Overview
Founded in 1905, Anchor Hocking is a leading marketer and manufacturer of a comprehensive line of glass products including beverageware, candle containers, servingware, ovenware, storageware, lighting components and other glass products sold under various brand names or as customized solutions for private label lines. Anchor Hocking is the second largest supplier of glassware in the United States. Its glassware products cross all price points through the retail, specialty (business-to-business), and hospitality channels.
Of the 1,500 people located in Anchor Hocking’s two primary production centers in Lancaster and Monaca PA, about 450 office workers depend upon their IT environment to support their use of the Microsoft suite, including Exchange 2010 as well as their Oracle ERP system to conduct their business.
Challenge
Anchor Hocking Manager of Infrastructure and Operations Michael Stewart reports that the company made a decision a few years back to insource their IT rather than outsource it.
At that time there were approximately 20 servers running internally, and another 20 or more located at a data center in Dallas, Tx. “We had servers down there that we needed to purchase up here. When we started looking at that infrastructure, what it would cost, and the overhead, it made sense to virtualize.”
Stewart proudly declares Anchor Hocking to be about 99% virtualized now. Today, instead of what would have been 40 to 50 physical servers, Stewart runs his VMware environment of 91 server sessions on only 5 virtualized servers. Two legacy applications require physical servers, bringing the physical count to only 7.
“The economies of scale just made it make sense that they come from physical to virtual and its been a pretty stable environment!” concludes Mike Stewart.
“VMware has their own reporting, they just don’t do a very good job of making that available right in your face. My biggest hurdle that I needed to look at was IOPS, if we’re overutilizing, who’s overutilizing. We started looking at products that were out there and VMTurbo gave me what I needed real time and was simple to use.”
IOPS (Input/Output Operations Per Second) is a common storage metric that helps managers understand how quickly a system can deliver stored data to users in applications, and which applications and users may be overutilizing specific resources.
Solution
“As far as the management is concerned, they want to see that everything is green,” explains Mike Stewart. But when it comes to his team, they want to see all the key metrics – CPU, memory utilization, IOPS, bandwidth utilization, and an ongoing day-to-day historical view of how everything is trending. “We have to make sure that we don’t have a server that is running crazy out there chewing up CPU for the past days, same with IOPS and memory.”
Stewart points out that while management doesn’t require any specific reporting from him, VMTurbo has allowed him to proactively keep them aware of the ongoing health and growth of their IT environment.
“Where we saw the benefit in our environment was in time management. I had a tech that was drilling down through VMware itself trying to find these reports and making sure we didn’t have devices running out of control. Across 91 servers it took him a good chunk of time to drill down through that. VMTurbo has that dashboard that, as soon as you log in, lets you know who’s having problems right away or if everything was okay. The time we saved at that we could devote to other tasks. We easily saved a day a week not hunting for this information as it was right there on the dashboard.”
Mike Stewart goes on to add, “The other issue is uptime. Finding and understanding the runaway VMs that were causing issues with memory. We had increased uptime which thereby increased performance of customer-facing and sales-facing equipment.”
Return on Investment within a Month
While he didn’t have specific data, from his own impressions, Mike Stewart feels very strongly about the value received from VMTurbo.
“VMware itself does every bit of this same reporting, you can find it, you’re just going to have to drill down and find it on each individual host, “explains Stewart. “The cost associated with VMTurbo was almost nominal. We had, easily within a month, return on investment in the sense of manpower not having to drill down through that kind of stuff. Everybody is out there writing tools to do this, but VMTurbo gave us an extremely simple interface to do a very difficult task.”
Benefits
- Key performance metrics from all servers are presented “right in your face” right out of the box
- Network managers save significant time by avoiding the need to drill down into ordinary management tools host by host
- Potential performance problems are identified and resolved early
- Aggregation of all servers into one dashboard saves time creating rapid return on investment