In Hadoop ecosystem components, which primarily use the Apache Log4j logging framework, reaching a configured log file size limit triggers a process called log rotation. The standard behavior for a size-based RollingFileAppender is to archive the current active log file (typically by renaming it, e.g., service.log becomes service.log.1) and then create a new, empty file with the original name (service.log). This ensures that logging continues without interruption and historical log data is preserved up to a configured number of backup files, effectively managing disk space usage.

1. Cloudera Data Platform (CDP) Documentation: In the section on "Configuring Log Rotation Policies

" the documentation states: "When a log file reaches a certain size or age

it is rolled over. The current log file is renamed and a new log file is created. You can configure the maximum size of each log file and the number of old log files to retain." This directly supports the process of archiving and creating a new file.

Source: Cloudera

"CDP Private Cloud Base - System Administrator Guide

" Section: "Configuring Log Rotation Policies."

2. Apache Logging Services Documentation (Log4j): The RollingFileAppender is the component responsible for this behavior. Its purpose is to roll over log files when a triggering policy

such as a size limit (SizeBasedTriggeringPolicy)

is met. The default rollover strategy involves creating archived copies of the log file.

Source: Apache Logging Services

"Log4j 2.x Documentation

" Page: "Appenders

" Section: "RollingFileAppender."

3. Apache Hadoop Documentation: The default log4j.properties configuration for Hadoop daemons specifies a RollingFileAppender (org.apache.log4j.RollingFileAppender). It uses parameters like MaxFileSize to trigger rotation and MaxBackupIndex to control how many archived files are kept

confirming the archive-and-recreate mechanism.

Source: Apache Hadoop 3.3.6 Documentation

"Hadoop: Setting up a Single Node Cluster

" Section: "Execution

" referencing the default etc/hadoop/log4j.properties configuration file.