A generic troubleshooting reference for LIMS/application server memory utilization incidents.

1. Issue

A production application server hosting a STARLIMS Application and Manager service began experiencing severe memory exhaustion. Memory utilization steadily climbed to approximately 99-100% on a server with substantial installed RAM, while CPU usage remained moderate, indicating the instability was memory-driven rather than processor-driven. As memory pressure increased, the server became unresponsive and had to be restarted manually on multiple occasions to restore service to end users.

Investigation of the running processes showed multiple IIS worker processes (w3wp.exe) and Manager service “Starlims worker process” whose memory consumption grew continuously over time until the server became unstable.  

2. Possible Root Cause

Common root causes of high worker process memory consumption are listed below. Most issues are resolved by addressing these points.

1. SQL queries with no proper WHERE clause — full table scans pulling large result sets into memory.
2. Connection pool leaks — connections/cursors not closed properly.
3. Inefficient JOINs — missing join conditions causing exponential row explosion.
4. In-memory aggregation of large datasets — doing GROUP BY/sorting in application code instead of pushing down to the database.