Collect Metadata with the Tracing Agent

The Native Image tool relies on the static analysis of an application’s reachable code at runtime. However, the analysis cannot always completely predict all usages of the Java Native Interface (JNI), Java Reflection, Dynamic Proxy objects, or class path resources. Undetected usages of these dynamic features must be provided to the native-image tool in the form of metadata (precomputed in code or as JSON configuration files).

Here you will find information how to automatically collect metadata for an application and write JSON configuration files. To learn how to compute dynamic feature calls in code, see Reachability Metadata.

Tracing Agent #

GraalVM provides a Tracing Agent to easily gather metadata and prepare configuration files. The agent tracks all usages of dynamic features during application execution on a regular Java VM.

Enable the agent on the command line with the java command from the GraalVM JDK:

$JAVA_HOME/bin/java -agentlib:native-image-agent=config-output-dir=/path/to/config-dir/ ...

Note: -agentlib must be specified before a -jar option or a class name or any application parameters as part of the java command.

When run, the agent looks up classes, methods, fields, resources for which the native-image tool needs additional information. When the application completes and the JVM exits, the agent writes metadata to JSON files in the specified output directory (/path/to/config-dir/).

It may be necessary to run the application more than once (with different execution paths) for improved coverage of dynamic features. The config-merge-dir option adds to an existing set of configuration files, as follows:

$JAVA_HOME/bin/java -agentlib:native-image-agent=config-merge-dir=/path/to/config-dir/ ...                                                              ^^^^^

The agent also provides the following options to write metadata on a periodic basis:

config-write-period-secs=n: writes metadata files every n seconds; n must be greater than 0.
config-write-initial-delay-secs=n: waits n seconds before first writing metadata; defaults to 1.

For example:

$JAVA_HOME/bin/java -agentlib:native-image-agent=config-output-dir=/path/to/config-dir/,config-write-period-secs=300,config-write-initial-delay-secs=5 ...

The above command will write metadata files to /path/to/config-dir/ every 300 seconds after an initial delay of 5 seconds.

It is advisable to manually review the generated configuration files. Because the agent observes only executed code, the application input should cover as many code paths as possible.

The generated configuration files can be supplied to the native-image tool by placing them in a META-INF/native-image/ directory on the class path. This directory (or any of its subdirectories) is searched for the file named reachability-metadata.json that is then automatically included in the build process. Not all of those files must be present. When multiple files with the same name are found, all of them are considered.

To test the agent collecting metadata on an example application, go to the Build a Native Executable with Reflection guide.

Conditional Metadata Collection #

The agent can deduce metadata conditions based on their usage in executed code. Conditional metadata is mainly aimed towards library maintainers with the goal of reducing overall footprint.

To collect conditional metadata with the agent, see Conditional Metadata Collection.

Agent Advanced Usage #

Caller-based Filters #

By default, the agent filters dynamic accesses which Native Image supports without configuration. The filter mechanism works by identifying the Java method performing the access, also referred to as caller method, and matching its declaring class against a sequence of filter rules. The built-in filter rules exclude dynamic accesses which originate in the JVM, or in parts of a Java class library directly supported by Native Image (such as java.nio) from the generated configuration files. Which item (class, method, field, resource, etc.) is being accessed is not relevant for filtering.

In addition to the built-in filter, custom filter files with additional rules can be specified using the caller-filter-file option. For example: -agentlib:caller-filter-file=/path/to/filter-file,config-output-dir=...

Filter files have the following structure:

{ "rules": [
    {"excludeClasses": "com.oracle.svm.**"},
    {"includeClasses": "com.oracle.svm.tutorial.*"},
    {"excludeClasses": "com.oracle.svm.tutorial.HostedHelper"}
  ],
  "regexRules": [
    {"includeClasses": ".*"},
    {"excludeClasses": ".*\\$\\$Generated[0-9]+"}
  ]
}

The rules section contains a sequence of rules. Each rule specifies either includeClasses, which means that lookups originating in matching classes will be included in the resulting configuration, or excludeClasses, which excludes lookups originating in matching classes from the configuration. Each rule defines a pattern to match classes. The pattern can end in .* or .**, interpreted as follows: - .* matches all classes in the package and only that package; - .** matches all classes in the package as well as in all subpackages at any depth. Without .* or .**, the rule applies only to a single class with the qualified name that matches the pattern. All rules are processed in the sequence in which they are specified, so later rules can partially or entirely override earlier ones. When multiple filter files are provided (by specifying multiple caller-filter-file options), their rules are chained together in the order in which the files are specified. The rules of the built-in caller filter are always processed first, so they can be overridden in custom filter files.

In the example above, the first rule excludes lookups originating in all classes from package com.oracle.svm and from all of its subpackages (and their subpackages, etc.) from the generated metadata. In the next rule however, lookups from those classes that are directly in package com.oracle.svm.tutorial are included again. Finally, lookups from the HostedHelper class is excluded again. Each of these rules partially overrides the previous ones. For example, if the rules were in the reverse order, the exclusion of com.oracle.svm.** would be the last rule and would override all other rules.

