[Question] Why is HugeGraph's memory usage steadily increasing?

### Problem Type (问题类型)

None

### Before submit

- [x] 我已经确认现有的 [Issues](https://github.com/apache/hugegraph/issues) 与 [FAQ](https://hugegraph.apache.org/docs/guides/faq/) 中没有相同 / 重复问题 (I have confirmed and searched that there are no similar problems in the historical issue and documents)

### Environment (环境信息)

- Server Version: 1.5.0 (Apache Release Version)
- Backend: RocksDB

### Your Question (问题描述)

Why does memory usage increase as HugeGraph is used in RocksDB mode? During this process, we created new graphs, inserted vertices and edges, and executed various queries. When memory usage approaches its limit, the HugeGraph service becomes unavailable or even crashes; how can this be prevented?

### Vertex/Edge example (问题点 / 边数据举例)

```javascript

```

### Schema [VertexLabel, EdgeLabel, IndexLabel] (元数据结构)

```javascript

```

No source file, test, or entry point is identified in the issue. Begin by investigating HugeGraph 1.5.0 with the RocksDB backend while creating graphs, inserting vertices and edges, and running queries; done means identifying the cause of the steadily increasing memory usage and documenting a reliable prevention or mitigation.

Written by the indexing model from the issue text.