Advanced Computing in the Age of AI | Monday, May 20, 2024

Tecplot Improves Storage Performance With PernixData 

PernixData today announced that Tecplot has successfully leveraged PernixData FVP to improve overall storage performance and support a growing virtualized environment.

In early 2013, Tecplot invested in additional compute resources for their infrastructure to support virtualized build systems and faster code compiling. While this provided a significant benefit to the software development team, the performance bottleneck shifted to the storage infrastructure. This also impacted other business critical systems that ran off of the shared storage system.

Tecplot faced the increasingly common dilemma of quickly outgrowing storage performance capabilities as their virtualized environment grew. To avoid a complete overhaul of their storage infrastructure, Tecplot turned to server side flash technology and selected PernixData FVP for its ability to support true write-back caching. PernixData FVP is tightly coupled with the hypervisor to provide a highly available pool of cache that provides the safety needed when accelerating writes, without compromising feature-sets in vSphere.

"We believed server side flash could provide a more scalable and cost-effective way of addressing our storage performance challenges. But, we needed a software solution to make it truly usable across our virtual data center," said Koehler. "PernixData helped us avoid a substantial SAN upgrade, which would have cost tens of thousands of dollars. Based on that fact alone, I believe flash virtualization is absolutely the right way to go."

PernixData FVP was easily integrated with Tecplot's existing infrastructure, from servers and storage to hypervisors and VMs. More importantly, the performance improvements reduced storage workloads from 100 percent around the clock to a 65 percent storage workload. In addition, adding the PernixData FVP technology delivered a 2.3 times performance improvement in the company's compiling application, above and beyond the initial performance gains from added compute power.

EnterpriseAI