If a device with slow memory leak runs for a long period of time, unexpected network exceptions might occur. This feature monitors the continuous memory usage growth and sends alarms to help identify whether memory leak occurs on the device.
This feature performs calculation based on the history data of the memory usage monitoring instance. When the memory usage of the device exceeds the level-2 threshold, the calculation is triggered. To use this feature, you must enable memory usage monitoring for the device, and configure and enable the level-2 threshold alarm.
To detect very slight memory leak, this feature supports the daily data growth rate and weekly data growth rate algorithms. If the memory usage grows within 30 days but the growth rate is too low to trigger alarms, the system automatically switches to the weekly data growth rate algorithm and performs calculation again.