Real-World Problems Solved By Performance Monitoring and Optimization

To gather insights on the state of performance optimization and monitoring today, we spoke to 12 executives from 11 companies that provide performance optimization and monitoring solutions for their clients.

Here's what they told us when we asked, "What real-world problems are you, or your clients, solving with performance optimization and monitoring?" 

• We have an ad services client monitoring service quality by monitoring latency and the flow of ads being served. The path through the internet varies. It can be influenced and changed. It’s important to be aware of the path being taken.
• Large-scale global automatictraffic measurement for static assets as well as video streaming to deliver business policy and intelligence are key. They should be driven by the end user perspective and actions should be mitigated in seconds. Clients want control, flexibility, scalability, and the peace of mind that comes with automation.
• Decrease time to resolution. When there’s an issue, collaborate quickly to resolve it. Time to resolution used to be measured in parts of an hour. Now it’s one to two minutes because we’re able to integrate with all other monitoring tools. We help people from different teams collaborate, once the issue is identified, via a shared space to get everyone on the same page quicker. Help close the loop with continuous learning. Identify the leading indicators of various problems.
• Understand how service providers are performing and affecting the delivery of your application from the data center down the chain to the customer. This includes the host environment, DNS services, and DDOS services. You need to appreciate and understand how each impacts performance. Think about how to plan and strategize to improve and optimize over time from various parts of the world and various platforms.
• Build a connected product. Thousands of devices in the field to stay connected is very difficult. Connection is a set of states and switches. A connection that is open hasn’t been specifically closed yet. You need to make reliable connections with greater connectivity than web networks. The further the server is away the more likely the packet is to get lost. Connected devices can’t hide problems. It can be difficult scaling problem and performance metrics.
• Make sure SLAs are being met for recommendation engines, fraud prevention, and other mission critical services that must be completed in a specific amount of time. An SLA management tool with auto-detect grouping similar jobs and looking for trend data on them. Able to alert on the SLA ahead of time to pinpoint why the speed is slow — more data, less resources, resolve, and meet the SLA time. DataOps teams (data teams and operations guys) have all apps running, full visibility into what’s running, data usage, resource usage, ability to see all the applications run. Give people full information at all levels for full-stack visibility at one place.
• The application load was increased tens of times over the course of an in-house beta testing round. Once it is released, it is estimated to raise hundreds or even thousands of times. To ensure smooth product start and enjoyable end-user experience, it is vitally important to check the application performance under the expected load conditions.
• Proactive problem avoidance is the primary reason IT organizations deploy our solution. They want to know about potential performance problems well in advance of applications end-users becoming aware. They want to know about latency issues, resource contention bottlenecks, and physical layer issues with their network infrastructure to take proactive and preventative actions. Cost optimization and ensuring that IT infrastructure investments are fully realized is also a problem our customers are solving, too. With our solution, they can drive the maximum performance and utilization for their IT infrastructure at the optimal cost, while at the same time minimizing risk. Because of this, they can accurately forecast and budget for future IT investments. IT teams can troubleshoot and resolve their toughest performance problems. The infallible accuracy of algorithmically-driven applied analytics gives teams an authoritative understanding of what’s happening, where it’s happening and why it’s happening. This eliminates the finger pointing and helps IT teams identify and remediate root causes exponentially faster, and with absolute certainty.
• On one project, we’re trying to ensure servers can sustain tens of thousands of concurrent users.. This is a real-time platform with stateful connected clients with database persistence. We could go higher but as I said previously, know your product and your load. Going higher could prove very costly in terms of development time. Monitoring would provide the answer if the results are worthwhile.
• When our software is slow, it means our customers can’t release their software into the world. So, we should make sure that we optimize how the binaries flow through our products, or else we might impact a customer releasing their software. This has a ripple effect that hurts software productivity. We believe you must release fast (or die!) so we can’t slow down that process or they will lose trust in us.
• For each product, it’s a little different. Our web applications are looking at CPU and memory usage, duration of queued request, user response times, page load times, and dozens of other performance indicators. These help us solve for customer experience and productivity, as every second they are waiting on information to be displayed is taking away from their ability to help their customers or colleagues. Our solution spends a lot of time reviewing speed of data delivery from end-user to the techs. They have automated tests that look at network quality, OS, and regional impact, just to name a few, on the quality of the remote control or meeting performance.

By the way, here's who we spoke to!

• Josh Gray, Chief Architect, Cedexis.
• Jeff Bishop, General Manager, ConnectWise Control.
• Bryan Jenks, CEO and Co-Founder, DropLit.io.
• Doru Paraschiv, Co-Founder, IRON Sheep TECH.
• Yoav Landman, Co-Founder and CTO, JFrog.
• Jim Frey, V.P. Strategic Alliances, Kentik.
• Eric Sigler, Head of DevOps, PagerDuty.
• Nick Kephart, Senior Director Product Marketing, ThousandEyes.
• Kunal Agarwal, CEO, Unravel Data.
• Len Rosenthal, CMO, Virtual Instruments.
• Alex Rysenko, Lead Software Engineer, and Eugene Abramchuk, Senior Performance Engineer, Waverley Software.