Metric Grouping Issues (MGI’s) occur when metrics that should be grouping, arent. You can read more about them in our public docs, but, in general, MGI’s result when the granularity of metric naming is too fine. One of the more common MGI”s we see occur when UUID’s are included in the metric name. This can cause a few code paths have thousands, hundreds of thousands or even millions of unique names.
We’ve added automatic detection for UUID explosions and added grouping based on standard 32 digit dash separated hex characters with more than 100 unique segments.
This automatic detection can take a list of Transactions looking like this
And automatically group them together to look like this, correctly identifying the product Transaction as the most time consuming.
This change is to help improve your APM experience by automatically detecting URI segments in Transaction names to correctly group your similar Transactions together, preserving trend data and correctly report the most time-consuming paths.