Improved organization of pbench output directory and detection of duplicate stages in nested stages

[minhancao]

Improved organization of pbench output directory to navigate easier for streams runs.

Example 3 streams run:

pbench command ran:

../pbench_mcao run -n testing_output_dir_change_%t -s http://127.0.0.1:8081 -o /root/oss_pbench/pbench/benchmarks/benchmark_base native_oss.json save_output.json save_json.json
 tpch/sf1_tpchstandard.json tpch/3streams.json > pbench_output2.txt 2>&
1

3streams.json:

{
    "description": "Launch 3 clients, each running from tpch.",
    "next": [
      "tpch_stream1.json",
      "tpch_stream2.json",
      "tpch_stream3.json"
    ],
    "query_files": [
      "queries_tpchstandard/query_01.sql"
    ]
}

tpch_stream1.json:

{
    "description": "Sequentially run 3 TPCH queries.",
    "next": [
      "tpch_stream1_child.json"
    ],
    "query_files": [
      "queries_tpchstandard/query_01.sql"
    ],
    "expected_row_counts": {
      "tpch_sf1000_": [
        4,
        100,
        10
      ]
    }
}

tpch_stream2.json, tpch_stream3.json:

{
    "description": "Sequentially run 3 TPCH queries.",
    "query_files": [
      "queries_tpchstandard/query_01.sql"
    ],
    "expected_row_counts": {
      "tpch_sf1000_": [
        4,
        100,
        10
      ]
    }
}

tpch_stream1_child.json:

{
    "description": "Sequentially run 3 TPCH queries.",
    "query_files": [
      "queries_tpchstandard/query_01.sql"
    ],
    "expected_row_counts": {
      "tpch_sf1000_": [
        4,
        100,
        10
      ]
    }
}

[minhancao]

@ethanyzhang I have changed the implementation to the way that we discussed and also improved on the checkStageLinks() to detect for duplicate stages in nested stages. Please review when you can, thanks!

[ethanyzhang]

If it is nil, you probably want to error out right?

[ethanyzhang]

Why? when the next stage is run, it will call this again right?