CASE STUDIES: Archiving tools relieve database bloat

By Kevin Komiega

—Data growth has joined death and taxes as one of the only guarantees in life and, as a result, database-specific archiving tools are growing in popularity. Sometimes referred to as "active archiving," these tools are only offered by a small number of vendors, but can have a big impact on database performance and backup windows.

The premise is simple: End users can realize better database performance and storage cost savings by taking the concept of information lifecycle management (ILM) and applying it to business transactions and application data.

As senior manager of business applications systems for Tektronix, a global developer of test, measurement, and monitoring equipment, Lois Hughes needed a leaner, meaner production environment to effectively support branch locations in 27 countries.

"Our database is growing by a substantial amount on a daily basis," says Hughes. "We cannot keep 10 or 15 years of data in our production environment, but because we have to meet the legal requirements of 27 different countries, I have to maintain data and make sure it's accessible for a certain period of time to meet local statutory requirements."

The growth of Tektronix' database was slowing down the replication process and gumming up the works for Hughes. Her team had decided to write its own database archiving tool until they came across a product from OuterBay Technologies called LiveArchive, a part of OuterBay's Application Data Management Suite (ADM) that relocates closed transactions and other infrequently accessed data to an online archive database. The process reduces the size of Hughes' Oracle 11i production system, thereby boosting application performance.

Hughes says she can now easily maintain two years' worth of financial, ERP, and human resources data in her production database and offload data to the archive once it turns three years old.

Tools like OuterBay's identify inactive or rarely accessed data based on retention policies and activity histories. Once data has been tagged for archiving the software automatically moves it to less-expensive tiers of storage while maintaining its referential integrity for easy retrieval.

Robert Skaljin, IT manager for TD Bank Financial Group, was also battling the performance and management issues of a bloated database until he began rolling out an archiving tool in his environment.

Skaljin was faced with an ever-growing PeopleSoft Human Capital Management (HCM) 8.8 database and a large pool of DASD storage devices. "All of our database activities were taking longer and longer to complete. It impacted a number of things, including backups and transaction performance," he says.

Skaljin recently began a rollout of Princeton Softech's Optim product to guard against performance lags and possible system failures. Like OuterBay's LiveArchive, Optim determines where application data is growing fastest and applies business rules to govern active, inactive, and reference data. Simply put, Optim decides when data needs to be available, where it should be stored, how long it should be stored, and who can access the files.

The ultimate goal for Skaljin is to maintain his production database at a manageable size to simplify maintenance and increase performance.

Evaluating vendors
OuterBay Technologies and Princeton Softech are not the only vendors specializing in active archiving software for databases. BridgeHead Technologies and Grid-Tools Ltd. entered the fray last December by announcing a deal that will combine BridgeHead's FileStore storage management technology with Grid-Tools' GT Datastore archiving software to create a database archiving tool.

Some issues to consider when evaluating database archiving tools are the level of integration (or lack thereof) with existing database software, the vendor's commitment to staying on top of the technology road map of a given database vendor and, of course, cost.

Key factors for Tektronix's Hughes were OuterBay's ability to stay with Oracle's patch level and its customization capabilities. Hughes also points to OuterBay's plans to archive data in XML format as another important piece of the puzzle. "We want an XML format so that we can get rid of the 40 or so instances of files from different software releases," says Hughes. "We don't want to be application-dependent when databases get too old."

TD Bank's Skaljin used similar criteria when evaluating Princeton Softech's software. "Integration with PeopleSoft was huge for us," he explains. "We looked at how quickly we could deploy an archiving solution and Softech had templates built and certified by PeopleSoft. That cut down on development time."

Skaljin also made sure Princeton Softech could meet his business retention and viewing requirements for a reasonable price.

Archiving vs. backup
Despite the obvious benefits of simplified database management and storage savings, database archiving is just beginning to garner attention in the IT industry. Some analysts believe the slow adoption of the technology could be due to confusion as to how the tools are different from traditional backup software.

"Database archiving is gaining in popularity, but I think there is confusion in the terminology that has not helped to differentiate active archiving from simple data backup," says Dianne McAdam, a senior analyst and partner with the Data Mobility Group research and consulting firm.

The tendency for IT users to confuse backup with archiving is not rare and it's not limited to database archiving. In fact, a recent survey commissioned by BridgeHead Software showed that while 77% of the users interviewed claim to archive data in some way, 35% of those respondents said their archiving process consisted of manual archiving using backup software. BridgeHead surveyed 165 IT managers from organizations across several industries.

McAdam believes that the number of end users who mistakenly use the terms backup and archiving interchangeably is disconcerting. "It shows there is a danger that many companies have neither the data protection nor the archiving systems in place that they believe they have," she says.

There is a simple way to differentiate between archiving and backup. According to McAdam, all of these technologies fall into three categories: active archiving software slims down databases by relocating inactive data; backup is the storage of data for the purposes of restoring lost or corrupted files; and deep archiving is the long-term storage of files to meet regulatory compliance, legal, or best practice requirements.

This article was originally published on January 18, 2006