10 changed files with 42 additions and 247 deletions
--- a/README.md
+++ b/README.md
@ -1,6 +1,6 @@
 # kv_bench (Mace vs RocksDB)

-Quick start for reproducible Mace vs RocksDB comparison. Full guide: [docs/repro.md](./docs/repro.md).
+Quick start for reproducible comparison. Full guide: [docs/repro.md](./docs/repro.md).

 ## 5-Minute Quickstart
 1. Set your storage root (any mount path, not hardcoded to `/nvme`):
@ -13,13 +13,13 @@ mkdir -p "${KV_BENCH_STORAGE_ROOT}"
 2. Initialize Python env once:

 ```bash
-cd "$HOME/kv_bench/scripts"
+cd /home/abby/kv_bench/scripts
 ./init.sh
 source ./bin/activate
-cd "$HOME/kv_bench"
+cd /home/abby/kv_bench
 ```

-3. Run baseline comparison (both engines append to the same CSV):
+3. Run baseline comparison (both engines write to the same CSV):

 ```bash
 rm -rf "${KV_BENCH_STORAGE_ROOT}/basic_mace" "${KV_BENCH_STORAGE_ROOT}/basic_rocks"
@ -29,32 +29,20 @@ mkdir -p "${KV_BENCH_STORAGE_ROOT}/basic_mace" "${KV_BENCH_STORAGE_ROOT}/basic_r
 ./scripts/rocksdb.sh "${KV_BENCH_STORAGE_ROOT}/basic_rocks" ./scripts/benchmark_results.csv
 ```

-4. Plot results:
+4. View and plot results:

 ```bash
 ./scripts/bin/python ./scripts/plot.py ./scripts/benchmark_results.csv ./scripts
 ```

-5. Print a direct comparison table from the CSV:
-
-```bash
-./scripts/bin/python ./scripts/compare_baseline.py ./scripts/benchmark_results.csv
-```
-
-## What Is Compared
- Comparison unit: rows with identical `workload_id`, `threads`, `key_size`, `value_size`, `durability_mode`, `read_path`
- Throughput metric: workload-level `ops_per_sec` (higher is better)
-  - `W1/W2/W3/W4`: mixed read+update throughput
-  - `W5`: mixed read+update+scan throughput
-  - `W6`: scan throughput (counted by scan requests, not scanned key count)
- Tail latency metric: workload-level `p99_us` (lower is better)
-  - This is the mixed p99 of all operations executed in that workload row, not per-op-type p99
-  - `W1/W2/W3/W4`: mixed read+update p99
-  - `W5`: mixed read+update+scan p99
-  - `W6`: scan p99
- Reliability gate: if `error_ops > 0`, debug that case before drawing conclusions
-
-Raw CSV path: `./scripts/benchmark_results.csv`
+## Fast Result Reading
+- Raw input CSV: `./scripts/benchmark_results.csv`
+- Key columns:
+  - `engine` (`mace` / `rocksdb`)
+  - `workload_id` (`W1..W6`)
+  - `ops_per_sec` (higher is better)
+  - `p99_us` (lower is better)
+  - `error_ops` (must be 0 before drawing conclusions)

 ## Phase Reports
 - Phase 1 (stability CV):
--- a/docs/repro.md
+++ b/docs/repro.md
@ -22,8 +22,8 @@ For the local profile, assume approximately:
 Before running, set paths and verify filesystem type/capacity:

 ```bash
-export KV_BENCH_STORAGE_ROOT="$HOME/kv_bench/target/repro_storage"
-export KV_BENCH_RESULT_ROOT="$HOME/kv_bench/target/repro_results"
+export KV_BENCH_STORAGE_ROOT=/home/abby/kv_bench/target/repro_storage
+export KV_BENCH_RESULT_ROOT=/home/abby/kv_bench/target/repro_results
 mkdir -p "${KV_BENCH_STORAGE_ROOT}" "${KV_BENCH_RESULT_ROOT}"

 df -hT "${KV_BENCH_STORAGE_ROOT}" "${KV_BENCH_RESULT_ROOT}"
@ -36,10 +36,10 @@ Requirements:

 ## 3. Initialization
 ```bash
-cd "$HOME/kv_bench/scripts"
+cd /home/abby/kv_bench/scripts
 ./init.sh
 source ./bin/activate
-cd "$HOME/kv_bench"
+cd /home/abby/kv_bench
 ```

 ## 4. Quick Baseline (W1~W6)
@ -68,17 +68,6 @@ Generate plots:
 ```

 ## 5. Phase Reproduction
-Default thread points now follow host CPU count:
- If CPU count is a power of two: `1 2 4 ... N` (e.g., `4 -> 1 2 4`, `8 -> 1 2 4 8`, `16 -> 1 2 4 8 16`)
- Otherwise: odd-number progression up to near-full CPU usage (e.g., `12 -> 1 3 5 7 9 11`)
-
-You can still override via `PHASE1_THREADS` / `PHASE2_THREADS_TIER_M` / `PHASE3_THREADS`.
