When your data is physically bound to a server, 50 percent of available server storage space isn't being utilised, 72 percent of overall storage space purchased isn't being used, and storage space can't be reallocated to other systems in need of additional storage, then you have a problem.
That was the view of Ann Raimondo, head of information technology at the Association of Teachers and Lecturers (ATL), an independent, registered trade union and professional association representing approximately 160,000 teachers, lecturers and support staff.
She's responsible for managing the IT infrastructure for the ever-expanding organisation, including deploying equipment, IT support and training for its 150 employees. It has offices in London, Belfast and Cardiff, plus a large volunteer base of remote workers throughout the UK who need IT systems and support.
Raimondo could see that ATL needed a robust infrastructure that would allow increased efficiency, flexibility and responsiveness by effectively connecting resources to business needs. The organisation also needed a solution that would allow for simple disaster recovery in the event of downtime.
Raimondo and the team looked to fulfil these needs with a virtualisation platform because of the technology’s ability to reduce memory and hardware dependency while increasing cost savings and business continuity abilities.
Specifically, ATL chose VMware virtual infrastructure because the product was already proven in the industry as delivering these benefits and making an immediate impact on technology and business processes.
"With VMware virtual infrastructure, we can now maximise the efficiency, flexibility and responsiveness of computing resources by consolidating our hardware footprint, reducing operational expenses and capital costs, and supporting an adaptive infrastructure," says Raimondo.
"With the results we’ve already seen, we’re very confident that VMware virtual infrastructure will continue to impact our organisation and allow us to operate effectively and efficiently."
The implementation resulted in the following benefits:
- Server consolidation. ATL consolidated from 22 servers to 11, reducing hardware requirements and costs by 50 percent.
- Flexibility and responsiveness. Prior to bringing in ESX Server, deploying a new server would require approximately three weeks for sourcing, ordering and implementing hardware. With VMware virtual infrastructure, this same process takes less than one hour.
- Lowered the cost of disaster recovery. The hardware independence of VMware virtual infrastructure helps mitigate failures caused by hardware and enables recovery from a disaster in a matter of minutes, matching and improving on user downtime expectations.
Avoiding a data recovery disaster
As a membership- and volunteer-based organisation, the impact of downtime for ATL is especially devastating and can result in significant operational difficulties across the organisation. ATL therefore wanted to ensure that it would receive the highest level of availability and disaster recovery -- but at a low cost.
ATL is no stranger to disaster. In 2004 alone, it faced air cooling problems, UPS failures and roof leakages, as well as melted tape drives, the failure of 15 36GB drives and signs of corruption in the document file server and Exchange system.
"The year 2004 could only be described as a year of near disaster," says Raimondo. "We felt we were on borrowed time and pushing our luck. To ensure that we didn’t go through another year like that, we embarked on extensive research and planning to design a new, robust infrastructure. That VMware virtual infrastructure would provide low-cost and advanced disaster recovery capabilities was crucial and has since ensured that we are not vulnerable to disasters of any nature."
The reason it's cheaper is that its virtual infrastructure is hardware-independent, so any physical server can serve as a recovery target for any virtual machine so repurposing under-utilised servers for recovery targets and disaster recovery testing saves the cost of duplicating all your hardware.
Shortly after implementation, VMware virtual infrastructure proved itself. On 19 July 2005, a power surge hit ATL, and it lost all power to its headquarters. ATL’s old UPS systems did not give the organisation sufficient time to shut down the systems cleanly, and member service and support was suffering as a result. "The impact on our organisation was serious," says Raimondo.
ATL’s IT team stayed late into the night hoping power would be restored so they could assess the damage and start the recovery process. When power was finally restored the next morning, ATL’s membership system was restored within 20 minutes. As the membership system had already been migrated onto a virtual platform, ATL was able to power on the ESX Server host system, do a file system integrity check and power on the virtual machine.
In contrast, ATL’s Exchange Server was not running in VMware virtual infrastructure, so it required numerous installations, configurations and restoration processes, which resulted in an entire day of system downtime along with dedicated resources from several key staff members.
"As head of IT, it is imperative that I choose to implement cost-effective and resilient technology solutions to ensure our organisation is not only making the best use of resources available, but that we are also providing our members and volunteers with quality service in order to retain and recruit members," says Raimondo.
"Perhaps the most significant benefit for me personally, however, is the ability to sleep well knowing our data and applications are in safe hands with VMware virtual infrastructure."