BalancePoint has already saved us significant time by identifying the root cause of server and storage problems and enabling faster, more informed decisions
Merkle is one of the nation’s largest and fastest growing customer relationship marketing agencies. For more than 20 years, Fortune 1000 companies and leading nonprofit organizations have partnered with Merkle to maximize the value of their customer portfolios. By combining a complete range of marketing, technical, analytical and creative disciplines, Merkle works with clients to design, execute and evaluate Integrated Customer Marketing™ programs.
Merkle’s IT department manages more than 300 servers hosting very large data warehouse solutions running in the range of 5-10+ TBs each with approximately half-a-petabyte (500TB) under management. The complexity of the infrastructure coupled with complex workloads generated by Very Large Databases (VLDBs) posed various challenges for the Database and IT teams. They really wanted a solution that provided end to end visibility from its applications down through to the storage that would allow them to troubleshoot, optimize and plan performance and capacity of the entire infrastructure.
Troubleshooting Troubles
After spending countless hours and days troubleshooting, the DBA and IT teams were tired manually correlating data to resolve system-wide issues and understand where the contention points were at various resource levels. Most of Merkle’s VLDB solutions are I/O bound and underlying storage is shared by multiple client implementations. Merkle’s HP Enterprise Virtual Arrays (EVA) handle so much workload that sometimes they cannot handle all of the work, creating bottlenecks.
BalancePoint Visibility and Analysis
Akorri BalancePoint provided visibility into the storage hotspots and identified the root cause of the problems. With the information that BalancePoint provided, Merkle’s DBA & IT departments were able to slice and dice certain servers and move them from one EVA to another. BalancePoint helped them utilize their EVAs more effectively. The hotspot analysis, application contention and data throughput analysis features were very useful to layout an optimized solution for Merkle’s clients. BalancePoint also maps database instances and schema elements, showing how they are deployed on server and storage systems and making it easy to troubleshoot and optimize database performance.
“BalancePoint has already saved us significant time by identifying the root cause of server and storage problems and enabling faster, more informed decisions,” said Atul Gandhi, Sr. Director, Enterprise Technology Group, Merkle Inc. “BalancePoint provides deeper statistics for the HP EVA platform and Cisco SAN switch support, which will likely give us an even deeper understanding to quickly resolve bottlenecks, contention or latency issues, and enable better performance from our data center.”
“BalancePoint manages across all of the different resources in the datacenter automatically correlating server and storage metrics and providing unique analysis. It sits in the background collecting all of the information and takes us to the next level to make our recommendations on how to solve problems. That’s where BalancePoint has helped us out big time,” continues Gandhi.
Results
• Reduced time and resources spent on troubleshooting and diagnostics
• Ensured optimal DBA & IT performance via daily performance intake
• Improved IT budget planning
•Helped lay out different data warehouses without introducing contention problems