-
-Default KV profiles in scripts:
- `P1`: `16/128`
- `P2`: `32/1024`
- `P3`: `32/4096`
- `P4`: `32/16384` (large-value group)

 ### 5.1 Phase 1
 ```bash
@ -88,6 +77,7 @@ rm -f "${KV_BENCH_RESULT_ROOT}/phase1_results.csv"

 WARMUP_SECS=10 MEASURE_SECS=20 REPEATS=2 \
 PHASE1_WORKLOADS="W1 W3 W6" \
+PHASE1_THREADS="1 6" \
 PHASE1_PROFILES="P2" \
 PHASE1_PREFILL_TIER_S_P2=200000 \
 ./scripts/phase1.sh "${KV_BENCH_STORAGE_ROOT}/phase1" "${KV_BENCH_RESULT_ROOT}/phase1_results.csv"
@ -101,6 +91,7 @@ rm -f "${KV_BENCH_RESULT_ROOT}/phase2_results.csv"

 WARMUP_SECS=10 MEASURE_SECS=20 REPEATS=2 \
 PHASE2_WORKLOADS_TIER_M="W1 W3 W6" \
+PHASE2_THREADS_TIER_M="1 6" \
 PHASE2_PROFILES="P2" \
 PHASE2_PREFILL_TIER_M_P2=500000 \
 RUN_TIER_L_REPRESENTATIVE=0 \
@ -122,6 +113,7 @@ rm -f "${KV_BENCH_RESULT_ROOT}/phase3_results.csv"

 WARMUP_SECS=10 MEASURE_SECS=20 REPEATS=2 \
 PHASE3_WORKLOADS="W1 W3" \
+PHASE3_THREADS="1 6" \
 PHASE3_DURABILITIES="relaxed durable" \
 PHASE3_KEY_SIZE=32 PHASE3_VALUE_SIZE=1024 PHASE3_PREFILL_KEYS=500000 \
 ./scripts/phase3.sh "${KV_BENCH_STORAGE_ROOT}/phase3" "${KV_BENCH_RESULT_ROOT}/phase3_results.csv"
--- a/maritx.md
+++ b/maritx.md
@ -0,0 +1,9 @@
+| 组合类型         | Key 大小          | Value 大小                | 典型场景                     | 测试目标                             |
+|------------------|-------------------|---------------------------|------------------------------|--------------------------------------|
+| 小 Key + 小 Value | 10-100 字节       | 100 字节 - 1 KB           | 缓存用户信息、配置、热点数据 | 高吞吐量（QPS）、低延迟、内存效率   |
+| 小 Key + 大 Value | 10-100 字节       | 1 KB - 数 MB              | 存储文档、图片缩略图、日志   | 网络带宽利用率、持久化性能、压缩效率 |
+| 大 Key + 小 Value | 1 KB - 10 KB      | 100 字节 - 1 KB           | 分布式锁、复杂命名空间       | Key 哈希性能、分片均衡性、元数据内存占用 |
+| 大 Key + 大 Value | 1 KB - 10 KB      | 数 MB - 100 MB            | 极端负载测试、BLOB 存储      | 系统极限性能、I/O 吞吐、GC 压力     |
+| 可变大小组合      | 10 字节 - 1 KB    | 100 字节 - 10 MB          | 混合业务负载（如电商系统）   | 稳定性、缓存命中率、碎片化影响      |
+| 超小 Value        | 10-50 字节        | 1-8 字节                  | 分布式锁、计数器（如 Redis 原子操作） | 高并发竞争性能、CAS 效率            |
+| 超大 Value        | 10-100 字节       | 接近系统上限（如 512 MB） | 单 Value 极限测试            | 单线程阻塞风险、序列化/反序列化瓶颈 |
--- a/scripts/compare_baseline.py
+++ b/scripts/compare_baseline.py
@ -1,90 +0,0 @@
-#!/usr/bin/env python3
-
-import argparse
-import sys
-import pandas as pd
-
-
-def main() -> int:
-    parser = argparse.ArgumentParser(
