We seem to be losing quite a few log records in a clustered web application writing to an NFS-mounted file with the configuration shown below.

We are running logback-1.0.0. Any suggestions?

      <appender name="file"

            class="ch.qos.logback.core.rolling.RollingFileAppender">

            <append>true</append>

            <prudend>true</prudend>

            <encoder>

                  <pattern>%d ${HOSTNAME} [%t] %-5p %c - {%marker} %m%n</pattern>

            </encoder>

            <rollingPolicy class="ch.qos.logback.core.rolling.TimeBasedRollingPolicy">

                  <maxHistory>10</maxHistory>

                  <fileNamePattern>/xxx/log-%d.log</fileNamePattern>

            </rollingPolicy>

      </appender>