Java Security Manager is a security component in the JVM that controls the permissions of your code. It defines the mapping of code source and permissions through the Policy file, checks sensitive operations at runtime, and throws a SecurityException if there is no authorization. Enabled to specify the policy file by adding the -Djava.security.manager and -Djava.security.policy parameters at startup. The policy file uses grant to define codeBase and its permissions. Common permissions include FilePermission, SocketPermission, RuntimePermission, and PropertyPermission. Applicable scenarios include running untrusted plug-ins, sandbox execution, and enterprise applications that require permission isolation. Pay attention to third-party libraries' dependency on permissions, risk of wildcard abuse, and security risks of AllPermission.

Java Security Manager is a mechanism in the Java platform for enhancing program security. It allows you to define fine-grained security policies that control what code can and cannot do. Although many frameworks and applications no longer enable Security Manager by default, it is still useful in certain scenarios (such as running untrusted code).

What is Java Security Manager?

Simply put, Security Manager is a component in the JVM. It checks whether the code has permission to perform certain sensitive operations at runtime, such as reading and writing files, establishing network connections, loading classes, etc. If there is no authorization, SecurityException will be thrown.

Its core is to define which code has what permissions through the Policy file . You can understand it as a "security guard" who is responsible for guarding your Java programs and preventing malicious or non-compliant operations from being executed.

How to enable Security Manager?

The way to enable Security Manager is simple, you only need to add two parameters when starting your Java application:

 java -Djava.security.manager -Djava.security.policy=my.policy MyApp

in:

• -Djava.security.manager means to enable Security Manager.
• -Djava.security.policy=my.policy specifies the policy file to use.

If you do not specify a policy file, the JVM uses the default security policy, which is usually very restrictive and may cause the program to not function properly.

How to write a Policy file?

Policy files are used to define the mapping relationship between code source and permissions. The format is roughly as follows:

 grant codeBase "file:/path/to/myapp/-" {
    permission java.io.FilePermission "/tmp/-", "read,write";
    permission java.net.SocketPermission "*:1024-65535", "connect,resolve";
};

The above configuration means:
All code from /path/to/myapp/ directory can read and write files under /tmp , and can connect to ports 1024 to 65535 of any host.

Some common permission types include:

• FilePermission : Controls file reading and writing
• SocketPermission : Control network connections
• RuntimePermission : Controls access to system resources, such as exiting the JVM
• PropertyPermission : Controls the reading and writing of system properties

When is Security Manager suitable?

Although Security Manager is not recommended for most modern application frameworks such as Spring Boot, it still works in the following scenarios:

• Run a plug-in or script submitted by the user (such as an online review system)
• Execute untrusted code in a sandbox environment (such as applets or modular plug-in systems)
• Enterprise-level applications that require fine control of permissions for different modules

It should be noted that Security Manager is only part of the Java security system and it cannot replace other security measures, such as code signatures, encrypted communications, etc.

Some things to pay attention to in small details

• If you are using third-party libraries, they may rely on certain system permissions, such as creating threads or accessing environment variables. If Security Manager does not release these permissions, the program will report an error.
• Policy files support wildcards, but do not abuse them, otherwise they may cause security vulnerabilities.
• Be especially careful when using AllPermission , which is equivalent to giving a green light to a certain code and skipping all security checks.

Basically that's it. Security Manager is not complicated, but it is easy to ignore details. If you're just developing a normal app, you might not need it; but once permission isolation or sandboxing is involved, it's a tool worth knowing.

The above is the detailed content of Understanding the Java Security Manager. For more information, please follow other related articles on the PHP Chinese website!