-        description="Compare mace vs rocksdb from benchmark_results.csv"
-    )
-    parser.add_argument(
-        "csv_path",
-        nargs="?",
-        default="./scripts/benchmark_results.csv",
-        help="Path to benchmark CSV (default: ./scripts/benchmark_results.csv)",
-    )
-    args = parser.parse_args()
-
-    df = pd.read_csv(args.csv_path)
-
-    required = {
-        "engine",
-        "workload_id",
-        "threads",
-        "key_size",
-        "value_size",
-        "durability_mode",
-        "read_path",
-        "ops_per_sec",
-        "p99_us",
-        "error_ops",
-    }
-    missing = required - set(df.columns)
-    if missing:
-        raise ValueError(f"Missing columns in csv: {sorted(missing)}")
-
-    keys = [
-        "workload_id",
-        "threads",
-        "key_size",
-        "value_size",
-        "durability_mode",
-        "read_path",
-    ]
-
-    ok = df[df["error_ops"] == 0].copy()
-    if ok.empty:
-        print("No rows with error_ops == 0, cannot compare.")
-        return 0
-
-    agg = ok.groupby(keys + ["engine"], as_index=False).agg(
-        ops_per_sec=("ops_per_sec", "median"),
-        p99_us=("p99_us", "median"),
-    )
-
-    piv = agg.pivot_table(
-        index=keys,
-        columns="engine",
-        values=["ops_per_sec", "p99_us"],
-        aggfunc="first",
-    )
-    piv.columns = [f"{metric}_{engine}" for metric, engine in piv.columns]
-    out = piv.reset_index()
-
-    for col in [
-        "ops_per_sec_mace",
-        "ops_per_sec_rocksdb",
-        "p99_us_mace",
-        "p99_us_rocksdb",
-    ]:
-        if col not in out.columns:
-            out[col] = pd.NA
-
-    out["qps_ratio_mace_over_rocksdb"] = (
-        out["ops_per_sec_mace"] / out["ops_per_sec_rocksdb"]
-    )
-    out["p99_ratio_mace_over_rocksdb"] = out["p99_us_mace"] / out["p99_us_rocksdb"]
-    out = out.sort_values(keys)
-
-    print(out.to_string(index=False))
-    print("\nInterpretation:")
-    print("- qps_ratio_mace_over_rocksdb > 1: mace has higher throughput")
-    print("- p99_ratio_mace_over_rocksdb < 1: mace has lower p99 latency")
-
-    return 0
-
-
-if __name__ == "__main__":
-    sys.exit(main())
--- a/scripts/mace.sh
+++ b/scripts/mace.sh
@ -9,8 +9,6 @@ fi

 script_dir="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
 root_dir="$(cd -- "${script_dir}/.." && pwd)"
-# shellcheck source=./thread_points.sh
-. "${script_dir}/thread_points.sh"

