Engine Works
Under the hood of Alteryx: tips, tricks and how-tos.
by Megan Dibble and Amelia Bong (@amelia_alteryx)
Introduction
Did you know that, on average, unexpected application downtime costs organizations $5,600 per minute (according to Gartner)? If your company has applications that are critical to business operations, a server failure could be extremely costly.
So how can you avoid a failure like this? By checking your server health regularly.
Setting up server health checks helps your business proactively avoid downtime costs and slowdowns. Reviewing data about your server’s performance can inform scaling and capacity planning activities. If your server isn’t performing well, then you wouldn’t want to load it with more and more business-critical operations. Reviewing this data can also help you spot unusual activity and avoid or at least manage risk.
It's not feasible to sit and watch server performance 24/7—that would be a tedious and expensive IT endeavor. And it can be time-consuming to analyze data on server health. This begs the question: how can you continuously monitor server health?
Auto Insights can help! Keep reading for an example of how Alteryx Auto Insights (AAI) is a great tool for server health checks.
Alteryx Server Use Case
At Alteryx, we can help perform Alteryx Server health checks for some customers. Taking the data from the customer’s server, we put it into AAI to analyze server capacity.
Once data is loaded into Auto Insights, we have to select our metrics. We selected server queue time as a metric for server capacity; a long queue time would mean the server was overloaded.
When we saw increased queue times, we would break down the metric to identify the time of day, workers, and execution type that contributed the most to these values. Here is an example of the mission summary view that Auto Insights provided:
It automatically shows us the most significant contributing factors to the increase, allowing us to make informed recommendations to customers.
Looking at the detailed stories, we could quickly see that a large proportion of the total minutes in the queue occurred in the morning. This allows us to let customers know how to level load workflow jobs on the server throughout the day to optimize server performance.
Under the “How was the total distributed?” section, we see the workflows that waited for the longest in the Alteryx Server queue.
All of the information Auto Insights provides allows us to make data-driven recommendations to help customers scale their operations, manage their schedules, and optimize their server performance. And since server performance should be monitored in an ongoing manner, Auto Insights is the perfect tool because with each data refresh, it will pull out the new key changes and insights automatically!
