fix(benchmarks): use nullMarkers in csvOptions and add missing request counter schema fields

[yuxin00j]

Summary

1. Change "nullMarker": "N/A" to "nullMarkers": ["N/A", ""] in cloudbuild/benchmarks/benchmarks_schema.json (matching macrobenchmarks_schema.json and subsystembenchmarks_schema.json) so BigQuery treats empty strings from missing columns in historical CSVs as NULL, preserving numeric types (INTEGER / FLOAT).
2. Add missing schema fields introduced in Adds cat file benchmarks #1052 (concurrency, requests_total, requests_per_op, requests_download, requests_object_get, requests_list, requests_other).

Problem

1. When BigQuery scans historical CSV files generated prior to feat(perf): add microbenchmark suite for cat_ranges #1035 under "sourceColumnMatch": "NAME", missing columns evaluate to empty strings (""). With "nullMarker": "N/A", BigQuery did not treat "" as NULL, failing ingestion when parsing "" into INTEGER columns (batch_size, max_gap, num_ranges, total_bytes). Using "nullMarkers": ["N/A", ""] resolves this while preserving INTEGER types.
2. Recent microbenchmark runs after Adds cat file benchmarks #1052 emit request counter metrics (concurrency, requests_total, requests_per_op, requests_download, requests_object_get, requests_list, requests_other) in CSV headers, which caused BigQuery ingestion to fail with Unmatched source column found when scanning newer CSVs.

[gemini-code-assist]

Code Review

This pull request changes the data types of several fields (batch_size, max_gap, num_ranges, and total_bytes) from INTEGER to STRING in the benchmarks schema. The reviewer suggests keeping these fields as INTEGER and instead configuring csvOptions to treat empty strings as NULL, which preserves schema quality and proper numeric types for querying.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.21%. Comparing base (581fd26) to head (45dc6bb).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1062   +/-   ##
=======================================
  Coverage   90.21%   90.21%           
=======================================
  Files          16       16           
  Lines        3749     3749           
=======================================
  Hits         3382     3382           
  Misses        367      367           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.