 # The runner creates per-case unique paths under this root; each path must not exist.
 db_root="$1"
@ -27,16 +25,9 @@ mkdir -p "$(dirname -- "${result_file}")"
 cargo build --release --manifest-path "${root_dir}/Cargo.toml"

 workloads=(W1 W2 W3 W4 W5 W6)
-mace_threads_raw="${MACE_THREADS:-$(default_thread_points)}"
-IFS=' ' read -r -a threads <<< "${mace_threads_raw}"
-if [ "${#threads[@]}" -eq 0 ]; then
-    printf "mace threads list must not be empty\n" >&2
-    exit 1
-fi
+threads=(1 6 12)
 profiles=(
-  "16 128"
  "32 1024"
-  "32 4096"
  "32 16384"
 )

--- a/scripts/phase1.sh
+++ b/scripts/phase1.sh
@ -9,8 +9,6 @@ fi

 script_dir="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
 root_dir="$(cd -- "${script_dir}/.." && pwd)"
-# shellcheck source=./thread_points.sh
-. "${script_dir}/thread_points.sh"
 python_bin="${PYTHON_BIN:-${root_dir}/scripts/bin/python}"
 if [ ! -x "${python_bin}" ]; then
    python_bin="${PYTHON:-python3}"
@ -25,15 +23,10 @@ repeats="${REPEATS:-3}"
 read_path="${READ_PATH:-snapshot}"

 phase1_workloads_raw="${PHASE1_WORKLOADS:-W1 W3 W6}"
-phase1_threads_raw="${PHASE1_THREADS:-}"
-if [ -z "${phase1_threads_raw}" ]; then
-    phase1_threads_raw="$(default_thread_points)"
-fi
-phase1_profiles_raw="${PHASE1_PROFILES:-P1 P2 P3 P4}"
-phase1_prefill_tier_s_p1="${PHASE1_PREFILL_TIER_S_P1:-44739242}"
+phase1_threads_raw="${PHASE1_THREADS:-1 12}"
+phase1_profiles_raw="${PHASE1_PROFILES:-P2 P3}"
 phase1_prefill_tier_s_p2="${PHASE1_PREFILL_TIER_S_P2:-6100805}"
-phase1_prefill_tier_s_p3="${PHASE1_PREFILL_TIER_S_P3:-1560671}"
-phase1_prefill_tier_s_p4="${PHASE1_PREFILL_TIER_S_P4:-392449}"
+phase1_prefill_tier_s_p3="${PHASE1_PREFILL_TIER_S_P3:-392449}"

 mkdir -p "${db_root}"
 mkdir -p "$(dirname -- "${result_file}")"
@ -53,30 +46,24 @@ fi

 profile_key() {
    case "$1" in
-        P1) echo 16 ;;
        P2) echo 32 ;;
        P3) echo 32 ;;
-        P4) echo 32 ;;
        *) printf "unknown profile: %s\n" "$1" >&2; exit 1 ;;
    esac
 }

 profile_val() {
    case "$1" in
-        P1) echo 128 ;;
        P2) echo 1024 ;;
-        P3) echo 4096 ;;
-        P4) echo 16384 ;;
+        P3) echo 16384 ;;
        *) printf "unknown profile: %s\n" "$1" >&2; exit 1 ;;
    esac
 }

 profile_prefill_tier_s() {
    case "$1" in
-        P1) echo "${phase1_prefill_tier_s_p1}" ;;
        P2) echo "${phase1_prefill_tier_s_p2}" ;;
        P3) echo "${phase1_prefill_tier_s_p3}" ;;
-        P4) echo "${phase1_prefill_tier_s_p4}" ;;
        *) printf "unknown profile: %s\n" "$1" >&2; exit 1 ;;
    esac
 }
--- a/scripts/phase2.sh
+++ b/scripts/phase2.sh
@ -9,8 +9,6 @@ fi

 script_dir="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
 root_dir="$(cd -- "${script_dir}/.." && pwd)"
-# shellcheck source=./thread_points.sh
