Elasticsearch is a search and analytics engine and vector database built on Java and Apache Lucene, part of Elastic's Search AI Platform. It requires a Java Virtual Machine (JVM) to run. The JVM provides a platform-independent runtime environment, allowing Elasticsearch to function across various operating systems. Understanding JVM memory management and garbage collection is critical for troubleshooting memory-related issues. These issues can manifest as java.lang.OutOfMemoryError or Exit code 137 due to excessive memory usage. This document explains JVM's role and how to correlate it with Elasticsearch APIs. The default JVM options in Elasticsearch are generally sufficient for most use cases, and modification is not recommended. A Java Virtual Machine (JVM) is an essential component of the Java Runtime Environment (JRE), bundled within the Java Development Kit (JDK). The JVM translates Java code into machine-executable instructions and manages memory, garbage collection, and security. JVM memory management involves the young generation (Eden, S0, S1) and the old generation, which stores long-lived objects. Garbage collection removes unused objects, reducing heap memory usage in a sawtooth pattern, which is the key to managing memory. The G1GC has a pause-time target for efficient garbage collection. The GET _nodes/_all/jvm and GET _nodes/stats/jvm APIs offer insights into Elasticsearch JVM configuration and metrics. Advanced users can use the jstat tool, part of the Java Development Kit, for real-time JVM statistics.