Ethan Banks, writing in Packet Pushers, examined the addition of user experience tracking as a new capability within...
Uila Inc.'s full-stack monitoring platform.
The new user experience tracking system is intended to find the root cause of infrastructure problems and identify issues with a high degree of certainty, Banks said. For example, Banks said Uila would track a web application by monitoring authentication, HTTP object loading, storage or database reads and host name resolution. Through its user experience tracking feature, Uila aims to help IT address user complaints about a slow network or slow response times.
In Banks' view, Uila's product is especially useful for any company provisioning applications on its infrastructure. And, through its reliance on hard data, it also helps limit the risk of different departments blaming each other when problems erupt. While Banks said he has not had a chance to demo Uila yet, he is particularly interested in the product's root cause analysis capabilities.
"Uila is historical. You can rewind -- go back in time to see what was going on in your infrastructure several hours ago when there was a problem. That helps deal with the 'incredible disappearing issue' all troubleshooters love to hate, because there never seems to be a problem when we're looking," he added.
Explore more of Banks' thoughts on Uila.
SD-WAN becomes increasingly mainstream
Andrew Lerner, an analyst with Gartner, said interest in SD-WAN continues to grow among Gartner customers. Avoiding the hype that's plagued other technological developments, SD-WAN, Lerner said, is actually delivering on its promise. Gartner estimated that SD-WAN has attracted more than 6,000 paying customers encompassing more than 4,000 implementations.
According to Lerner, the benefits of SD-WAN are manifold, providing users with a good return on investment and competitive cost, along with excellent performance and high availability. That said, Lerner said vendors still have to work to plug functionality gaps and also address the issue of heterogeneous connectivity.
Additionally, Lerner said, SD-WAN often involves more hardware than its name suggests. And while SD-WAN systems can often be easier to implement, some place a burden of complexity on branch offices.
Dig deeper into Lerner's thoughts on SD-WAN.
Tackling complexity in leaf-spine fabrics
Ivan Pepelnjak, writing in ipSpace, shared his thoughts based on a conversation with a networking professional who claimed it would be easier to replace elements of a large fabric with large chassis switches.
According to Pepelnjak, professionals need to distinguish between implicit and explicit complexity. In the example of leaf-spine fabrics, explicit complexity stems from dozens of different switches interconnected by fiber cables. While fewer explicit connections thanks to larger chassis switches sounds like a great idea, Pepelnjak reminded his readers that "it depends."
Chassis switches pose their own set of challenges. Automating the data center fabric does not necessarily make management easier. Large chassis switches are valuable, and engineers never want them to fail -- upgrading a single component of a large fabric can be accomplished easily with quick failure detection instead.
Additionally, restarting a 1RU switch with a single application-specific integrated circuit is faster than attempting to restart a complex, distributed system overseen by two supervisors and dozens of line cards. Pepelnjak shared a few other thoughts from engineers about large data center fabrics. He recommended making the failure domain as small as possible, focusing on the scalability of the control plane and being sure to always remember the ways in which redundant architectures can fail.
Read more of Pepelnjak's insights on leaf-spine fabrics.
Monitoring the experience of end users is vital for SaaS providers
Will SD-WAN replace edge routers?
Data centers shift to leaf-spine fabrics