Open albilu opened 3 years ago
Hi,
this error typically occurs, when the thread that generates the tests is in garbage collection, when it is supposed to stop. Often this can be prevented by setting the max heap space differently. Note that both to much and not enough heap is possible. I typically use heap sizes ~2 - 4g when running EvoSuite.
Garbage collection is managed by the JVM. Therefore, we can hardly do anything to prevent this from happening (,if it is the cause)
How to set the heap size of the JVM: https://docs.oracle.com/cd/E19563-01/819-4438/gavou/index.html
Hi, so tried by modifying the JVM heap like this: "C:\Program Files\Java\jdk1.8.0_181\bin\java" -Xms2048m -Xmx4096m -jar C:\Users\xxx\Downloads\evosuite-1.1.0.jar -target build\classes
but still i got the issue.
Not sure if this is really a memory issue, but the command line C:\Program Files\Java\jdk1.8.0_181\bin\java" -Xms2048m -Xmx4096m -jar C:\Users\xxx\Downloads\evosuite-1.1.0.jar -target build\classes
only sets the memory for the master process. EvoSuite will spawn a client process in which it runs test generation, and that will use the system default. You can explicitly change that using the -mem
option, e.g.:
java -jar evosuite-1.1.0.jar -mem 4096 -target build/classes
If this doesn't help, is the AtomicLogHistory
class available somewhere to reproduce this?
Context
Running Evosuite on a bunch of classes
Steps to Reproduce
EvoSuite Arguments
1-"C:\Program Files\Java\jdk1.8.0_181\bin\java" -jar C:\Users\xxx\Downloads\evosuite-1.1.0.jar -Duse_separate_classloader=false -Dreflection_start_percent=0.0 -Dp_reflection_on_private=1.0 -setup build\classes build\public-package-jars*.jar build\cluster\modules\ext*.jar
2-"C:\Program Files\Java\jdk1.8.0_181\bin\java" -jar C:\Users\xxx\Downloads\evosuite-1.1.0.jar -target build\classes
Current Result
The test generation stuck on a file and it is impossible to continue
[MASTER] 15:27:04.790 [logback-2] ERROR ConcreteClassAnalyzer - Removing class from inheritance tree
Expected result
Test generation Completed
Additional info
Issue is the same when increasing the timeout by using -Dextra_timeout=180 Issue is the same as #96