UPDATE最新版本的 dropwizard 支持开箱即用的日志配置
我在尝试使用单独的文件设置 Dropwizard (0.8.4) 时遇到了同样的问题。我遇到了同样的问题。所以我更深入地挖掘并找到了适合我的解决方案（不是最干净的，但我似乎无法以不同的方式工作）。
问题是LoggingFactory#configure自动将每个附加程序添加到根目录。这不是很理想，因此需要覆盖。我所做的是：
- 覆盖
LoggingFactory.
这有点混乱，因为有一些东西需要复制：（这是我的实现：
import java.io.PrintStream;
import java.lang.management.ManagementFactory;
import java.util.Map;
import javax.management.InstanceAlreadyExistsException;
import javax.management.MBeanRegistrationException;
import javax.management.MBeanServer;
import javax.management.MalformedObjectNameException;
import javax.management.NotCompliantMBeanException;
import javax.management.ObjectName;
import org.slf4j.LoggerFactory;
import org.slf4j.bridge.SLF4JBridgeHandler;
import com.codahale.metrics.MetricRegistry;
import com.codahale.metrics.logback.InstrumentedAppender;
import com.fasterxml.jackson.annotation.JsonIgnore;
import com.fasterxml.jackson.annotation.JsonProperty;
import com.google.common.collect.ImmutableMap;
import ch.qos.logback.classic.Level;
import ch.qos.logback.classic.Logger;
import ch.qos.logback.classic.LoggerContext;
import ch.qos.logback.classic.PatternLayout;
import ch.qos.logback.classic.jmx.JMXConfigurator;
import ch.qos.logback.classic.jul.LevelChangePropagator;
import ch.qos.logback.classic.spi.ILoggingEvent;
import ch.qos.logback.core.Appender;
import ch.qos.logback.core.util.StatusPrinter;
import io.dropwizard.logging.AppenderFactory;
import io.dropwizard.logging.LoggingFactory;
public class BetterDropWizardLoggingConfig extends LoggingFactory {
@JsonIgnore
final LoggerContext loggerContext;
@JsonIgnore
final PrintStream configurationErrorsStream;
@JsonProperty("loggerMapping")
private ImmutableMap<String, String> loggerMappings;
private static void hijackJDKLogging() {
SLF4JBridgeHandler.removeHandlersForRootLogger();
SLF4JBridgeHandler.install();
}
public BetterDropWizardLoggingConfig() {
PatternLayout.defaultConverterMap.put("h", HostNameConverter.class.getName());
this.loggerContext = (LoggerContext) LoggerFactory.getILoggerFactory();
this.configurationErrorsStream = System.err;
}
private Logger configureLevels() {
final Logger root = loggerContext.getLogger(org.slf4j.Logger.ROOT_LOGGER_NAME);
loggerContext.reset();
final LevelChangePropagator propagator = new LevelChangePropagator();
propagator.setContext(loggerContext);
propagator.setResetJUL(true);
loggerContext.addListener(propagator);
root.setLevel(getLevel());
for (Map.Entry<String, Level> entry : getLoggers().entrySet()) {
loggerContext.getLogger(entry.getKey()).setLevel(entry.getValue());
}
return root;
}
@Override
public void configure(MetricRegistry metricRegistry, String name) {
hijackJDKLogging();
final Logger root = configureLevels();
for (AppenderFactory output : getAppenders()) {
Appender<ILoggingEvent> build = output.build(loggerContext, name, null);
if(output instanceof MappedLogger && ((MappedLogger) output).getLoggerName() != null) {
String appenderName = ((MappedLogger) output).getLoggerName();
String loggerName = loggerMappings.get(appenderName);
Logger logger = this.loggerContext.getLogger(loggerName);
logger.addAppender(build);
} else {
root.addAppender(build);
}
}
StatusPrinter.setPrintStream(configurationErrorsStream);
try {
StatusPrinter.printIfErrorsOccured(loggerContext);
} finally {
StatusPrinter.setPrintStream(System.out);
}
final MBeanServer server = ManagementFactory.getPlatformMBeanServer();
try {
final ObjectName objectName = new ObjectName("io.dropwizard:type=Logging");
if (!server.isRegistered(objectName)) {
server.registerMBean(new JMXConfigurator(loggerContext, server, objectName), objectName);
}
} catch (MalformedObjectNameException | InstanceAlreadyExistsException | NotCompliantMBeanException
| MBeanRegistrationException e) {
throw new RuntimeException(e);
