For a company that supplies some of the leading bread brands in the UK and sends products to bakeries around the country on a daily basis, having a fail-safe order processing system is a must.
Allied Bakeries claims approximately a one-third share of the UK’s bread market. It makes Kingsmill, Allinson, Sunblest and Burgen, as well as a range of morning goods, rolls and part-baked breads to retailers, at its 13 bakeries situated around the country. All orders are handled and dispatched daily through an electronic processing system, which must remain up and running 24 hours a day.
Loss of trade
Dave Rowland, head of infrastructure at Allied Bakeries, says: “If the order processing procedure were to fail, it would result in immediate loss of trade and reputation, not only for the group and individual bakeries, but also for the hundreds of retail outlets that we supply. This, coupled with the short life-cycle of our products, means that any downtime in the system – notably server failure and the ensuing loss of data – is simply not a viable option.”
In the case of Allied Bakeries, its IT Shared Services Centre and each bakery within the company has a number of servers and back-up servers. These are installed for order process control and other day-to-day, critical applications.
This means that Allied Bakeries has a network of around 200 servers, which all need to be running almost continuously, all of the time.
Allied Bakeries felt that it lacked a coherent IT strategy for cases when servers failed and so proceeded to look for a remedy. It required a technical solution that would enable it to develop a bespoke server and data back-up system, covering all 13 bakeries and its IT Shared Services Centre based in Liverpool.
To mitigate the risk of a breakdown in the order processing system, Allied Bakeries
decided to seek a back-up data replication process. This ensures that a contingency plan can spring into effect in the event of a system failure, without the loss of any vital information.
It also enables the company to resume normal operations within minutes of a break down, without any risk to its performance and reputation.
After looking at a number of options, Mr Rowland chose NSI Software’s Double-Take – a data replication product that is supplied exclusively to the European market by Worcester-based Sunbelt Software. Double-Take provides continuous real-time data replication and server failover, which means if one server fails, data can be switched over to another.
The product is one of Microsoft’s preferred data replication engines, due to its ability to replicate vital data in real-time to a back-up or failover server located anywhere in the world, says Sunbelt
According to Mr Rowland: “After evaluating a number of data replication products, we found Double-Take to be easy to manage and applicable to all our Windows platforms. It just seemed to work out-of-the-box – it did what it said it would do.”
Benefits of the Double-Take system, according to Sunbelt:
- Reliable, real-time replication
- No file pre-scan required to identify changed files (after initial mirror)
- Preservation of file level security
- Open file handling
- Effective replication between systems linked by low band-width connections
- The ability to replicate between non-identical hardware and different operating systems
- One-to-one, one-to-many and many-to-one replication capability
- Minimal impact on performance of source and target systems
- Cost effectiveness
Sunbelt also provides specialist knowledge and support for Double-Take, and, in the case of Allied Bakeries, was able to adapt this to deliver the right solution for the baking firm.
Sunbelt’s UK sales director Ian Masters says: “It was clear from the outset that Allied Bakeries required a bespoke solution, so we spent time training the relevant staff on the technology behind these systems (SQL and Biztalk) and gave them the knowledge needed to design their own solution around Double-Take.”
Allied Bakeries’ Mr Rowland adds: “Sunbelt provided excellent evaluation, testing and deployment of Double-Take, as well as being very supportive throughout product training. The way we are using the system at the moment is that each secondary server monitors its primary server and, if it sees a problem, it will raise the alarm with the network centre. The operator will then decide whether the primary server is going to fix itself and make a decision to failover or not. This helps the overall management of all servers and acts as an insurance policy.”