Knowledge Base Article

Tech Support Cracks the Case: Sleeping on the Job

Symantec channel partners are dedicated. One such contractor called in to Symantec support when she was having difficulty upgrading a local government site on the West coast to the latest version of Backup Exec (BE). It was her responsibility that Thursday to upgrade the customer's 20 servers from BE 11d to 12.0, migrate the data, and then run over 100 scripted backup jobs. The problem was that although the BE upgrade seemed successful; the 100 backup jobs previously defined were gone. The library was missing as well. She spent most of Thursday night on the phone with two Symantec support engineers attempting to restore the jobs and maintain the upgrade. "If I can't restore these jobs," she said, "I'll have to recreate them from scratch over the weekend." Losing patience—and sleep—she built a make-shift cot in a cubicle at the client site and tried to get some rest as Symantec support escalated the case.

Steven, a veteran BE support specialist on the East coast, contacted her the next day. "She was upset, she was tired," Steven recalled. "If you have to escalate cases like this, people tend to get upset, which is understandable." However, Steven was able to talk her down and get her back on track to begin solving the problem. "I just need to get the main server up and running," she pleaded. "If I have to stay here all weekend to get the jobs recreated, I will."

Tools of the trade

Steven took things one step at a time. "Our first goal was to get that main server up and running and recognizing the hardware," he said. Since the site was working in a shared storage environment, the parent control admin server had to be properly configured before upgrading the child servers. Up to this point, she had worked with the other techs to re-install BE with no success. Her jobs and library were missing. "The previous support technicians had tried one method of fixing the problem," he said. "And in the process she lost her library and backup jobs." Steven wanted to try a different approach. "What I tried was less obvious." Steven knew that BEUTILITY.EXE can make manual changes to a BE installation that are not accessible during an automated installation. He suspected that the parent server had been upgraded when backup services were still running, causing an install of default user-mode device drivers. In Windows 2008, these default drivers might not detect the hardware correctly. Using Registry editing, Steven walked her through changing a Registry key to install kernel mode drivers. After a reboot of the system, BE finally recognized the hardware and her jobs. She was ecstatic, according to Steven. Once Steven was satisfied that the parent server was working, they used the BEMIG.EXE migration tool to migrate the data catalog and database. Thankfully, this ran without errors. The library and jobs were intact.

Home for the weekend

Working through the upgrade for all machines, Steven wanted to make sure that the recovered jobs would run. By this time, he'd been on the phone, off and on, for most of the day with the contractor, and she had been on site for almost 36 hours. "I stayed well after hours to get her up and running," Steven said, "But I was working from home that day and didn't need to buck traffic or anything. I figured it was the least I could do to help her out." By 7 p.m. Eastern Time, all seemed well. Both were glad to see progress. As the first jobs ran successfully she said to Steven, "If you were here right now I'd give you a big hug." She had her weekend back.

Published 16 years ago
Version 1.0

Was this article helpful?

No CommentsBe the first to comment