Delfont Mackintosh Theatres (Delfont) own and manage 7 London theatres and sells tickets for these and other theatres. Delfont is committed to providing the best facilities possible and to deliver the best possible service both front-of-house and backstage. A key element in their ability to provide this service is the capability of their technology. Delfont has relied upon Managed Networks to ensure their IT is dependable since January 2002.
With nearly 200 staff using desktops and with the business open seven days a week, Delfont requires predictable and consistent access to its IT applications. This has been delivered using clustered servers and a fat client for the bulk of the services, including Microsoft Office, Adobe and Sage suite for payroll and accounts. In addition, the Box Office application, ENTA from Galathea, is delivered using Citrix. A Disaster recovery site mirrors the primary and is able to be up and running within 4 hours. Managed Networks ensures the delivery of this IT, providing a Service Desk for not only the infrastructure but as first line for all of the applications.
With many of the servers coming towards the end of their warranty periods, Managed Networks, as a part of the normal account management process, suggested that Delfont should review their infrastructure set up and needs. It was agreed that a project should be undertaken to:
1. Enable easy deployment of new servers and therefore of new applications, to increase the flexibility and speed of response
2. Reduce the level of support maintenance and minimise ongoing support costs
3. Reduce the overall physical footprint
4. Ensure that the current resilience and redundancy stays intact
5. Maximise the return on investment
Managed Networks designed a solution centred on a high availability blade-based system and fibre switch SAN. Microsoft’s Hyper-V virtualisation technology allowed us to deliver multiple guest (virtual) servers on a reduced number of host (physical) servers to replace those that were at end of life. The easy deployment for new applications meant that, for example, the transport server for Exchange, which needs to be done as an independent server and not as a cluster, could be deployed on the blades reducing the need for a separate box.
To deliver this, Managed Networks virtualised the infrastructure on Dell’s M1000 blade enclosure and M600 blades, with Dell Equallogic AX4-5 storage arrays. Two Blades are running as the file and mail cluster and four are running Hyper-V to host all other virtual servers. In addition, a SAN for storage space has been installed with a backup replicating the data continually and then removed each night. This configuration will then replace the existing servers as and when they go out of warranty which will happen over the next five years.
Phase One was the installation of the core infrastructure. Managed Networks achieved Phase 1 by building the hardware on the Managed Networks site and stress testing in their controlled environment. The built system was then relocated to Delfont’s data room and racked up. This only required about one third of a rack and was accommodated within the existing racks by removing some of the end of life servers. The system then ran in parallel.
A key objective was to ensure no down time for users during the transfer. The system is always in a live state and is used during both the weekend and normal business hours. Bearing in mind the nature of their work and the Box Office opening hours, the only window was on Saturday nights from 10pm for twelve hours. So Managed Networks used this time and split the key transfers to live over two weekend shifts. The total programme, including design, procurement of the hardware, build, test and go live took four months. The final review included Managed Networks drafting a questionnaire to enable Delfont’s IT manager to get user feedback.
Subsequent Phases will occur at regular intervals as each old server goes out of warranty. However the next Phase of the project is currently underway. This is to update the Disaster Recovery site to reflect this change. This will be achieved by redeploying the out of warranty hardware to deliver a cost effective yet robust facility.
The deployment of the core infrastructure is complete and stable. To date, the project is ahead of schedule. The system has been tested and redundancy features current and newly introduced have all tested correctly. Subsequent Phases are now underway.
Like for like hardware replacement would have taken significantly longer with greater project cost. The project delivers over a 10% cost reduction over a five year period. This required a small investment in the first year, but by the end of the second year, this investment will have been fully recovered and cost gains achieved. In addition, there will be further benefits from the enhanced flexibility in deployment of new applications.
Ongoing maintenance costs have also been reduced by approximately 20%, with less onsite maintenance required and an overall reduction in the time required. Similarly, space required has been reduced with much greater capacity fitting within the existing physical facility. The project objectives have been one hundred percent achieved.