1. Don’t Know The Possible Impact of Deployments
Fast, short release cycles make it hard to keep track of your application and infrastructure landscape. A small change deep down in the infrastructure may seem innocuous, but can have a large impact over time. Today’s monitoring solutions should be able to provide insight which systems could be impacted before you deploy new software.
Ask yourself: do I know the possible business impact when I execute this deployment?
2. You Need to Jump from Tool to Tool to Root-Cause Issues
When issues occur, you have to manually correlate siloed data sets across multiple tools. This process requires too much time and resources and is not suited for today’s dynamic and elastic environments. This is definitely a sign that you should start looking for a better monitoring solution.
3. You Rely on The IT Superman a.k.a. Brent
Remember Brent? Brent is a character in The Phoenix Project (a novel about IT and DevOps), who seems to be critical for every project and is badly overcommitted. Brent, or as we call it ‘The IT Superman’, knows every system, every line of code and every dependency across the IT landscape. When issues occur, Brent needs to be involved to solve it. He represents a human bottleneck for the IT organization.
To accelerate and scale IT operations it’s important that everyone in the IT organization has knowledge about the systems they maintain and its dependencies across other teams and systems.
4. When You Receive a Dozen Notifications From Multiple Tools For a Single Problem
According to a survey by Enterprise Management Associates (EMA), over 65% of the enterprise organizations have more than 10 monitoring tools. Buy a tool and become more proactive they thought. But as more monitoring tools were deployed, more events and alerts were generated for the same problem.
If an alert is triggered somewhere in your IT landscape, chances are many more alerts about components related to the original one will go off as well. The resulting "alert storm" can overwhelm IT operators, obfuscating the true cause of the problem and crippling the team's ability to respond.
All of those tools were never meant to work together, it's like they speak a different language.
5. War Room Meetings Are Essential For You To Diagnose Issues
A war room has become an essential approach for large enterprise IT organizations to diagnose complex issues. The approach of the war room is to put together a group of people from different IT siloes, put them in one room and set them to finding a solution to a critical issue.
But war rooms do more harm than good. They make teams less productive, they are reactive instead of proactive and it’s hard to get everyone on the same page as each person has their own monitoring solutions and thus speaks a different ‘language’. This process is painful, time-consuming and often fruitless.
6. There’s a Gap Between Application Performance and Infrastructure Management
Today’s Application Performance Monitoring (APM) solutions focus on identifying and diagnosing application-performance issues caused by the application code, such as Java and PHP. Issues caused by infrastructure teams are not visible to APM solution because these tools primarily look at the application code for custom application servers.
Infrastructure teams have the same visibility-problem. They lack application visibility and don’t have a clue of how infrastructure configurations could affect application performance.
Monitoring is at the core of any digital enterprise. Today's monitoring solution can help but, it's not always about the tools. It's also about people, culture and collaboration across teams.
What other signs or mistakes do you see other teams and IT organizations making?