}
configureInstrumentation(root, metricRegistry);
}
private void configureInstrumentation(Logger root, MetricRegistry metricRegistry) {
final InstrumentedAppender appender = new InstrumentedAppender(metricRegistry);
appender.setContext(loggerContext);
appender.start();
root.addAppender(appender);
}
}
正如您所看到的，我遗憾地不得不复制/粘贴一些私人成员和方法以使事情按预期工作。
我添加了一个新字段：
@JsonProperty("loggerMapping")
private ImmutableMap<String, String> loggerMappings;
这允许我为每个记录器配置映射。这不是开箱即用的，因为我无法获取名称（dropwizard 默认附加程序名称，非常不方便......）
因此，我添加了一个新的记录器，在我的例子中，它还进行主机名替换，这是我出于不同原因需要的。为此我覆盖了旧的FileAppenderFactory并实现我自己的接口MappedLogger。此处实现：
import com.fasterxml.jackson.annotation.JsonProperty;
import com.fasterxml.jackson.annotation.JsonTypeName;
import ch.qos.logback.classic.LoggerContext;
import ch.qos.logback.classic.spi.ILoggingEvent;
import ch.qos.logback.core.FileAppender;
import ch.qos.logback.core.rolling.RollingFileAppender;
import io.dropwizard.logging.AppenderFactory;
import io.dropwizard.logging.FileAppenderFactory;
@JsonTypeName("hostnameFile")
public class HostnameFileAppender extends FileAppenderFactory implements AppenderFactory, MappedLogger {
private static String uuid = UUID.randomUUID().toString();
@JsonProperty
private String name;
public void setCurrentLogFilename(String currentLogFilename) {
super.setCurrentLogFilename(substitute(currentLogFilename));
}
private String substitute(final String pattern) {
String substitute = null;
try {
substitute = InetAddress.getLocalHost().getHostName();
} catch (UnknownHostException e) {
System.err.println("Failed to get local hostname:");
e.printStackTrace(System.err);
substitute = uuid;
System.err.println("Using " + substitute + " as fallback.");
}
return pattern.replace("${HOSTNAME}", substitute);
}
@Override
public void setArchivedLogFilenamePattern(String archivedLogFilenamePattern) {
super.setArchivedLogFilenamePattern(substitute(archivedLogFilenamePattern));
}
@Override
public String getLoggerName() {
return name;
}
}
请注意，为了添加新的 json 类型，您必须遵循 JavaDocAppenderFactory（将 Meta-inf 添加到类路径并使新的附加程序可发现）
到目前为止一切顺利，我们现在有了一个可以获取记录器映射的配置，我们有一个可以采用可选名称的记录器。
在配置方法中，我现在将这两者联系在一起：
for (AppenderFactory output : getAppenders()) {
Appender<ILoggingEvent> build = output.build(loggerContext, name, null);
if(output instanceof MappedLogger && ((MappedLogger) output).getLoggerName() != null) {
String appenderName = ((MappedLogger) output).getLoggerName();
String loggerName = loggerMappings.get(appenderName);
Logger logger = this.loggerContext.getLogger(loggerName);
logger.addAppender(build);
} else {
root.addAppender(build);
}
}
为了向后兼容，我保留了默认行为。如果没有定义名称，追加器将被添加到根记录器中。否则，我将解析类型记录器并根据需要向其添加附加程序。
最后但并非最不重要的一点是旧的 yaml 配置：
logging:
# The default level of all loggers. Can be OFF, ERROR, WARN, INFO, DEBUG, TRACE, or ALL.
level: INFO
loggers:
"EVENT" : INFO
loggerMapping:
# for easier search this is defined as: appenderName -> loggerName rather than the other way around
"eventLog" : "EVENT"
appenders:
- type: console
threshold: ALL
logFormat: "myformat"
- type: hostnameFile # NOTE THE NEW TYPE WITH HOSTNAME RESOLVE
currentLogFilename: /Users/artur/tmp/log/my-${HOSTNAME}.log
threshold: ALL
archive: true
archivedLogFilenamePattern: mypattern
archivedFileCount: 31
timeZone: UTC
logFormat: "myFormat"
- type: hostnameFile
name: eventLog # NOTE THE APPENDER NAME
currentLogFilename: something
threshold: ALL
archive: true
archivedLogFilenamePattern: something
archivedFileCount: 31
timeZone: UTC
logFormat: "myFormat"
- type: hostnameFile
currentLogFilename: something
threshold: ERROR
archive: true
archivedLogFilenamePattern: something
archivedFileCount: 31
timeZone: UTC
logFormat: "myFormat"
正如您所看到的，我将事件附加器映射到事件记录器。这样，我的所有事件最终都保存在文件 A 中，而其他信息则保存在其他地方。
我希望这有帮助。可能不是最干净的解决方案，但我认为 Dropwizard 目前不允许此功能。
