Bad Object Array seen in Netty

[Yichun Ding]

Hi,

Our system has a memory usage issue on a Java application on version openjdk 11.0.7. Based on jxray report of the memory dump, I see two chunks of memories were not seen after the application was started initially:

18,576Kb (17.1%): Object[]: 9,216 / 100% of empty 18,576Kb (17.1%)
 ↖io.netty.util.internal.shaded.org.jctools.queues.MpscArrayQueue.buffer
↖io.netty.buffer.PoolThreadCache$SubPageMemoryRegionCache.queue
↖io.netty.buffer.PoolThreadCache$MemoryRegionCache[]
↖io.netty.buffer.PoolThreadCache.tinySubPageHeapCaches
↖j.l.r.Finalizer.referent
↖j.l.r.Finalizer.{next}
↖Java Static java.lang.ref.Finalizer.unfinalized

5,031Kb (4.6%): Object[]: 2,496 / 100% of empty 5,031Kb (4.6%)
 ↖io.netty.util.internal.shaded.org.jctools.queues.MpscArrayQueue.buffer
↖io.netty.buffer.PoolThreadCache$SubPageMemoryRegionCache.queue
↖io.netty.buffer.PoolThreadCache$MemoryRegionCache[]
↖io.netty.buffer.PoolThreadCache.tinySubPageDirectCaches
↖io.netty.buffer.PooledUnsafeDirectByteBuf.cache
↖io.netty.util.Recycler$DefaultHandle.value
↖io.netty.util.Recycler$DefaultHandle[]
↖io.netty.util.Recycler$WeakOrderQueue$Link.elements
↖io.netty.util.Recycler$WeakOrderQueue.tail
↖{j.u.WeakHashMap}.values
↖Object[]
↖io.netty.util.internal.InternalThreadLocalMap.indexedVariables
↖io.netty.util.concurrent.FastThreadLocalThread.threadLocalMap
↖j.l.Thread[]
↖j.l.ThreadGroup.threads
↖org.apache.logging.log4j.core.util.Log4jThreadFactory.group
↖j.u.concurrent.ThreadPoolExecutor.threadFactory
↖Java Static org.apache.logging.log4j.core.jmx.Server.executor

Is this a known issue?

Thanks.