open
  1 (866) 866-2320 Resources Events Blog

3 Ways to Use Evolven to Stop Performance Issues

Blog

3 Ways to Use Evolven to Stop Performance Issues

About

This content is brought to you by Evolven. Evolven Change Analytics is a unique AIOps solution that tracks and analyzes all actual changes carried out in the enterprise cloud environment. Evolven helps leading enterprises cut the number of incidents, slash troubleshoot time, and eliminate unauthorized changes. Learn more

 

Large-scale IT outages from major organizations are making headlines every week.

In today’s world, any disruption can bring huge parts of life to a halt for both the organization and its customers. Yet even the most sophisticated technology companies have experienced major outages recently.

Facebook suffered its biggest IT outage since 2008 after a misconfiguration. An expired certificate in Let’s Encrypt's IT environment caused outages for Amazon, Google, Cisco, and others. An IT outage at United Airlines forced them to ground hundreds of flights across the US and Canada.

image

 

It is a well-known fact in IT that change is causing most of these IT outages.

This can include changes to the application, configuration, infrastructure, data, workload, capacity, or other aspects of a business system or IT environment supporting it.

That’s why IT teams are starting to recognize the importance of change awareness, and are using Evolven’s Change Control Platform to achieve this.

Evolven provides intelligent visibility into actual granular changes and differences across IT environments by automatically detecting and analyzing most granular updates across end-to-end hybrid cloud environments, whether executed manually or automatically, planned or unplanned.

With Evolven, DevOps and IT operations teams have reduced the number of incidents and problems occurring across their environments and reduced mean-time-to-resolution (MTTR).

Here’s 3 Use Cases for how IT is using Evolven to Stop Performance Issues:

1. Root Cause Analysis

Evolven accelerates MTTR by zooming in on the actual changes and differences (planned or unplanned) that it has detected before an issue happened.

Using machine-learning based analytics, Evolven automatically narrows down a list of these changes into a small subset of most probable root causes, even if the overall number of changes detected in the environment before the issue is vast.

Thus, when an issue occurs, IT can use Evolven to quickly view all granular changes that happened before the incident, prioritized by their probability to cause the issue.

The range of the analyzed changes can be extended by IT specialists to even look for less relevant changes that still could influence target environments (e.g., show me all suspicious changes that occurred in a specific environment for the last month), ensuring that no stone remains unturned.

The correlated changes can be automatically added to an incident ticket in a service desk or to an alert or event detected by a monitoring tool integrated with Evolven.

Real-world examples:
  • Performance of the system significantly decreased on Monday after the weekend maintenance. Evolven highlighted a routine change executed by DBA that dropped a database index and forgot to rebuild it.
  • An application update worked in the test environment but failed after production deployment. A comparison of environments with Evolven found that one of the application subfolders was set with the wrong permissions.
  • Change executed in a microservice did not align with container configuration causing intermittent failures downstream. Evolven connected the dots informing DevOps engineers about the misalignment. 

image

 
2. Early Detection of High-Risk Changes

IT teams are also using patented Evolven’s AI-based risk analysis engine to avoid incidents by detecting high-risk changes early.

Typically, changes executed in an environment do not cause an issue immediately.

There is a time gap between the deployment and manifestation of the issue that ranges from minutes to weeks. The more significant the gap, the more difficult it is to connect a change as a root cause to the issue.

Evolven assesses different risk markers of each granular change it detects using a mix of machine learning, statistic and heuristics algorithms, and combining the results through a complex decision tree.

Evolven continuously adjusts this risk based on the new data collected, time passed, observed behavior of a business system, and underlying environment where the change was detected.

Real-world examples:
  • A middleware specialist updated the application server's configuration, decreasing connection timeout to improve the system's performance. Evolven benchmarks the value of the changed parameter across the environments, identifying that this value is abnormal. It alerts the specialists that the change is risky and can potentially lead to stability issues.
image
 
3. Detection of Unauthorized Changes

Knowing and acting on unauthorized changes is critical to mitigating compliance, security, and stability risks.

A change management process enabled by a change management platform expects all changes to pass through registration, review, and approval steps. Changes that bypass this process introduce operational risk, as they were not assessed during the planning phase and not coordinated with the schedule of other changes.

If teams responsible for their IT environments’ performance and availability are not aware of unauthorized changes, incidents and problem resolution are lengthier and more complicated.

Evolven alerts IT of unauthorized changes by automatically correlating actual changes executed in an IT environment with approved change records in the service desk, or approved automated deployments in the automation platform.

Any actual change that cannot be correlated to existing records is flagged and reported as unauthorized.

Real-world examples:
  • A DevOps engineer resolving a critical issue in production made a change directly on one of the hosts and forgot to report the change later. Evolven reported the change as unauthorized, triggering a review. As a result, the change was documented and productized.
  • An administrator patching servers distributed a patch to an environment that was not in the approved scope and could interrupt a critical business application. Evolven detected out of scope changes and timely alerted the operations team to stop the patch.
  • A team implementing a change request did not finish deployment in time and decided to continue after the approved time slot. Evolven detected this activity and notified the change manager, who was able to coordinate the effort without any service interruptions.
image

Contact Evolven here to see the Evolven Change Control technology in action.

About the Author
Nicole Prybula

Field Marketing Manager