We have logback configured with this configuration:
<br><br>   &lt;appender name=&quot;FILE&quot; class=&quot;ch.qos.logback.core.rolling.RollingFileAppender&quot;&gt;
<br>      &lt;Append&gt;true&lt;/Append&gt;
<br>      &lt;file&gt;logs/process.log&lt;/file&gt;
<br>      &lt;rollingPolicy class=&quot;ch.qos.logback.core.rolling.TimeBasedRollingPolicy&quot;&gt;
<br>         &lt;FileNamePattern&gt;logs/process-%d{yyyy-MM-dd_HH}.%i.log&lt;/FileNamePattern&gt;
<br>         &lt;TimeBasedFileNamingAndTriggeringPolicy class=&quot;ch.qos.logback.core.rolling.SizeAndTimeBasedFNATP&quot;&gt;
<br>            &lt;MaxFileSize&gt;1MB&lt;/MaxFileSize&gt;
<br>         &lt;/TimeBasedFileNamingAndTriggeringPolicy&gt;
<br>      &lt;/rollingPolicy&gt;
<br>       &lt;layout class=&quot;ch.qos.logback.classic.PatternLayout&quot;&gt;
<br>            &lt;Pattern&gt;%date|%level|%logger{10}|%msg|%throwable{5}%n&lt;/Pattern&gt;
<br>      &lt;/layout&gt;
<br>   &lt;/appender&gt;
<br><br><br>The thing is that in our production environment (HP-UX) and
under heavy load, logging was working fine for a time, and suddenly
stop working.
<br><br>From the status we see this:
<br>INFO in c.q.l.co.rolling.helper.RenameUtil - Renaming file
[/gp/log/process.log] to [/gp/log/process.2009-12-16_11.22.log], ERROR
in ch.qos.logback.core.rolling.RollingFileAppender[FILE] - IO failure
in appender java.io.InterruptedIOException: , WARN in
ch.qos.logback.core.rolling.RollingFileAppender[FILE] - Attempted to
append to non started appender [FILE]., WARN in
ch.qos.logback.core.rolling.RollingFileAppender[FILE] - Attempted to
append to non started appender [FILE]., WARN in
ch.qos.logback.core.rolling.RollingFileAppender[FILE] - Attempted to
append to non started appender [FILE]., INFO in
c.q.l.co.rolling.helper.RenameUtil - Renaming file
[/gp/log/application.log] to [/gp/log/application.2009-12-16_11.log]]