[
https://issues.apache.org/jira/browse/LOG4J2-3185?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17443434#comment-17443434
]
ASF subversion and git services commented on LOG4J2-3185:
---------------------------------------------------------
Commit 843ca800ba60464059a22e9e12edcaf5ecd92d95 in logging-log4j2's branch
refs/heads/release-2.x from Volkan Yazici
[ https://gitbox.apache.org/repos/asf?p=logging-log4j2.git;h=843ca80 ]
LOG4J2-3060 LOG4J2-3185 Fix error-propagation logic in DefaultErrorHandler.
> DefaultErrorHandler can not share values across threads
> -------------------------------------------------------
>
> Key: LOG4J2-3185
> URL: https://issues.apache.org/jira/browse/LOG4J2-3185
> Project: Log4j 2
> Issue Type: Bug
> Components: Appenders, Core
> Affects Versions: 2.14.1
> Reporter: mzbonnt
> Priority: Critical
> Fix For: 2.15.0
>
>
> My application used a logger with AsyncAppender and set the blocking
> attribute to false. The example code as following.
> {code:java}
> // Some example logger usage for catching all throwables and log it to disk
> file
> try {
> // my application code, which may throw an exception
> } catch (Throwable t) {
> logger.error("", t);
> }
> {code}
> If there was too much throwable generated, then the the blocking queue in the
> async appender will be full, and will never enqueue any more. Thus, it will
> execute DefaultErrorHandler#error(String) method.
> {code:java}
> // org.apache.logging.log4j.core.appender.AsyncAppender
> @Override
> public void append(final LogEvent logEvent) {
> if (!isStarted()) {
> throw new IllegalStateException("AsyncAppender " + getName() + "
> is not active");
> }
> final Log4jLogEvent memento = Log4jLogEvent.createMemento(logEvent,
> includeLocation);
> InternalAsyncUtil.makeMessageImmutable(logEvent.getMessage());
> if (!transfer(memento)) {
> if (blocking) {
> if (AbstractLogger.getRecursionDepth() > 1) { // LOG4J2-1518,
> LOG4J2-2031
> // If queue is full AND we are in a recursive call, call
> appender directly to prevent deadlock
> AsyncQueueFullMessageUtil.logWarningToStatusLogger();
> logMessageInCurrentThread(logEvent);
> } else {
> // delegate to the event router (which may discard,
> enqueue and block, or log in current thread)
> final EventRoute route =
> asyncQueueFullPolicy.getRoute(dispatcher.getId(), memento.getLevel());
> route.logMessage(this, memento);
> }
> } else {
> error("Appender " + getName() + " is unable to write primary
> appenders. queue is full");
> logToErrorAppenderIfNecessary(false, memento);
> }
> }
> }
> {code}
> Application usually runs with multi threads, if the AsyncAppender inner queue
> was full, then the DefaultErrorHandler may be executed across threads.
> However, the exceptionCount filed and lastException fields were not volatile
> variables, it may cause their value were not shared across threads.
> {code:java}
> // org.apache.logging.log4j.core.appender.DefaultErrorHandler
> private static final int MAX_EXCEPTIONS = 3;
> private static final long EXCEPTION_INTERVAL =
> TimeUnit.MINUTES.toNanos(5);
> private int exceptionCount = 0;
> private long lastException = System.nanoTime() - EXCEPTION_INTERVAL - 1;
> @Override
> public void error(final String msg) {
> final long current = System.nanoTime();
> if (current - lastException > EXCEPTION_INTERVAL || exceptionCount++
> < MAX_EXCEPTIONS) {
> LOGGER.error(msg);
> }
> lastException = current;
> }
> {code}
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
