[Improvement] TraceId Implementation for Audit Logs in Gravitino

### What would you like to be improved?

Currently, for Gravitino's audit logs, there is no concept of traceId that runs through the entire user request, processing, and response return. Especially since we may later connect Gravitino to various engines, and in our actual production practice, the logs and requests related to the engine will include a traceId, which may contain specific information corresponding to an ETL task's appId, user, etc. As a link in the call chain of a task, Gravitino also hopes to include this traceId in the entire call, processing, and response chain, so that a full link tracking of a task can be carried out, and it is known which task a failure specifically corresponds to. In particular, it is very necessary to include this traceId in all logs related to a request.

### How should we improve?

_No response_

No files, tests, or entry points are identified. Start by locating Gravitino's request, processing, response, and audit-log paths, then determine how a traceId should flow through them and what evidence would show that all request-related logs can be correlated.

Written by the indexing model from the issue text.