AuditLog logfile retention problem

[mmisztal1980 (Maciek Misztal)]

Hello,

We are running Elastic Stack v6.5.3 and have configured the auditlog to emit events to the logfile sink.
We'd like to ensure that the logfiles are kept with a 3-day retention. Therefore we have attempted to apply the following log4j configuration:

appender.audit_rolling.type = RollingFile
appender.audit_rolling.name = audit_rolling
appender.audit_rolling.fileName = ${sys:es.logs.base_path}${sys:file.separator}${sys:es.logs.cluster_name}_audit.log
appender.audit_rolling.layout.type = PatternLayout
appender.audit_rolling.layout.pattern = {\
                "@timestamp-auditlogevent":"%d{ISO8601}"\
                %varsNotEmpty{, "node.name":"%enc{%map{node.name}}{JSON}"}\
                %varsNotEmpty{, "node.id":"%enc{%map{node.id}}{JSON}"}\
                %varsNotEmpty{, "host.name":"%enc{%map{host.name}}{JSON}"}\
                %varsNotEmpty{, "host.ip":"%enc{%map{host.ip}}{JSON}"}\
                %varsNotEmpty{, "event.type":"%enc{%map{event.type}}{JSON}"}\
                %varsNotEmpty{, "event.action":"%enc{%map{event.action}}{JSON}"}\
                %varsNotEmpty{, "user.name":"%enc{%map{user.name}}{JSON}"}\
                %varsNotEmpty{, "user.run_by.name":"%enc{%map{user.run_by.name}}{JSON}"}\
                %varsNotEmpty{, "user.run_as.name":"%enc{%map{user.run_as.name}}{JSON}"}\
                %varsNotEmpty{, "user.realm":"%enc{%map{user.realm}}{JSON}"}\
                %varsNotEmpty{, "user.run_by.realm":"%enc{%map{user.run_by.realm}}{JSON}"}\
                %varsNotEmpty{, "user.run_as.realm":"%enc{%map{user.run_as.realm}}{JSON}"}\
                %varsNotEmpty{, "user.roles":%map{user.roles}}\
                %varsNotEmpty{, "origin.type":"%enc{%map{origin.type}}{JSON}"}\
                %varsNotEmpty{, "origin.address":"%enc{%map{origin.address}}{JSON}"}\
                %varsNotEmpty{, "realm":"%enc{%map{realm}}{JSON}"}\
                %varsNotEmpty{, "url.path":"%enc{%map{url.path}}{JSON}"}\
                %varsNotEmpty{, "url.query":"%enc{%map{url.query}}{JSON}"}\
                %varsNotEmpty{, "request.body":"%enc{%map{request.body}}{JSON}"}\
                %varsNotEmpty{, "action":"%enc{%map{action}}{JSON}"}\
                %varsNotEmpty{, "request.name":"%enc{%map{request.name}}{JSON}"}\
                %varsNotEmpty{, "indices":%map{indices}}\
                %varsNotEmpty{, "opaque_id":"%enc{%map{opaque_id}}{JSON}"}\
                %varsNotEmpty{, "transport.profile":"%enc{%map{transport.profile}}{JSON}"}\
                %varsNotEmpty{, "rule":"%enc{%map{rule}}{JSON}"}\
                %varsNotEmpty{, "event.category":"%enc{%map{event.category}}{JSON}"}\
                }%n

appender.audit_rolling.filePattern = ${sys:es.logs.base_path}${sys:file.separator}${sys:es.logs.cluster_name}_audit-%d{yyyy-MM-dd}.log
appender.audit_rolling.policies.type = Policies
appender.audit_rolling.policies.time.type = TimeBasedTriggeringPolicy
appender.audit_rolling.policies.time.interval = 1
appender.audit_rolling.policies.time.modulate = true
appender.audit_rolling.strategy.action=Delete
appender.audit_rolling.strategy.type=DefaultRolloverStrategy
appender.rolling.strategy.action.condition.nested_condition.type=IfLastModified
appender.rolling.strategy.action.condition.nested_condition.age=3D

However we haven't observed the old log files being deleted.
Can you point us to a correct solution?