Could you develop an advanced distributed lock that supports reentrant locks across the jvm within same call chain?
Is it possible to acquire locks based on the trace ID, for example, by providing the trace ID as input, or by providing a trace ID version from the tlog.