Staying alive

Business continuity planning is not an easy job, but someone has to do it. NME looks at how organisations in the Middle East approach the subject, and what the key issues are when considering how to keep a business going.

  • E-Mail
By  Eliot Beer Published  October 2, 2006

|~|flooded200.jpg|~||~|Every IT manager has a to-do list – a collection of projects that he or she would like to get done, or that need to get done but aren’t quite urgent enough, or that are going to get started any day now, just as soon as the manager gets around to it. For many Middle Eastern organisations, business continuity planning may be languishing somewhere on this list. “Business continuity is not a priority for enterprises; with the economy today, the priority is growth,” says Abbas Taher, product manager at UAE incumbent telco Etisalat. “From all the IT departments we see, their priority is to sustain operation and sustain the business. For you to do business continuity, you need to come to the subject with a very cold mind and you need the consultants in place.” It is easy to duck the issue of business continuity – as Taher points out, it takes a “cold” mind to consider all the possible eventualities that might befall a business, and how best an organisation should marshal its people and resources if they happen. But it is possible to express business continuity in terms of defined metrics. By working out the cost of implementing business continuity plans – say, constructing a second site complete with servers, and maintaining a high-speed data link between the sites – then comparing it to the cost of not having a second site in the event of certain scenarios – fire, floods, power failure – whatever is probable in the region under consideration. It then becomes just a matter of probability – if the chances that a disabling disaster will occur during the lifetime of the second site are sufficient that it becomes worthwhile to implement, then the issue of continuity simply becomes another business calculation among many. “Business continuity planning depends on exposure and available resources,” says Carrie Higbie, global network applications market manager for Siemon. “Some companies get together and offer to provide redundant equipment for each other. The newer trend is to outsource a disaster centre or a redundant data centre. This option, while optimal, is not always fiscally possible. “The idea is to provide 30 days worth of offsite operations, assuming that if your facility was completely destroyed you could find a suitable replacement in 30 days,” she adds. “With this in mind, important pieces of information and other tangibles should be stored offsite with your backups.” Networking infrastructure is a critical issue for business continuity. Where most plans rely on the presence of a second, remote site in case of a primary site failure, the question of how to link the sites together, at sufficient speeds to allow mirroring of all mission-critical data, is of prime importance. “What’s happening more and more, the majority of employees don’t work at the same site as the data centre – they work in remote offices,” says Pierre Langlois, general manager for Central and Eastern Europe, the Middle East and Africa (CEEMEA) at Packeteer. “So being able to establish quick links between the remote offices and the data centre is also important – making sure the same data is saved at each location is critical if there is a problem and data from one of the secondary sites needs to be used.” Packeteer, and other WAN optimisation vendors, can play a valuable role in ensuring only changed data is mirrored, saving valuable bandwidth. By using intelligent caching and compression systems, as well as quality of service, switching to remote sites in the event of a problem is also less painful, from a bandwidth perspective. “Data needs to be mirrored very quickly to a remote site; it’s not only a question of saving money, but it’s also a question of time,” Langlois adds. “If anything happens, even a data loss of a few minutes’ business can be very costly – millions of dollars, potentially.”||**|||~|abbas200.jpg|~|“Business continuity is not a priority for enterprises; with the economy today, the priority is growth.” Abbas Taher, product manager, Etisalat.|~|How much data, and how efficient the link required, vary greatly depending on the applications running and the level of redundancy required. For example, organisations looking to ensure completely seamless failover to a remote site will need to invest more in their link than others. Server virtualisation technologies, where two or more servers share or mirror a workload, are some of the most demanding consumers of network bandwidth. Marathon Technology’s virtualisation solution, as one example, requires 155Mbit/s of bandwidth with less than 5ms latency to run at full capacity, although it can still be used with as little as 55Mbit/s – still a significant amount for a longer-range link. “The majority of our deployments like that are in a campus environment, with multiple facilities in the same area, or where organisations own their own fibre, and have some dark fibre they can utilise,” says Nick Turnbull, European sales director for Marathon. The Marathon system is aimed at organisations with very high business continuity demands; while it virtualises two servers into one, it does not share capacity between the pair – it mirrors it. This means, from a cost perspective, an organisation has two servers doing the work of one. But because the servers can be separated geographically, if one goes down there is no loss of data, and the applications will continue running without any downtime at all. Turnbull explains: “An example I used with a customer recently – a global customer, with thousands of people travelling and using email – if they had a high-end clustered solution running in a building, and they lost that building, they would lose everything. With Marathon, you can have one server in the main facility, then several miles down the road have another server. If something happens to that first building, and the server is lost, no-one around the world will notice anything – the other server in the other building would continue to operate, with no performance hit.” Performance is one of the factors which organisations need to cost: what level of performance is it worth maintaining in the event of a problem. CA’s Abdul Karim Riyaz, regional director for the company’s storage and protection business, asks if performance takes too big a hit, is it really business continuity? Business continuity can be looked at as yes, the system is available, but the performance you’re getting is very low,” he says. “To me, that’s a lack of business continuity.” He says that bandwidth is one of the key infrastructure issues businesses need to look at, and that in the Middle East, it is one of the key weaknesses. Riyaz does see new initiatives coming from regional operators as well as new entrants into the telecoms market, and hopes to see progress in this area within the near future. “You can have the best software and the best processes in place, but if the underlying infrastructure doesn’t measure up to the necessary speeds, you will definitely have some kind of loss between switchover,” he says.