-. "${script_dir}/thread_points.sh"
 python_bin="${PYTHON_BIN:-${root_dir}/scripts/bin/python}"
 if [ ! -x "${python_bin}" ]; then
    python_bin="${PYTHON:-python3}"
@ -28,23 +26,13 @@ tier_l_repeats="${TIER_L_REPEATS:-1}"

 phase2_workloads_tier_m_raw="${PHASE2_WORKLOADS_TIER_M:-W1 W2 W3 W4 W6}"
 phase2_workloads_tier_l_rep_raw="${PHASE2_WORKLOADS_TIER_L_REP:-W1 W3 W6}"
-phase2_threads_tier_m_raw="${PHASE2_THREADS_TIER_M:-}"
-if [ -z "${phase2_threads_tier_m_raw}" ]; then
-    phase2_threads_tier_m_raw="$(default_thread_points)"
-fi
-phase2_threads_tier_l_rep_raw="${PHASE2_THREADS_TIER_L_REP:-}"
-if [ -z "${phase2_threads_tier_l_rep_raw}" ]; then
-    phase2_threads_tier_l_rep_raw="$(default_thread_points)"
-fi
-phase2_profiles_raw="${PHASE2_PROFILES:-P1 P2 P3 P4}"
-phase2_prefill_tier_m_p1="${PHASE2_PREFILL_TIER_M_P1:-134217728}"
+phase2_threads_tier_m_raw="${PHASE2_THREADS_TIER_M:-1 6 12}"
+phase2_threads_tier_l_rep_raw="${PHASE2_THREADS_TIER_L_REP:-1 12}"
+phase2_profiles_raw="${PHASE2_PROFILES:-P2 P3}"
 phase2_prefill_tier_m_p2="${PHASE2_PREFILL_TIER_M_P2:-18302417}"
-phase2_prefill_tier_m_p3="${PHASE2_PREFILL_TIER_M_P3:-4682013}"
-phase2_prefill_tier_m_p4="${PHASE2_PREFILL_TIER_M_P4:-1177348}"
-phase2_prefill_tier_l_p1="${PHASE2_PREFILL_TIER_L_P1:-208783132}"
+phase2_prefill_tier_m_p3="${PHASE2_PREFILL_TIER_M_P3:-1177348}"
 phase2_prefill_tier_l_p2="${PHASE2_PREFILL_TIER_L_P2:-28470427}"
-phase2_prefill_tier_l_p3="${PHASE2_PREFILL_TIER_L_P3:-7283132}"
-phase2_prefill_tier_l_p4="${PHASE2_PREFILL_TIER_L_P4:-1831430}"
+phase2_prefill_tier_l_p3="${PHASE2_PREFILL_TIER_L_P3:-1831430}"

 mkdir -p "${db_root}"
 mkdir -p "$(dirname -- "${result_file}")"
@ -71,20 +59,16 @@ fi

 profile_key() {
    case "$1" in
-        P1) echo 16 ;;
        P2) echo 32 ;;
        P3) echo 32 ;;
-        P4) echo 32 ;;
        *) printf "unknown profile: %s\n" "$1" >&2; exit 1 ;;
    esac
 }

 profile_val() {
    case "$1" in
-        P1) echo 128 ;;
        P2) echo 1024 ;;
-        P3) echo 4096 ;;
-        P4) echo 16384 ;;
+        P3) echo 16384 ;;
        *) printf "unknown profile: %s\n" "$1" >&2; exit 1 ;;
    esac
 }
@ -94,18 +78,14 @@ prefill_for() {
    local profile="$2"
    if [ "${tier}" = "tier-m" ]; then
        case "${profile}" in
-            P1) echo "${phase2_prefill_tier_m_p1}" ;;
            P2) echo "${phase2_prefill_tier_m_p2}" ;;
            P3) echo "${phase2_prefill_tier_m_p3}" ;;
-            P4) echo "${phase2_prefill_tier_m_p4}" ;;
            *) printf "unknown profile: %s\n" "${profile}" >&2; exit 1 ;;
        esac
    elif [ "${tier}" = "tier-l" ]; then
        case "${profile}" in
-            P1) echo "${phase2_prefill_tier_l_p1}" ;;
            P2) echo "${phase2_prefill_tier_l_p2}" ;;
