Remember having to package log files for IT to troubleshoot apps? Everything has changed. We now need complete visibility from anywhere. In the past, IT support received a trouble ticket from an unhappy end user and began troubleshooting the issue by asking the now, grumpier end user, to package their log files and send to the support center along with more details about versions, OS’s, and their last steps before things went bad. If things were even slightly complicated, the end user had to upload files to an FTP site.
Realistically, this exchange of log information worked pretty well for years—and still does in some environments, especially when the app is installed locally in the end user’s environment. If the end user is incompetent at using the log packager, the support staff just does a quick WebEx, Lync, or some other share-my-screen tool, into the client environment to solve the problem. With an app expert on one end and a relatively simple, client-server app with actionable logging data on the other, app performance is addressed quite nicely.
Things have changed. A lot. Now hundreds and thousands of users and businesses rely on much more complicated apps that get accessed from around the globe from multiple devices. These mission critical apps that we now use have a bunch of tiers including app servers, web servers, databases, and service calls to cloud components, and even to gadgets or probes that we’ve named the Internet of Things (IoT). Can you imagine having an end user round up the logs from this type of environment?
Now more than ever, end users are accessing apps from a more modern HTML5-enabled web interface that looks relatively simple on the front end, yet extremely complex on the back end. The end user might even be more competent, but this type of troubleshooting requires a better solution. As far as driving business, these apps are constantly changing and adding features as new business ideas proliferate our landscape. Uber, Airbnb, and Trip Advisor—just to name a few—have changed the way many of us now travel. The app is the center of their business.
So now you’ve got IT operations staff with a bigger burden, having an extraordinarily more complex task of rounding up the logs from machine data all over the data center and beyond. To add to the challenge, they’ve got to index it all together into something actionable to troubleshoot.
Without a viable application performance monitoring (APM) tool in place, enterprise IT staff has trouble finding which tier of the app failed. All they know is that something is not working and the end user is at risk for abandoning the activity—a sale is not made, a delivery is not scheduled, or a commission check is not sent.
If you were in charge of this IT environment, what would you do? Best practice IT operations analytics (ITOA) strategies include log analysis as a key component. For example, let’s say you have ten or more web servers running the front end of your apps, and an end user reports that they’re getting page unavailable errors just when they’re ready to complete an action. It would be very time consuming to go through each server’s log to figure out which one was causing the issue.
Using the right tool, you could find this issue in seconds and then correlate it to a root cause. IT Data Analytics tools should include:
- The ability to quickly search on any log data in real time.
- Presenting that data in an actionable dashboard that helps to pinpoint an issue.
- The option to pivot to a broader understanding of the environment.
Gartner Analyst and VP of Research, Will Capelli believes that within the IT Operation Analytics (ITOA) landscape, platforms with combinations of analysis engines and pattern discovery technologies will become the next version of the monitoring manager and the single pane-of-glass for IT operations management in general.
These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.