You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
- This project contains a stress version of the Idle Micro Benchmark test which aims to test the Idle detection and management feature available in J9 vm. For more information on the feature and related -XX JVM options, visit : https://www.ibm.com/support/knowledgecenter/SSYKE2_8.0.0/com.ibm.java.lnx.80.doc/diag/appendixes/cmdline/commands_jvm_xx.html.
3
+
- The test has been written to have alternating active and idle cycles.
4
+
- Active cycles are periods of time when the test will perform CPU and Memory Intensive operations to consume CPU cycles and Heap Memory.
5
+
- Idle cycles are longer periods of time where the test sleeps/ performs no activity.
6
+
- Idle detection and management are enabled/disabled via newly added JVM options.
7
+
- The Idle detection feature is available on all platforms whereas Idle management feature is supported only on Linux x86 platforms.
8
+
- The test exercises the newly added JVM options and checks if they behave as expected under stress conditions.
9
+
10
+
- There are 3 variations to this test
11
+
12
+
1. MinIdleWaitTime
13
+
- Tests the -XX:IdleTuningMinIdleWaitTime option. This is the minimum amount of time the application needs to be idle to be detected as Idle.
14
+
- Applicable on all platforms.
15
+
- Uses the following Java arguments while running the test
- Tests the -XX:+IdleTuningCompactOnIdle and when enabled, GC performs compaction on the heap during idle cycle to create contiguious free memory spaces for newer allocations.
26
+
- Applicable only on Linux_x86-32 and Linux_x86-64 platforms as of now.
27
+
- Uses the following JVM arguments while running the test
0 commit comments