JVM memory management------Garbage collector parameters detailed explanation
This article is the last one related to GC. This time LZ just lists the key parameters related to the garbage collector in hotspot JVM, as well as the explanation of each parameter. Without further ado, let’s get started.
Garbage collector selection parameters
UseSerialGC: Enable this parameter to use serial & serial old collector (client mode default value).
UseParNewGC: Enable this parameter to use ParNew & serial old collector (not recommended).
UseConcMarkSweepGC: Turn on this parameter to use ParNew & CMS (serial old is the substitute) collector.
UseParallelGC: Turn on this parameter to use the parallel scavenge & parallel old collector (server mode default value).
UseParallelOldGC: Turn on this parameter to use the parallel old collector in the old generation (this parameter is no longer useful after JDK1.5).
Parameters related to the size of each JVM memory area
Xms: the initial value of the heap. The default is 1/64 of the physical memory, and the maximum does not exceed 1G.
Xmx: The maximum value of the heap. The default is 1/4 of the physical memory, and the maximum does not exceed 1G.
Xmn: The size of the new generation.
Xss: Thread stack size.
PermSize: Initial size of the permanent generation. The default is 1/64 of the physical memory, and the maximum does not exceed 1G.
MaxPermSize: Maximum permanent generation size. The default is 1/4 of the physical memory, and the maximum does not exceed 1G.
NewRatio: The ratio of the new generation to the old generation. For example, if it is 3, the new generation occupies 1/4 of the heap and the old generation occupies 3/4.
SurvivorRatio: Adjust the ratio of the eden area and the survivor area in the new generation. The default is 8, that is, the eden area is 80% of the size, and the two survivors are 10% of the size respectively. (Note: This parameter setting is a method to solve the problem of halving the memory of the replication algorithm in the chapter explaining the replication algorithm. The eden area is the 80% part in the chapter of the replication algorithm, and the survivor area is two 10% That part.)
Garbage collector performance general parameters
PretenureSizeThreshold: Increase the object size of the young generation. The default is 0. For example, if it is set to 10M, objects exceeding 10M will not be allocated in the eden area, but will directly enter the old generation.
MaxTenuringThreshold: The maximum age for promotion to the old generation. The default is 15. For example, if it is set to 10, the object will be placed in the old generation after 10 ordinary GC times.
DisableExplicitGC: Disable System.gc().
Parallel collector parameters
ParallelGCThreads: The number of threads opened during recycling. The default is equal to the number of CPUs.
GCTimeRatio: Set the throughput of the system. For example, if it is set to 99, the GC time ratio is 1/1+99=1%, which means the required throughput is 99%. If it cannot be satisfied, the size of the new generation will be reduced.
MaxGCPauseMillis: Set the maximum pause time for garbage collection. If the setting value cannot be met, the new generation size will be reduced first. If it is still not met, the throughput will be sacrificed.
Concurrent collector parameters
CMSInitiatingOccupancyFraction: The memory ratio that triggers the CMS collector. For example, 60% means that when the memory reaches 60%, CMS concurrent collection will start.
UseCMSCompactAtFullCollection: This has been mentioned before and is used to send memory deflation after each time the CMS collector cleans up garbage.
CMSFullGCsBeforeCompaction: Set to trigger a memory compaction after several CMS garbage collections.
Conclusion
The GC related series is basically over. This article is just a list. Then let’s enter the source code world of virtual machines.
The above is the detailed explanation of JVM memory management------garbage collector parameters. For more related content, please pay attention to the PHP Chinese website (www.php.cn)!

Java is widely used in enterprise-level applications because of its platform independence. 1) Platform independence is implemented through Java virtual machine (JVM), so that the code can run on any platform that supports Java. 2) It simplifies cross-platform deployment and development processes, providing greater flexibility and scalability. 3) However, it is necessary to pay attention to performance differences and third-party library compatibility and adopt best practices such as using pure Java code and cross-platform testing.

JavaplaysasignificantroleinIoTduetoitsplatformindependence.1)Itallowscodetobewrittenonceandrunonvariousdevices.2)Java'secosystemprovidesusefullibrariesforIoT.3)ItssecurityfeaturesenhanceIoTsystemsafety.However,developersmustaddressmemoryandstartuptim

ThesolutiontohandlefilepathsacrossWindowsandLinuxinJavaistousePaths.get()fromthejava.nio.filepackage.1)UsePaths.get()withSystem.getProperty("user.dir")andtherelativepathtoconstructthefilepath.2)ConverttheresultingPathobjecttoaFileobjectifne

Java'splatformindependenceissignificantbecauseitallowsdeveloperstowritecodeonceandrunitonanyplatformwithaJVM.This"writeonce,runanywhere"(WORA)approachoffers:1)Cross-platformcompatibility,enablingdeploymentacrossdifferentOSwithoutissues;2)Re

Java is suitable for developing cross-server web applications. 1) Java's "write once, run everywhere" philosophy makes its code run on any platform that supports JVM. 2) Java has a rich ecosystem, including tools such as Spring and Hibernate, to simplify the development process. 3) Java performs excellently in performance and security, providing efficient memory management and strong security guarantees.

JVM implements the WORA features of Java through bytecode interpretation, platform-independent APIs and dynamic class loading: 1. Bytecode is interpreted as machine code to ensure cross-platform operation; 2. Standard API abstract operating system differences; 3. Classes are loaded dynamically at runtime to ensure consistency.

The latest version of Java effectively solves platform-specific problems through JVM optimization, standard library improvements and third-party library support. 1) JVM optimization, such as Java11's ZGC improves garbage collection performance. 2) Standard library improvements, such as Java9's module system reducing platform-related problems. 3) Third-party libraries provide platform-optimized versions, such as OpenCV.

The JVM's bytecode verification process includes four key steps: 1) Check whether the class file format complies with the specifications, 2) Verify the validity and correctness of the bytecode instructions, 3) Perform data flow analysis to ensure type safety, and 4) Balancing the thoroughness and performance of verification. Through these steps, the JVM ensures that only secure, correct bytecode is executed, thereby protecting the integrity and security of the program.


Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

WebStorm Mac version
Useful JavaScript development tools

SublimeText3 Chinese version
Chinese version, very easy to use

Dreamweaver CS6
Visual web development tools

SAP NetWeaver Server Adapter for Eclipse
Integrate Eclipse with SAP NetWeaver application server.

MantisBT
Mantis is an easy-to-deploy web-based defect tracking tool designed to aid in product defect tracking. It requires PHP, MySQL and a web server. Check out our demo and hosting services.
