About Craig Group Ltd
Craig Group Ltd, a global shipping and energy services company, combines the strength of four specialist divisions with the depth of 75 years' experience in the areas of offshore support vessels, oilfield procurement, fishing and leisure. One of Scotland's top 100 companies, Craig Group Ltd employs more than 1,000 people, has offices in Aberdeen, Houston and Cape Town, and generates revenues of £110 million (US$163 million).
Challenge
As an international company with interests in shipping and the oil and gas sector, Craig Group Ltd operates a 24/7 business. Users log into corporate IT systems from the UK, USA and South Africa—so ensuring the availability of these systems is a key priority.
"Our IT systems are reliable and available in everyday operation, but we wanted to achieve a more workable and robust disaster recovery strategy," said Bruce Catto, Group IT Manager at Craig Group Ltd. "We had a standard overnight backup policy, but in the event of a disaster, actually restoring all of our systems onto new hardware would have been a time-consuming process, resulting in several days of downtime for our users."
When one of the company's server rooms narrowly escaped damage during a winter flood, the urgency of implementing a more practical disaster recovery solution became apparent. However, with a complex infrastructure involving both physical and virtual servers, the IT team at Craig Group Ltd thought that a simple, cost-effective solution to cover all servers would be difficult to find.
Novell Solution
Craig Group Ltd decided to consult Evision, one of its most trusted technology partners.
"Evision has been an infrastructure partner of ours for several years, and understands our business, systems and requirements very well," said Catto. "Before the meeting, my colleagues and I were discussing what our ideal disaster recovery solution would be—the outcome was a single solution that could copy both our physical and virtual servers offsite in real time. When Evision came into the meeting and presented PlateSpin Forge® from Novell, we were stunned: it was exactly what we had been discussing moments earlier, but we didn't think would be possible in a single solution."
Novell loaned a demonstration model to Craig Group Ltd for a month of comprehensive testing.
"We tested all the capabilities of PlateSpin Forge, and were impressed," said Catto. "We saw how easy it was to transfer images of our physical and virtual servers to the PlateSpin Forge and fail over to them. We also proved that we could restore workloads from the appliance to a new physical server, even if it was a different model from the original hardware. As a result, we decided to go ahead with a full implementation."
Within three months, Evision assisted the IT team to implement two PlateSpin Forge appliances and a Dell SAN at a data centre belonging to IFB, Craig Group Ltd's communications provider.
Data from 20 Microsoft Windows servers (some physical, others virtualised using VMware) at two of the company's Aberdeen offices is replicated to PlateSpin Forge according to predefined policies. Snapshots of the company's file servers, ERP system, vessel management system, Microsoft SQL Server databases and Microsoft Exchange servers are taken every hour. A Citrix environment and other systems that change less frequently are replicated once a day.
"We have now instituted a monthly disaster recovery testing plan, which will become part of our standard working practice," said Catto. "The tests are easy to run—it's only a matter of a few mouse clicks—and have been very successful so far. They give us, our users and the board of directors peace of mind. We have also recently been audited by KPMG, who were very impressed by the solution we now have in place, and the ease with which it can be tested and brought into production when required."
Results
Thankfully, Craig Group Ltd has never experienced a genuine disaster recovery situation—but the value of the PlateSpin Forge solution is still apparent to the IT team.
"The last test we did, we performed a fail-over to the PlateSpin Forge and asked users to log on from all over the world," said Catto. "Even under load, the performance was excellent."
Above all, the improvement in recovery time and recovery point objectives (RTO and RPO) is significant.
"With PlateSpin Forge, we can get all our systems up and running again within a few hours, whereas previously it would have been a time-consuming and extensive operation which would have taken several days," said Catto. "Moreover, we can restore to system images that are only an hour old, instead of having to revert to the previous night's backup and potentially losing up to a day's worth of data. In the event of a real disaster, our users would be able to log into the applications they need from home, and the business could continue almost as normal. For our company, that reassurance is absolutely priceless."