“We see the greatest impact of archiving with Structured Data Manager (SDM) on databases of over 100GB,” says Peter Rospond, Operations Manager for HPE IT Center of Excellence (CoE). “With hundreds – or thousands – of users across the globe submitting online queries simultaneously, performance counts. Records need to be retrieved with lightning speed. Having millions of unnecessary, aging records lying around in the database impacts query times and – when users are kept waiting – productivity.”
With rapidly growing data volumes, HPE IT (previously HP IT) needed to find a way to optimize their data management lifecycle to meet their Service Level Agreements (SLAs). They also needed to ensure data compliance along with fast, easy retrieval when required. The deployment of Database-to-Database (D2D) archiving using SDM archived historical data from the production databases on high-performance, more expensive Tier 0 storage, to Tier 1 and Tier 2 storage. It also ensured seamless access to archived data for easy search and retrieval.
“When SDM was originally implemented (before HP split into Hewlett Packard Enterprise and HP Inc.) the initial impact was huge,” reflects Rospond. “Three production instances in Atlanta, Austin, and Houston supported over 70 business-critical applications across 180 source databases and environments, running a total of 1800 D2D archive jobs. We archived over 100 billion rows – equating to more than 83TB of compressed archive data with a typical compression ratio of 3:1 – from the production databases over a five-year period, reducing query times by up to 89%. We also reduced our backup/restore window at that time by up to 48%, and the duration of full database backups to tape by 37%.”
A reduction in storage capacity was one of the key objectives of deploying SDM, with savings multiplied across database clones used for high availability, disaster recovery, and standby instances. The SDM implementation resulted in Tier 0 storage capacity being reduced by up to 48%1.
“During the company split we took the opportunity to further optimize the SDM environment,” continues Rospond. “We created a completely new SDM environment for the HPE applications and migrated the existing archives across to the new system. The whole process was quick and seamless.”
“Now we’ve reduced the number of archive jobs for the HPE environment to 140 covering 51 applications – 24 of which are mission-critical – across 123 source databases/ environments,” states Rospond. “Since the split in November 2015, we’ve already archived over 1.3 billion rows, reducing our Tier 0 storage requirements by an additional 15%. This has contributed to reduced resource contention on our database servers, increasing query performance by an additional 43%.”
Rospond sums it up: “Data volumes are going to continue to grow. But SDM enables us to manage that data efficiently and cost-effectively. By intelligently archiving millions of records using SDM, we’ve significantly increased query performance across all of HPE’s applications, driving increased productivity while realizing significant dollar savings.”