motte, GBM’s head of projects, professional services. “In most cases users go for the redundancy option because the chance of an actual disaster occurring is extremely low,” he adds. Others agree with de la Motte and suggest that if a firm backs up its data regularly and keeps it offsite then this should be sufficient to protect most local firms from going out of business should the unthinkable occur. “There is an argument to be made for assessing the requirements and assessing the risk and viability of different solutions rather than taking an ‘off the shelf’ disaster recovery solution,” says Stefan Niemitc, data management sales manager for the South Eastern EMEA at Sun Microsystems. “For example, doing something manually can fit with the volumes of data and the distances involved,” he adds. Another factor stemming the adoption of disaster recovery is a lack of awareness. While this typically applies to smaller businesses that are less IT savvy, it can also creep into the enterprise mindset, especially in companies that have still failed to realise the value of their data. “Almost all companies can give you a detailed ledger of the spending of company assets but none can give you a figure of the value of the information they depend on,” says Fryer. “IMT has conducted audits on customer disaster recovery plans and every time we find that the lack of user education, availability of skilled personnel and top management commitment is missing,” he adds. However, even if a local firm understands the need for a disaster recovery plan and feels it is necessary to implement one, there is still the issue of cost. While some, such as OAB and Al Futtaim, have taken the expenditure in their corporate stride, others struggle to justify the cost of investing in redundant hardware and the business process/consultancy work required to create an effective disaster recovery set up. “Disaster recovery is not cheap and the solutions can be expensive… Therefore companies can find it difficult to justify the cost. As a result, they first go in for a back up solution before heading into replication before going into clustering and full disaster recovery,” says Ancel Fernandes, executive vice president of marketing & audit at STME. Another cost, and potentially the most prohibitive due to the fact that redundant hardware can still be utilised in the production environment, is the expense of transferring data between sites, especially if a proposed disaster recover site is to be located in a different country. “One of the main blocks that has come up in terms of disaster recovery is the cost of connecting two sites together. Ideally speaking, people should be using fibre, but that cost prohibitive and it is out of the budget range for most people,” says Satish Shenoy, sales manager for Veritas Software in the Middle East. “The lines are expensive and the lines are not always reliable,” confirms Mahmoud Soubra, vice president of marketing & business development at Saudi Arabian ISP Awalnet. While the expense of international data communications does vary in the region, it is widely acknowledged to be too high everywhere. The main reason for this is the monopoly status of most data carriers in the region. “In the Arab world, transferring data is only cost effective if the offices are in the same country. Once you move cross borders, because we have the bottlenecks of cross border international gateways where all the data comms traffic and voice traffic is monopolised by one single operator, the costs of leased lines is much more prohibitive,” explains Jawad Abbassi, president of Arab Advisors Group. The expense of transferring data around the region for disaster recovery purposes has already prevented some companies investing in live mirror sites. For example, Niemitc reports that one of Sun’s local oil & gas clients looked into a full solution but decided that the cost of transporting the data was too expensive to justify creating a disaster recovery centre. “The actual communications infrastructure was so expensive that it did not justify the business requirements so they are effectively using tapes and bikes,” he says. ||**||Outsourcing data recovery|~||~||~|In addition to forcing local enterprises down the path of manual disaster recovery, the expanse of transferring data between production sites and disaster recovery centres has also given rise to the introduction of outsourced disaster recovery in the Middle East. Primarily the domain of hosting companies in the region, such a solution set comprises not just hosting but additional consultation, more services and extra guarantees. For example, in Saudi Arabia, Awalnet will mirror data for its customers. Within the UAE Etisalat’s internet data centre offers disaster recovery services for its customers, as does ASP Gulf, which offers to replicate clients’ mission critical data to a secure, remote data centre in realtime for a few thousand dollars a month. “There is an advantage in terms of cost to outsourcing disaster recovery because companies that outsource their data reduce their running costs by between 30% and 40%,” says Mohammed Amin, regional manager for EMC Middle East. Another company offering outsourced disaster recovery in the region is CA-ME, which recently signed a deal with global disaster recovery outsourcer SunGard and has been pushing its services ever since. “SunGuard was looking for a partner in the Middle East because they did not want to come to the market directly at this time. They were looking for opportunities in the region that would leverage their experience of providing business continuity and disaster recovery in the region,” explains Qadir. “Outsourcing disaster recovery means there is a recovery site and the customer would have certain parts of their business that are critical placed in the disaster recovery centre and from that they would be able to recover. The difference between SunGard and other providers is that it is more focused towards both the business and the IT,” he adds. Another provider currently focusing on the business side of outsourced disaster recovery is eHosting DataFort (EHDF). While the DIC-based firm’s hosting pedigree is reasonably well established, it is also promoting other services, such as its disaster recovery consulting and hot seat provision. “Technology is only one means of creating full disaster recovery capabilities and we talk to a customer and help them identify their business critical needs and how we can address their priorities. Hosting is just one of the services we offer and we talk about more than that,” says Kuldeep Bhatnagar, director of corporate business at EHDF. “We also have mobile connectivity so users can access their data even if their building is not there. In line with this we also offer disaster recovery seats, so that users can come to our facility and access their data if they need to,” he explains. EHDF has already recruited a number of clients for its outsourced disaster recovery solution and Bhatnagar names Dubai Bank, Doha Bank and Bank Muscat as just three of the region’s banks that have availed themselves of the company’s services. Another finance company to have signed up is American Express Middle East, which is using EHDF’s disaster recover hosting and connectivity, along with its hot seats. “Our agreement [with EHDF] helps us ensure that our customers will have continuous access to their accounts in a secure manner even in the event of a disaster,” says Trevor Stokes, president & CEO of American Express Middle East. However, while CA-ME and EHDF believe the outsourced model will be a profitable one — the former predicts that the KSA market alone is worth millions of dollars while the latter is building a third data centre to handle demand — others are less convinced. For example, Sun’s Niemitc suggests the region’s longstanding aversion to handing over mission critical data to third parties will slow adoption, while STME’s Fernandes suggests a combination of poor service level agreements (SLAs) in the region and end users’ desire to retain control will also hamper uptake. “By investing in their own assets and their own people end users get full control over their own environment. Also, I’m not sure that there are the stringent SLAs available in the Middle East from these outsourcing providers that the clients can go by,” he says. Whether outsourced disaster recovery becomes the norm or local users eschew the idea and find the money necessary to build their own mirror sites, one thing is for sure: As information technology plays an increasingly key role in the day-to-day business of companies and they come to understand the true value of their data, disaster recovery of some form will become paramount. As Fernandes says: “We have a long way to go in the Middle East in terms of disaster recovery solutions and organisations have started to realise that disaster recovery is not cheap. However, they are also learning that the benefits are significant should a disaster befall an organisation.” ||**||

Add a Comment

Your display name This field is mandatory

Your e-mail address This field is mandatory (Your e-mail address won't be published)

Security code