Agencies can modernize as they optimize.
IT optimization has emerged as a means to modernization in the federal government. Determined to “work smarter not harder,” 25 percent of tech leaders rate their IT environment as “fully” optimized, up from 9 percent who reported the same progress in 2016, according to a survey from IDG Research Services. But optimization isn’t just about getting the most modern tech on your networks; it’s also about making the tech you already have work as fast and efficiently as possible.
Federal initiatives such as Cloud First, FedRAMP, Shared Services and PortfolioStat encourage a “do more with less” mindset, and according to the Office of Management and Budget’s Data Center Optimization Initiative, 24 agencies are currently virtualizing systems to improve performance and boost energy efficiencies.
The vast majority of attention focuses on “big picture” projects: tackling a major component of the IT infrastructure/ecosystem to reap the rewards of massive cost savings and productivity increases. However, there is considerably less discussion about the smaller, day-to-day user interactions which create frustrating bottlenecks and lengthy downtime that’s simply tolerated as an unavoidable reality.
It’s important organizations think about optimizing networks beyond their major IT projects, including in ways that enhance the end-user experience such as faster application load times, shorter processing times and more efficient networks overall. By implementing three low-cost optimization techniques, federal organizations have been able to reduce load times for Internet Explorer from 5.9 seconds to 2, and drop Microsoft Outlook load time from 75 seconds to 4. In other cases, federal organizations have been able to cut the standard Windows log-on from nearly 12 minutes to less than 4 minutes. In other instances, as much as 230 TB of data were removed off the wide-area network in a single week on account of optimization efforts, reducing the overall queue and improving performance for the end users. Here’s a look at three common ways other federal agencies can optimize to reduce end-user wait times, frustration and inefficiency.
Network Fragmentation: One of the more esoteric system functions is in the way that data travels from point to point in packets. If these packets are too big, users can experience network fragmentation, which refers to the process of breaking inbound data packets into smaller pieces. Think of a network device as a window that can only accommodate so much data within one frame. If you’re trying to jam in the equivalent of a jigsaw puzzle the size of a dining room table through a comparatively tiny window, it won’t work. So you break the jigsaw into pieces to push everything through, and then reassemble the pieces on the other side. But this takes time and often leads to data transmission delays, hindering user productivity and performance.
The best way to optimize network fragmentation is to avoid it. Every network has an overall Maximum Transmission Unit that dictates the frame size by limiting the volume of data processed in a frame. IT solutions today enable you to analyze your maximum frame size, and then set your MTU to accommodate all server traffic that exists below this size. In the process, there is less breaking up of data packets to make them fit and reduced latency overall. Aside from the time needed to determine and set the MTU, optimizing network fragmentation doesn’t require any special tools or software, making it a very low-cost effort.
Server Message Block Protocol: SMB protocol is the protocol used to retrieve and move data. When you copy and paste a file or move a file from one place to another, you are using SMB. That said, an ever-proliferating abundance of this traffic—along with increasing security requirements—is slowing down the process to the point where you sometimes feel like you can drink a cup of coffee while you wait to call up a PowerPoint sent by a co-worker.
What’s needed are available solutions which help SMB “cut to the chase.” Traditionally, SMB involves a lot of “questions asked” between nodes and the distant end, each question arriving with addresses and headers that add bulk to the traffic. A node will ask “Can I …?” And the distant end will respond with, “Can I ... what?” Then the node will ask “Can I transfer …?” And the distant end will respond with “Go ahead.” And that conversation is repeated. Obviously, this is a rather stilted and inefficient conversation. With the right solution in place, multiple exchanges can be reduced to one, so the node simply has to ask “Can I transfer this PowerPoint file to this machine?” and the distance end sends it all over without all the chatter.
Hypertext Transfer Protocol and Hypertext Transfer Protocol Secure: HTTP and HTTPS define how data is formatted, secured, and transmitted on the web. Typically, web pages aren’t a single file but consist of multiple, separate files which servers must request and retrieve separately, frequently leading to delays in accessing information. Much like SMB, there is chatter which uses significant time and causes users to wait due to all the extra headers.
To address this, you should invest in wide area optimization appliance that recognizes what changed within every transmission. This would work like “tracked changes” in a Word file. When you make edits for, say, a 4,500-word doc and send to someone for review, you shouldn’t ask them to go over the entire, lengthy doc. That’s why you apply tracked changes, so your colleague jumps right on those and ignores the rest.
When we think of IT optimization, we often envision extensive efforts that will require months of planning, execution and testing as part of an organization’s modernization strategy. But we can’t overlook the smaller, easier adjustments which will make a huge difference in improving the technical performance and the user experience. Time is money, after all, and the minutes saved per transaction—when multiplied over thousands of federal users pursuing daily tasks over a year or longer—will amount to a large chunk of return-on-investment-benefiting savings. That’s the kind of optimization that any agency can—and should—take on.
Terry Bailey is a senior network architect for NetCentrics Corp.