Data Storage Tiering Gets Logistics Company on the Road Again

Matt Sekelsky approaches every network systems challenge with the same mindset: no preconceived opinions.

matt-sekelsky-4

Matt adopted this approach after serving 10 years with Huber & Associates in a variety of roles, from pre-sales to management to development and now senior systems engineer. “There are a million different reasons people choose one IT solution over another,” Matt says. “When an issue arises, clients are looking for a supportive partner, not a know-it-all. I just want to help in whatever capacity the IT team needs, be that as an expert or simply as a second set of eyes.”

Road Rules

For a large trucking company, inadequate storage capacity was the problem Matt’s team was engaged to investigate. Overall transportation logistics were “out of whack” because the existing systems could not process time-sensitive data queries quickly enough to answer questions such as “what’s the fastest route from Point A to Point B?” Additionally, headquarters was literally running out of physical rack space to store the data. The right technology solution would allow the company to avoid a costly facility expansion while maintaining its fast, efficient and reliable competitive advantage.

Storage Tiering by IBM®

To address physical space concerns, the team selected IBM SAN Volume Controller (SVC). The SVC allowed the system to identify data duplications and then tier and compress data to deliver increased capacity in a significantly smaller footprint, reducing the requirements from 42 rack units to approximately 10 rack units. Engineers also replaced the slower, spinning-disk-driven Storage Area Network (SAN) with the extremely fast IBM FlashSystem 900® for the fastest tier of disk. The company would have had to add hundreds of disk drives to the older system for it to achieve similar performance. The SVC and the Flash drives combine to deliver a “smart” system that achieves outstanding performance and reliability by automatically flagging frequently accessed data and migrating queries between storage tiers based on real-time usage analysis patterns.

3 Things Matt Knows for Sure

  1. Always operate from facts. If you think your system is slow, resist the tendency to guess at the reason(s). Take the time to conduct an assessment to really understand the bottlenecks.
  2. Set performance parameters to address specific shortcomings, i.e., more input/output operations per second (IOPS), increased transactions, latency, etc. After setting performance parameters, model the correct solution.

Conduct a full cost/benefit analysis. Your final decision should take into account the entire return on investment, i.e., the value of speed as a competitive advantage and the cost savings of rack space improvements as they impact facility and energy expenses.

Get more information about Huber & Associates data storage solutions now.