Вячеслав Блинов «Java Garbage Collection: A Performance Impact»
0 likes586 views
This document discusses Java garbage collection and provides an overview of common GC algorithms, their performance impacts, and basic tuning strategies. It describes how the generational heap is divided and explains that GC pauses can significantly impact performance. Different algorithms like the serial, parallel, CMS and G1 collectors are introduced along with considerations for choosing a collector based on heap size, CPU usage, and pause requirements. Guidelines are provided for sizing the heap and generations as well as enabling adaptive sizing.
Вячеслав Блинов «Java Garbage Collection: A Performance Impact»
- 1. Java Garbage Collection: A Performance Impact Blynov Viacheslav Lead Software Developer 24 November 2014 www.luxoft.com
- 3. Introduction Agenda Garbage Collection Overview Java GC Algorithms Basic GC Tuning www.luxoft.com
- 5. GC Purpose Objects that are referenced are said to be live Objects that are no longer referenced are considered dead and termed garbage Garbage collector is responsible for: allocating memory ensuring that any referenced objects remain in memory recovering memory used by objects that are no longer reachable from references in executing code www.luxoft.com
- 7. GC performance impact Collector needs computational resources (CPU cycles) to perform garbage collection As garbage collection involves moving objects in memory a collector must ensure that no thread is using these objects www.luxoft.com The pauses when all application threads are stopped are called stop-the-world pauses These pauses generally have the greatest impact on the performance of an application, and minimizing those pauses is the key consideration when tuning GC.
- 14. Summary all GC algorithms divide the heap into old and young generation all GC algorithm employ stop-the-world approach to clearing objects from young generation, which is usually a very quick operation www.luxoft.com
- 15. www.luxoft.com
- 16. “Client” and “Server” JVM types Depending on underlying hardware platform and version JVM can act as “client” of “server” VM. This affect the choice of JIT compiler and default GC algorithm. “client” platform is usually 32-bit and has 1 CPU “server” platform is usually 64-bit (but 32-bit is also possible) and has several CPUs www.luxoft.com
- 17. GC Algorithms Serial garbage collector (-XX:+UseSerialGC) Throughput collector (-XX:+UseParallelGC , -XX:+UseParallelOldGC) CMS collector (-XX:+UseConcMarkSweepGC, -XX:+UseParNewGC) G1 collector (-XX:+UseG1GC) www.luxoft.com
- 18. Serial garbage collector default collector for client-class platforms (32-bit JVMs on Windows or single-processor www.luxoft.com machines) uses single thread to process heap stops all application threads for both minor and full GC Usage cases: no low-pause requirements “client-style” single-CPU environment very small heap (few hundred MBs) several JVMs running on single platform (number of JVM > number of available CPUs)
- 19. Throughput (parallel) garbage collector default collector for server-class machines (multi-CPU Unix machines or any 64-bit JVM) utilizes multiple threads for garbage collection to gain speed and minimize pauses stops all application threads for both minor and full GC -XX:+UseParallelGC enables multi-threaded collection of young generation and single-threaded www.luxoft.com old-generation collection/compaction -XX:+UseParallelOldGC enables multi-threaded collection of young generation and multi-threaded old-generation collection/compaction Usage cases: multi-CPU are available large heap size and many object created/discarded
- 20. CMS (Concurrent Mark Sweep) collector designed to eliminate long pauses associated with full GC cycles stops all application threads during minor GC uses different algorithm to collect young generation (-XX:+UseParNewGC) uses one or more background threads to periodically scan through the old generation and discard unused objects. This makes CMS a low-paused collector do not perform any compaction in case of CPU unavailability and/or heap fragmentation – fallback to serial collector by default does not collect permgen Usage cases: low pause requirement and available CPU resources in case of single-CPU machine can be used with -XX:+CMSIncrementalMode (deprecated in Java 8) www.luxoft.com
- 21. G1 (Garbage First) garbage collector designed to process large heaps (more than 4 Gb) with minimal pauses divides the heap into separate regions performs incremental compaction of old generation by copying data between regions www.luxoft.com
- 22. G1 (Garbage First) garbage collector www.luxoft.com
- 23. Summary (choosing GC algorithm) Serial GC is best only for application with heap <= 100 Mb Batch jobs which consume all available CPUs will get better performance with concurrent collector Batch jobs which DON’T consume all CPUs could get better performance with throughput collector When measuring response time the choice between throughput and concurrent collectors depends on CPU availability Most of the time CMS should overperform G1 for heaps < 4 Gb For large heaps G1 is better because of the way it can divide work between different threads and heap regions www.luxoft.com
- 24. www.luxoft.com
- 25. Sizing the heap Too small heap -> too much time spent in GC Main rule – never to specify heap more than the amount of available physical memory -Xms – initial heap size -Xmx – maximum heap size www.luxoft.com
- 26. Sizing the Generations -XX:NewRatio=N – sets the ration of young generation to old -XX:NewSize=N – sets the size of young generation -XX:MaxNewSize=N – maximum size for young generation www.luxoft.com
- 27. Sizing Permgen and Metaspace Java 7: -XX:PermSize=N -XX:MaxPermSize=N Java 8: -XX:MetaspaceSize=N -XX:MaxMetaspaceSize=N www.luxoft.com
- 28. Adaptive Sizing JVM can try to find optimal performance according to its policies and configuration Adaptive sizing controls how the JVM alters the ratio of young generation to old Adjusting generation sizes is base on GC algorithms attempts to meet their pause goals Adaptive tuning can be disabled for small performance boost (usually not recommended) Command line argument differs for different GC algorithms. For, example for throughput collector: -XX:+UseAdaptiveSizePolicy – whether to use adaptive policy (true by default) -XX:MaxGCPauseMillis=nnn – maximal GC pause we can tolerate -XX:GCTimeRatio=nnn - hint to the virtual machine that it's desirable that not more than 1 / (1 + nnn) of the application execution time be spent in the collector www.luxoft.com