The regexRules section also contains a sequence of rules. Its structure is the same as that of the rules section, but rules are specified as regular expression patterns which are matched against the entire fully qualified class identifier. The regexRules section is optional. If a regexRules section is specified, a class will be considered included if (and only if) both rules and regexRules include the class and neither of them exclude it. With no regexRules section, only the rules section determines whether a class is included or excluded.

For testing purposes, the built-in filter for Java class library lookups can be disabled by adding the no-builtin-caller-filter option, but the resulting metadata files are generally unsuitable for the build. Similarly, the built-in filter for Java VM-internal accesses based on heuristics can be disabled with no-builtin-heuristic-filter and will also generally lead to less usable metadata files. For example: -agentlib:native-image-agent=no-builtin-caller-filter,no-builtin-heuristic-filter,config-output-dir=...

Access Filters #

Unlike the caller-based filters described above, which filter dynamic accesses based on where they originate, access filters apply to the target of the access. Therefore, access filters enable directly excluding packages and classes (and their members) from the generated configuration.

By default, all accessed classes (which also pass the caller-based filters and the built-in filters) are included in the generated configuration. Using the access-filter-file option, a custom filter file that follows the file structure described above can be added. The option can be specified more than once to add multiple filter files and can be combined with the other filter options, for example, -agentlib:access-filter-file=/path/to/access-filter-file,caller-filter-file=/path/to/caller-filter-file,config-output-dir=....

Specify Configuration Files as Arguments #

A directory containing configuration files that is not part of the class path can be specified to native-image via -H:ConfigurationFileDirectories=/path/to/config-dir/. This directory must directly contain reachability-metadata.json or the formerly-used individual metadata files (jni-config.json, reflect-config.json, proxy-config.json, serialization-config.json, and resource-config.json). A directory with the same metadata files that is on the class path, but not in META-INF/native-image/, can be provided via -H:ConfigurationResourceRoots=path/to/resources/. Both -H:ConfigurationFileDirectories and -H:ConfigurationResourceRoots can also take a comma-separated list of directories.

Injecting the Agent via the Process Environment #

Altering the java command line to inject the agent can prove to be difficult if the Java process is launched by an application or script file, or if Java is even embedded in an existing process. In that case, it is also possible to inject the agent via the JAVA_TOOL_OPTIONS environment variable. This environment variable can be picked up by multiple Java processes which run at the same time, in which case each agent must write to a separate output directory with config-output-dir. (The next section describes how to merge sets of configuration files.) In order to use separate paths with a single global JAVA_TOOL_OPTIONS variable, the agent’s output path options support placeholders:

export JAVA_TOOL_OPTIONS="-agentlib:native-image-agent=config-output-dir=/path/to/config-output-dir-{pid}-{datetime}/"

The {pid} placeholder is replaced with the process identifier, while {datetime} is replaced with the system date and time in UTC, formatted according to ISO 8601. For the above example, the resulting path could be: /path/to/config-output-dir-31415-20181231T235950Z/.

Trace Files #

In the examples above, native-image-agent has been used to both keep track of the dynamic accesses on a JVM and then to generate a set of configuration files from them. However, for a better understanding of the execution, the agent can also write a trace file in JSON format that contains each individual access:

$JAVA_HOME/bin/java -agentlib:native-image-agent=trace-output=/path/to/trace-file.json ...

The native-image-configure tool can transform trace files to configuration files. The following command reads and processes trace-file.json and generates a set of configuration files in the directory /path/to/config-dir/:

native-image-configure generate --trace-input=/path/to/trace-file.json --output-dir=/path/to/config-dir/

Interoperability #

The agent uses the JVM Tool Interface (JVMTI) and can potentially be used with other JVMs that support JVMTI. In this case, it is necessary to provide the absolute path of the agent:

/path/to/some/java -agentpath:/path/to/graalvm/jre/lib/amd64/libnative-image-agent.so=<options> ...

Experimental Options #

The agent has options which are currently experimental and might be enabled in future releases, but can also be changed or removed entirely. See the ExperimentalAgentOptions.md guide.

Native Image Configure Tool #

When using the agent in multiple processes at the same time as described in the previous section, config-output-dir is a safe option, but it results in multiple sets of configuration files. The native-image-configure tool can be used to merge these configuration files:

native-image-configure generate --input-dir=/path/to/config-dir-0/ --input-dir=/path/to/config-dir-1/ --output-dir=/path/to/merged-config-dir/

This command reads one set of configuration files from /path/to/config-dir-0/ and another from /path/to/config-dir-1/ and then writes a set of configuration files that contains both of their information to /path/to/merged-config-dir/. An arbitrary number of --input-dir arguments with sets of configuration files can be specified. See native-image-configure help for all options.