||**|||~|higbie200.jpg|~|“Some companies get together and offer to provide redundant equipment for each other.” Carrie Higbie, global network applications market manager, Siemon.|~|Riyaz also comments that regional organisations have adapted to this state of affairs, by compromising on what they choose to keep running: “If you look at the business continuity and disaster recovery sites and the business continuity plans some regional organisations have, they do not have a ready-made, complete secondary site made available. What they have done is in some cases they have taken some of their key applications, and made sure they have a redundant site made available for them.” Etisalat’s Abbas Taher says that the telco is committed to cutting its prices, and says he believes that as far as leased lines go, Etisalat’s prices are comparable with rates around the world. No pricing information was immediately available, though. As far as Etisalat’s data centre offerings go, though, the telco has cut its prices dramatically: five years ago, the firm charged 180 dirhams (US$49) per megabyte – today it charges 20 dirhams ($5.45) for the same amount. The data centre approach is one which can make a lot of sense for companies looking to protect their business, according to Taher. Not only are their immediate cost savings from the point of view of an enterprise not having to maintain its own secondary site, but specialist data centres can also offer a much higher level of business continuity infrastructure than enterprises would otherwise be able to implement. By way of example, Etisalat’s two data centres – with three more coming – have multiple power feeds from different sources, as well as generators in case the main lines fail. The centres also feature high levels of redundancy throughout. But even if organisations do decide to invest in an outsourced business continuity solution, they still need to make sure their continuity plans are up to scratch. Taher says one of the most common mistakes he sees in regional enterprises is the assumption that backup is enough. “The problem is, if there’s a disaster you don’t just lose your backup, but you lose your equipment, and possibly your people,” he explains. “Assume a fire happens; when the alarm goes off, the IT staff won’t go to their department to retrieve the backup, they will get out of the building – life is more important than data. But this means you can’t do anything to restore your data. “To do a good business continuity plan, you need to consider people, operations, how much data will be lost since the last snapshot – everything,” he concludes.||**|||~|nbad200.jpg|~|The National Bank of Abu Dhabi currently has a second site for business continuity purposes, and is looking to leverage it further in the future.|~|The National Bank of Abu Dhabi (NBAD) recently implemented a range of business continuity systems and processes, including high-end storage from EMC. Hossam El Kobrosy, the bank’s technical services manager, talks about the firm’s approach. “Part of our strategy is protection of the customer data – financial institutes always look for security as much as we can. So we look at protection of data itself internally- to keep it offline, duplication, whatever we can do with the technology right now – helping our business continuity. Anything could happen, and we want to protect the financial data of our customers.” It is clear that NBAD and El Kobrosy, takes business continuity very seriously; the above comments demonstrate how both the bank and the individual have accepted that – despite everyone’s best efforts – things can and will go wrong. “Business continuity is a long process, and it’s not easy, but it starts from the data itself, the protection of the data – backup, replication, high-availability systems, remote sites, communication between data centres,” says El Kobrosy. “And then you do another layer to ensure continuity, to safeguard against manual mistakes, human mistakes. It does not need to be a full disaster that happens – you still need to have business continuity, you have to plan for all aspects of how to keep the dataflow going and services available to your customers.” Specifically, NBAD has implemented an external site for core operations, which the bank hopes to expand in the near future. At the bank’s head office the IT and operations staff have installed uninterruptible power supplies and generators, to ensure the primary site stays up as far as is possible. As regards the bank’s network infrastructure, NBAD’s technical services manager says it has deployed high-availability systems throughout its infrastructure. It also prioritises its traffic, making sure its business-critical applications are given the necessary amounts of bandwidth. Part of any business continuity plan is testing that the emergency procedures and systems do actually work; NBAD’s departments test all available systems by simulating a failure in one area, then checking other business areas for levels of functionality. El Kobrosy says there are still some systems which cannot be tested, but the bank is working to change this as soon as possible. On the importance given to business continuity planning in the Middle East, El Kobrosy does not subscribe to the view that the region lags behind: “I wouldn’t say at all that business continuity and disaster recovery are not taken seriously in the Middle East. It used to be the case, but since 2001, and even before with what happened to Kuwait, and since with Iraq, the knowledge has been transferred that we need to look after our data. “The growth of the economy has required this kind of business continuity, and we have to adhere to the new standards and the governance which are coming onto the market – the most astute organisations are trying to implement it ahead of it arriving in the Middle East.”||**||

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