The British race team has used NetBackup since 1997 when it was sold by Open Vision, and backs up to tape and a virtual tape library (VTL) in its Enstone, U.K.-based data center. Lotus began complementing that protection this year, first with a Symantec Backup Exec 3600 appliance at the track, and then with two NetBackup 5220 appliances in the data center.
Auto racing is a data-intensive business. According to CIO Graeme Hackland, Lotus F1 generates about 25 GB to 50 GB for each of its 20 races per year. It also must protect data generated at the factory during tests and simulations. One high-resolution camera inside a car's gear box tracks gear changes and generates 12 GB for six seconds of footage, for example. The team also stores back office data needed to run its business.
Hackland said Lotus F1 has about 500 TB of storage on NetApp FAS filers. "For a small manufacturing company, that's huge," he said. "There are lots of areas where we manage vast amounts of data."
Lotus F1 has about 550 employees, and approximately 45 IT people accompany the racing team to the track for races. The team also ships storage arrays to each race, but problems during shipping to the Singapore Grand Prix in 2011 almost left Lotus without IT services during the race.
That gave Hackland the idea to buy a Backup Exec appliance to use at the track. The team put its first integrated backup appliance into service last summer.
"We needed something lightweight with a small footprint at the track," he said. "We're using 1.8 TB. Without dedupe, it would be more than 5 TB."
Previously, the team had to carry spare servers for each application used at the track. And each server had to be maintained with the latest software.
"Before the appliance, we had no backup at the track," Hackland said. "Now we can back up and recover VMware at the track, and that gives us a bit of disaster recovery. We hold the original images and as we go through the race we're taking backups of all the images."
He said the easy installation (an hour to unbox the unit and begin backing up) and outstanding deduplication ratio on the Backup Exec appliance prompted him to add the NetBackup 5220 appliances to the data center to protect around 100 virtualized Microsoft Windows servers. Lotus F1 maintains its Backup Exec 7.5 standalone software, but Hackland said he has greatly reduced reliance on tape and its NetApp VTL.
"With the appliance, you don't have the operating system or the media server to maintain," he said. "And they're easy to set up. We're starting to use our VTL less now that we have the NetBackup appliance. We can keep more of the backups on the appliance, and then push it to disk and then put it to tape. It's incredible how few tapes we buy now. I used to buy hundreds of tapes every quarter. Now we only place an order once or twice a year."
Hackland said his team still does monthly and annual backups to tape, but daily and incremental backups go to either the VTL or appliance. "We're expecting we will in the end move everything to the appliance," he said. "We were being a little cautious, because we hadn't used an appliance from Symantec before."
Hackland said in the last two weeks Lotus F1 backed up 3 TB of data to the NetBackup appliances that would have required 39 TB without dedupe. "I was staggered by that number," he said. Lotus F1 backs up a lot of data that dedupes well -- VMDKs along with race car design data that has minor changes each day.
Hackland said Symantec's built-in intelligence for handling virtual machines (VMs) is also a great benefit. His NetBackup appliance has policies to automatically protect new VMs and track which VMs have been deleted. "So it isn't continuously polling trying to back up deleted VMs," he said. "That reduces administration time."
Lotus F1 ships NetApp arrays to whatever track it is racing at. The scare at Singapore last year came after the storage arrays were damaged in shipping and didn't work after they arrived. Hackland sent an engineer to Singapore with spare disks, but by the time he arrived, NetApp's service team fixed the problem. Disaster was averted, but it left the team thinking how vulnerable it was.
"It gave us a real fright," Hackland said. "We had virtualized and taken backups of all machines, but those backups were back here and there was no way we could get them to Singapore in time. We ran the risk of not being able to provide any IT services for the race.
"I think we had been a bit complacent because all the physical servers have their own hard drives and it's easy to swap one out if something goes wrong. But when all your VMs are sitting on one piece of storage … We had stopped mirroring storage, so we went back to that and added the backup appliance to keep us safe."