Glossary Updates12 new terms added to the glossaries · October 2, 2026, 22:44 CEST
AI TechDocKnowledge

Glossary · Automation software engineering and architecture

Memory leak

German: Speicherleck

In software engineering, a memory leak is a defect in which a program keeps allocated memory it no longer needs, so available memory shrinks over time until performance degrades or the program fails.

  • Software engineering

In one sentence

A memory leak is a defect where a program never releases memory it no longer needs, so free memory shrinks until the system slows or fails.

Example

An HMI application allocates a new buffer for every alarm message but never frees it; after three weeks of operation the panel freezes.

How it applies

  • Operation: Leaks are typical long-run failures: the software passes every short test and fails only after days or weeks of continuous operation, which is normal for machines and plants.
  • Verification: Endurance tests, memory profilers and Dynamic code analysis find leaks; monitoring memory use in the field through Observability tooling shows a slow upward trend before failure.
  • Documentation: Known leaks and their workarounds (for example a scheduled restart) belong in the Release note and in the troubleshooting section of the manual until a fix is released.

Memory leak vs. dangling pointer

A memory leak keeps memory that should be released; a Dangling pointer refers to memory that was already released. Both are errors in managing Dynamic memory allocation, but a dangling pointer typically causes corrupted data or crashes, while a leak causes gradual exhaustion.

By knowledge.aitechdoc.world · Published September 26, 2026 · Last reviewed

Source: AI TechDoc Knowledge editorial definition, based on common software engineering practice

Definitions follow the cited standards and specifications. Where a source is a copyrighted publication, such as an ISO, IEC or EN standard, the definition is a close paraphrase, not a verbatim quotation, so as not to infringe copyright. We recommend reading the original publication. The sections “How it applies” are editorial commentary by AI TechDoc Knowledge and are not part of any standard.

Seen a mistake? Send us a note!