Unnecessary ERROR level logging

When start a multiple nodes environment, you may find the following error message:
ERROR [org.infinispan.remoting.InboundInvocationHandlerImpl] Defined caches: [StreamingDeviceCache, TOPOLOGY_QAMNAMETOSERVICEGROUPID, TOPOLOGY_DEVICE, VAID, ASSETINFO_LOCAL_CACHE, CONTENT_VOLUME_LOCAL_CACHE, TOPOLOGY_EDGEDEVICEINPUT, TOPOLOGY_SERVICEGROUP, ODRM_SESSION, eigAllocation, PAID, clientIdToSessionId, exclusion, TOPOLOGY_STREAMINGDEVICE, sopAllocation, TOPOLOGY_QAM, qamAllocation, STREAMING_CONTENT, content, SERVICE_GROUP, billing, accountIdToSessionId, TOPOLOGY_EDGEDEVICE, session, POILIST, TOPOLOGY_SOPGROUP, VAID_LOCAL_CACHE]

This is actually not always an ERROR if the current node is not fully started with all caches initialized yet.
It confuses our QA people, maybe we can change it to WARN level.