Benchmark RLI flow with a large table to improve performance

[hudi-bot]

High level context

Benchmark RLI for tables on an existing table with large number of record keys (~100B). 
Incrementally ingest about 10GB of data, MoR table, partitioned with ~500 partitions. 

Use Hfile size of 2GB.

• Ensure the bootstrap of RLI works as expected.
• Measure the read and write latencies for the RLI index
• Find and measure all bottlenecks
• Report any issue with the core indexing or RLI or MDT DAGs.

JIRA info

• Link: https://issues.apache.org/jira/browse/HUDI-9283
• Type: Improvement
• Fix version(s):
  • 1.1.0