Taming The JVM
13 likes1,166 views
This document discusses techniques for tuning the Java Virtual Machine (JVM) to improve application performance. It covers JVM configuration options like heap size, garbage collection algorithms, and generation sizing. Tools for memory analysis are presented, including jmap, JConsole, and JRockit Mission Control. Future improvements like the G1 garbage collector in JDK 7 and appliance solutions are also mentioned. A case study demonstrates how profiling and adjusting JVM settings helped resolve performance issues affecting a Java EE application.
Taming The JVM
- 1. Taming the JVM Tuning Tools/Techniques Mike Prasad Or downright grimacing! ..can be adventurous!
- 2. Objectives ! Application challenges ! JVM Configuration ! Heap Analysis Tools ! Future/Alternate Solutions ! Case Study
- 3. Application Challenges ! Java environment dramatically changed ! Multi-core, multi-GB environments ! Understanding JVM important
- 4. Java Memory ! Divided into two areas – Stack & Heap Stack ! associated with methods ! intermediate calcs Heap ! All objects created using “new” keyword ! GC responsible for cleaning dead/unwanted objects ! Heap sizes controlled by command line options
- 5. Java Garbage Collector ! Objects either in young or tenured generations ! Minor GC when young objects die ! Surviving objects moved to tenured gen ! Full GC when tenured gen needs to be collected http://java.sun.com/docs/hotspot/gc5.0/gc_tuning_5.html
- 6. Generation Sizing ! Default values typically used ! This not always sufficient ! Requires adjusting generation size
- 7. VM Options New Space Behavioral Stack -XX:NewSize -XX:-UseConcMarkSweepGC new gen heap size old gen concurrent mark-sweep -Xss -XX:MaxNewSize -XX:-UseParallelGC size set stack parallel garbage collection max new gen heap size -XX:-UseSerialGC serial garbage collection -XX:NewRatio ratio of new/old gen size -XX:SurvivorRatio eden/survivor space size Heap Debugging -Xms initial heap size Perm Space -XX:-PrintConcurrentLocks -Xmx maximum heap size Prints concurrent locks -XX:PermSize initial perm size -XX:-PrintGCTimeStamps Print timestamps at gc -Xmn -XX:MaxPermSize max perm generation heap size -XX:-PrintTenuringDistributionyoung size Print tenuring age info -Xnoclassgc eliminates load/unload overhead
- 8. Memory Analysis Tools ! JMap ! JConsole ! Visual GC ! JRockit Mission Control ! Others ! JStat ! HPROF
- 9. jmap ! Prints memory–related statistics ! Other options ! –heap ! –histo ! -permstat ! To programmatically monitor use: - java.lang.Runtime.totalMemory() - java.lang.Runtime.freeMemory()
- 10. jmap-heap Used to obtain : ! gc name ! algorithm details ! #threads used for parallel gc ! heap configuration ! heap usage summary
- 11. jmap -histo ! Class-wise histogram of the heap ! Prints out: ! the number of instances ! total amount of memory consumed ! the fully qualified class name
- 12. jmap -permstat ! Get perm gen statistics ! Configuring perm gen size important ! Web apps load large number of classes (JPs etc) ! “too many” classes = OutOfMemoryError
- 13. JConsole ! Monitoring and management tool ! JMX–compliant attaches to a running JVM ! JDK 5.0 and above
- 14. Visual GC ! Monitors ! Garbage Collector ! Compiler ! Class loader ! Superior to JConsole
- 15. JRockit Mission Control ! Associated with JRockit JVM ! Tool Suite : – visualize gc and other perf stats – runtime performance profiling tool – memory-analysis tool
- 16. Future ! G1 a new GC in JDK 7 Supports ! Parallelism : Uses all available CPUs ! Generational: same as HotSpot GC's ! Compaction : performs heap compaction
- 17. Appliance Solution ! Theoretical heap limit of 32-bit JVM is 4G ! Azul appliance provides ! hardware-assisted pauseless gc feature ! Up to 864 cores with 768GB heap! ! Xfer workload from app server to appliance
- 18. Case Study ! JEE App – Struts/EJB ! On WebLogic ! High volume app ! Overnight performance issues ! Full GCs every 1.5 minutes
- 19. Case Study Original VM Settings Attempted # of Settings -Server -XX:-UseConcMarkSweepGC –Xms2048M ProblemPizza + ? solved Coke = solution -XX:-UseParallelGC –Xmx2048M , , нет , nein, αριθ. -XX:-UseHailMaryGC (made it –Xmn512M worse!) Problem was with the code!
- 20. Moral FinallyStory Of the Apply rigor forJVM, test, the developer! test Don’t lookthe unit tamein the wrong place! Spare – garbage profile, load Why women live longer!