Currently, there is an issue with the calculation of query duration metric (searchQueryExecutionMaxTime) in the infinispan mbean for statistics. The metric being generated do not accurately reflect the actual query duration, leading to inaccurate performance analysis and misleading insights.

The problem lies in the formula used to calculate query duration metrics
In the BaseEmbeddedQuery class, specifically in the execute() method on line 82, the recordQuery function is called with only the startTime argument. However, it should actually be System.nanoTime() - startTime to align with the implementation in the IndexedQueryImpl class.