ExakTime (formerly TimeSummit) Connect, our cloud-based time tracking software, frees you from buying software as well as operating or managing your own hardware resources.
That’s a major plus.
SaaS software is great because affords a lighter load for customers, but it does introduce a whole new aspect of management on the back end.
ExakTime’s software development (a.k.a. Special Ops) team no longer oversees individual servers (oh, how quaint) where all our customers’ data was once briefly stored while it traveled from the field to their computers.
Instead, we now manage a series of applications which dynamically add and remove Microsoft Azure servers based on customer workload. With over 20 individual components able to scale independently, we could be utilizing anywhere from 20 to 250 servers on any given day.
How We Monitor
With a computing environment that’s constantly changing, we use an array of tools to monitor operations.
We have a dashboard that helps us quickly spot usage spikes in our cloud software throughout our many-thousand-count customer base, or areas of heavy usage. Key metrics like ExakTime Connect response times, CPU usage, database performance, and memory consumption send out automatic alerts when any of our configured thresholds are exceeded.
A series of tests simulating user activity on internal sites also helps us monitor the environment. Every few minutes, our friendly virtual worker bot logs into a test site and creates a test punch in the Time Card Detail area. In the event of a failure or slow response time, the bot engages the alerting system to start the notification process of a potential issue.
The alerting system then works through an escalation process consisting of emails, text messages, mobile app notifications, and voice calls to ensure someone is looking at the issue, regardless of the hour of day or night.
In addition to the above, we have a variety of tools observing real-time activity in ExakTime Connect at all times. This allows our team to check where users are on the site, track trends for page times, and observe if problem areas are developing.
We can also see at a glance how many users are on the error page (known by our team as the “Houston” page) if a problem has occurred. We routinely monitor this page during deployments as the production site is upgraded in every effort to avoid impacting customer activity.
We have continued to and will keep on adding to our arsenal of monitoring capabilities to ensure a positive experience on ExakTime Connect.
Refining our system monitoring is key to making sure we detect issues and radio into Houston before the customer does, so we can resolve those issues as quickly as humanly possible (with a little help from technology, of course).
Brad Douglass is a Senior Software Engineer for ExakTime. He brings over 20 years of experience to building and operating software applications. With technology constantly advancing, Brad enjoys learning about new technology and using it to produce better software products.