Regular benchmark for typical scenarios

As our products now tested and used by more and more users, we should have a benchmark suite that provide regular report for typical scenarios, monitoring the CPU, Memory, Result and timing metrics, in order to not give user **surprise**.

This could be a large project that including all components, but we could start in most urgent scenarios.

This doesn't need to be extensive but should capture the most used and typical test suites.

- [ ] GAE
- [ ] Builtin algorithms
- [ ] Networkx algorithms
- [ ] GIE
- [ ] Vineyard storage
- [ ] Groot storage
- [ ] GLE

...others

Start by reviewing the existing GAE, GIE, and GLE test entry points, including Builtin algorithms, Networkx algorithms, Vineyard storage, and Groot storage. Determine which urgent and typical scenarios can be measured first. Done means a regular benchmark report covers CPU, memory, result, and timing metrics for the selected scenarios.

Written by the indexing model from the issue text.