Situation
A production cluster under strain
The client processes payments in real time and relied on RabbitMQ as its messaging tier. In production the cluster showed performance and scaling problems.
Approach
Understand first, tune second
The first step is always to understand the overall infrastructure and the actual use cases RabbitMQ serves, not to start tuning blind. Working closely with the client's engineers, the team stabilized the cluster.
A follow-up architectural review then looked at how the platform should be built to stay reliable and responsive for the client's users.
Modernization
Link to modernization
Messaging problems are often a symptom of the system around the broker. Fixing them in small, checked steps is the same approach described in the 8 principles. For a documented project, read the American National case study.
See also Application modernization services.
Note
About this example
This page is an anonymized engagement example. It names no client and states no metrics. The documented case study on this site is American National.