Guarantee usefulness of `db-analyser` ledger ops result.

[dnadales]

We need to establish that the benchmark-ledger-ops analysis of db-analyser are useful in spotting potential performance regressions or optimizations. For that, we need its results to be reproducible and meaningful.

We need to run the benchmark-ledger-ops analysis with the same build, data, machine, and workload and corroborate if the runs have small enough variance among each other. Additionally, we need to make sure the input data is suitable to clearly reveal any measurable changes wrt. ledger ops.

Requires https://github.com/input-output-hk/ouroboros-consensus/issues/230

Acceptance criteria

• A set of input data is identified that:
  • Reveals known performance problems.
  • Exhibits low variance among runs.

Subtasks

• [ ] Validate that benchmark-ledger-ops exhibits the same mutator times as a running node. (eg Issue https://github.com/IntersectMBO/ouroboros-consensus/issues/1272)
• [ ] Ensure we use the same RTS flags as cardano-node only for benchmark-specific invocations (which can be done in tools such as beacon or invocation scripts).
• [ ] Calibrate and generate a useful set of input data (P&T)
• [ ] Feed this data to the variance analysis of P&T.

[mgmeier]

To enable establishing a posible correlation of this data with that gathered from system level benchmarks, a good starting point for input data are those exact ChainDBs that are produced as a result of running our benchmarking profiles.