Reporting tools address backup problems

By Sonia R. Lelii

It's no secret that backups have become a high priority among IT executives. But, for the most part, the process of backing up data to tape still leaves storage administrators guessing whether their backups have been successful.

According to analysts, the instability of backing up and restoring data is so dismal that it ranks as one of the two biggest complaints among storage administrators (along with the amount of time it takes to provision volumes). In fact, analysts report that on average 40% of backups fail for a variety of reasons.

"Companies are living in the dark, not knowing how many incremental backups have been successful until they go to recover data," says Arun Taneja, consulting analyst and founder of the Taneja Group consulting firm. "The problem is huge and the big vendors are not addressing it well."

But a number of smaller vendors are shipping software tools that address users' backup monitoring and reporting issues. Examples include Bocada, SysDM, and Tek-Tools, which have designed what some analysts describe as storage resource management (SRM) for tape. These reporting and monitoring products delve deep into the backup infra-structure to collect data from multiple points. The software then puts the data into a common format so that storage administrators can quickly isolate problems that are hindering backups as well as identify ways to get better utilization from their backup software.

Support for backup applications from different vendors varies, but some backup reporting tools support a combination of applications from leading backup software vendors such as Computer Associates (ARCserve), EMC/Legato (NetWorker), IBM (Tivoli Storage Manager), and Veritas (NetBackup, Backup Exec).

For example, Tek-Tools' Profiler for Backup, which is a module in the company's Profiler Rx software suite, supports backup applications from Computer Associates, IBM/Tivoli, Syncsort (Backup Express), and Veritas. Profiler for Backup can also be used to report on StorageTek's ACSLS.

Like some other backup reporting tools, Tek-Tools' Profiler for Backup confirms the success (or failure) of backups and reports on backup windows, how much and how often data is being backed up, how long backups take, as well as other troubleshooting, trending, and forecasting functions. Profiler for Backup also monitors network components such as hosts, host bus adapters, SAN switches, and tape drives, according to Ken Barth, president and CEO at Tek-Tools.

SysDM's flagship product, WysDM for Backup, supports backup software from EMC/Legato, IBM/Tivoli, and Veritas. WysDM provides application-centric performance management for distributed computing. The software performs backup monitoring and analysis and helps pinpoint chronic problems such as "soft errors" on tape drives and server performance problems, as well as monitoring network components.

"Our software goes beyond reporting to analysis," says Allan Atkinson, SysDM's CEO. "It looks at the kernel of the operating system to get details on firmware levels and, for example, shows when you didn't meet the SLAs on your backup window because the throughput of your backup software was not working well."

Atkinson says that backup applications often get blamed even when they are not the cause of a failed backup. "Backup software touches every piece of the enterprise infrastructure, so it's a very complicated problem," he says.

Adding to the problem, backup administrators only control the backup servers, tape libraries, and backup software. They often have no control over the clients or the network. More important, data centers have expanded to include hundreds of servers that need to be backed up on a regular basis while the demand to keep applications running around the clock has increased, which has led to shorter backup windows.

"All the backup applications work," says Liam Scanlan, chief product designer and architect at Bocada, "but the complexity of backup environments has grown."

All of the major suppliers of backup applications have backup reporting tools, but analysts say these products are either not comprehensive enough or don't support other vendors' backup applications. (Computer Associates is an exception because its reporting tools support backup applications from vendors such as EMC/Legato, IBM/Tivoli, and Veritas.)

Even as backups have become more critical, storage administrators have poor visibility and control. In the past, many companies tried to resolve this issue by developing in-house, proprietary software to monitor and analyze tape backups.

According to Taneja, a solid backup monitoring/reporting tool covers three areas. It monitors the entire backup environment, collects and analyzes data so administrators can address backup problems, and produces reports for administrators.

"It's this level of sophistication that's been missing," explains Taneja. "All of the leading backup application vendors have reporting tools for their software, but the tools often aren't very intelligent and/or they don't support other vendors' backup software. And they all deliver information in a different way; there's no common reporting format."

Unlike some backup reporting/monitoring tools, Bocada's BackupReport software is "agent-less" so backup administrators do not have to install agents on each backup server. BackupReport "talks" to a master server, which functions as a "translator" for a variety of backup applications, including CA's ARCserve, EMC/Legato's NetWorker, IBM's Tivoli Storage Manager, and Veritas' Backup Exec. The product includes three interrelated modules that deliver customized reports on how the backup system is working.

BackupReport focuses on three key disciplines: extracting data from multiple sources, turning the data into a common format, and building a visual representation of the data for administrators. Other functions include error reporting, reporting on backup success/failure and the percent of time that tape devices are in operation, and producing chargeback reports.

Virgin Atlantic Airways, based in the UK, has been using Bocada's BackupReport since November 2003. When the software was first installed, Virgin Atlantic determined that only 65% of its backups succeeded. Today, that figure has improved to 85% to 90%.

One of the first problems the software helped with related to why a particular batch of servers continuously failed during the backup process. The culprit was a specific switch connected to the servers. "We never made that connection," says Tim Graham, team leader of Virgin Atlantic's Data Systems Management group. "We just saw all the server backups failing. With BackupReport we discovered they were all on the same network segment."

Virgin Atlantic has almost 200 servers across its main offices in Crawley, West Sussex, and 12 regional offices worldwide. The company uses both Computer Associates' ARCserve and Veritas' NetBackup. According to Graham, Bocada's software gives Virgin Atlantic better visibility of the overall environment and enables the company to make better strategic decisions on its infrastructure.

This article was originally published on September 01, 2004