pferraro Thanks!

The default settings with WildFly is used without much changes. There is no other cache regions added.

<cache-container name="hibernate" default-cache="local-query" module="org.hibernate">

                    <transport lock-timeout="120000"/>

                    <local-cache name="local-query">

                        <locking isolation="READ_COMMITTED"/>

                        <transaction mode="NONE" locking="OPTIMISTIC"/>

                        <eviction strategy="LRU" max-entries="10000"/>

                        <expiration max-idle="100000"/>

                    </local-cache>

                    <invalidation-cache name="entity" mode="SYNC">

                        <locking isolation="READ_COMMITTED"/>

                        <transaction mode="NON_XA" locking="OPTIMISTIC"/>

                        <eviction strategy="LRU" max-entries="10000"/>

                        <expiration max-idle="100000"/>

                    </invalidation-cache>

                    <replicated-cache name="timestamps" mode="ASYNC">

                        <locking isolation="READ_COMMITTED"/>

                        <transaction mode="NONE" locking="OPTIMISTIC"/>

                        <eviction strategy="NONE"/>

                    </replicated-cache>

                </cache-container>

Currently in 2 hours he same lag occurred when tested. Which settings need to be adjusted to control cache expiration?

Here is the L2C metrics printed during the operation:

[Server:ltchie-node-00] 20:24:59,495 INFO  [org.hibernate.engine.internal.StatisticalLoggingSessionEventListener] (default task-8) Session Metrics {

[Server:ltchie-node-00]     15330440 nanoseconds spent acquiring 48 JDBC connections;

[Server:ltchie-node-00]     4266808 nanoseconds spent releasing 48 JDBC connections;

[Server:ltchie-node-00]     16230818 nanoseconds spent preparing 48 JDBC statements;

[Server:ltchie-node-00]     47383156 nanoseconds spent executing 48 JDBC statements;

[Server:ltchie-node-00]     0 nanoseconds spent executing 0 JDBC batches;

[Server:ltchie-node-00]     96615 nanoseconds spent performing 1 L2C puts;

[Server:ltchie-node-00]     1512024 nanoseconds spent performing 28 L2C hits;

[Server:ltchie-node-00]     1162267 nanoseconds spent performing 28 L2C misses;

[Server:ltchie-node-00]     913458 nanoseconds spent executing 1 flushes (flushing a total of 52 entities and 228 collections);

[Server:ltchie-node-00]     15298375 nanoseconds spent executing 31 partial-flushes (flushing a total of 712 entities and 712 collections)

[Server:ltchie-node-00] }

No, I think it should be higher. So the existence of an entry in local hibernate cache depends on  max-entries & max-idle.