What happens if there is insufficient RAM and the server starts swapping?

• A. Confluence performance improves
• B. No effect on the JVM
• C. Swapping can interfere with the JVM garbage collector and slow Confluence
• D. Swapping prevents crashes

Answer: (C)

When the system runs out of RAM, the operating system starts swapping memory pages to disk. The JVM relies on fast in-memory access for its heap and native structures, so once pages are swapped out, the JVM has to pay the cost of disk I/O to fetch them back. This slows down garbage collection because the collector spends more time waiting on memory pages, causing longer pause times and reduced throughput. With GC and application threads contending for memory while some pages are on disk, Confluence can become noticeably slower and less responsive, especially under load.

Swapping does not improve stability or prevent crashes. It can mask memory pressure temporarily, but it often makes memory-related issues worse by introducing latency and the risk of longer GC pauses or OOM events when the system can’t reclaim memory quickly enough.

In short, swapping interrupts memory access patterns, slows garbage collection, and degrades Confluence performance rather than helping it.