            P3) echo "${phase2_prefill_tier_l_p3}" ;;
-            P4) echo "${phase2_prefill_tier_l_p4}" ;;
            *) printf "unknown profile: %s\n" "${profile}" >&2; exit 1 ;;
        esac
    else
--- a/scripts/phase3.sh
+++ b/scripts/phase3.sh
@ -9,8 +9,6 @@ fi

 script_dir="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
 root_dir="$(cd -- "${script_dir}/.." && pwd)"
-# shellcheck source=./thread_points.sh
-. "${script_dir}/thread_points.sh"
 python_bin="${PYTHON_BIN:-${root_dir}/scripts/bin/python}"
 if [ ! -x "${python_bin}" ]; then
    python_bin="${PYTHON:-python3}"
@ -25,10 +23,7 @@ repeats="${REPEATS:-5}"
 read_path="${READ_PATH:-snapshot}"

 phase3_workloads_raw="${PHASE3_WORKLOADS:-W1 W3 W6}"
-phase3_threads_raw="${PHASE3_THREADS:-}"
-if [ -z "${phase3_threads_raw}" ]; then
-    phase3_threads_raw="$(default_thread_points)"
-fi
+phase3_threads_raw="${PHASE3_THREADS:-1 12}"
 phase3_durabilities_raw="${PHASE3_DURABILITIES:-relaxed durable}"
 key_size="${PHASE3_KEY_SIZE:-32}"
 value_size="${PHASE3_VALUE_SIZE:-1024}"
--- a/scripts/rocksdb.sh
+++ b/scripts/rocksdb.sh
@ -10,8 +10,6 @@ fi
 script_dir="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
 root_dir="$(cd -- "${script_dir}/.." && pwd)"
 rocksdb_dir="${root_dir}/rocksdb"
-# shellcheck source=./thread_points.sh
-. "${script_dir}/thread_points.sh"

 db_root="$1"
 result_file="${2:-${script_dir}/benchmark_results.csv}"
@ -28,16 +26,9 @@ mkdir -p "$(dirname -- "${result_file}")"
 (cd "${rocksdb_dir}" && cmake --build --preset release)

 workloads=(W1 W2 W3 W4 W5 W6)
-rocksdb_threads_raw="${ROCKSDB_THREADS:-$(default_thread_points)}"
-IFS=' ' read -r -a threads <<< "${rocksdb_threads_raw}"
-if [ "${#threads[@]}" -eq 0 ]; then
-    printf "rocksdb threads list must not be empty\n" >&2
-    exit 1
-fi
+threads=(1 6 12)
 profiles=(
-  "16 128"
  "32 1024"
-  "32 4096"
  "32 16384"
 )

--- a/scripts/thread_points.sh
+++ b/scripts/thread_points.sh
@ -1,48 +0,0 @@
-#!/usr/bin/env bash
-
-# Shared helpers for choosing default thread points from host CPU count.
-
-detect_logical_cpus() {
-    local detected
-
-    if detected="$(nproc 2>/dev/null)"; then
-        :
-    elif detected="$(getconf _NPROCESSORS_ONLN 2>/dev/null)"; then
-        :
-    else
-        detected=1
-    fi
-
-    if [[ ! "${detected}" =~ ^[0-9]+$ ]] || [ "${detected}" -lt 1 ]; then
-        detected=1
-    fi
-
-    printf "%s\n" "${detected}"
-}
-
-is_power_of_two() {
-    local n="$1"
-    [ "${n}" -gt 0 ] && [ $((n & (n - 1))) -eq 0 ]
-}
-
-default_thread_points() {
-    local cpu_count="${1:-$(detect_logical_cpus)}"
-    local points=()
-    local t
-
-    if is_power_of_two "${cpu_count}"; then
-        t=1
-        while [ "${t}" -le "${cpu_count}" ]; do
-            points+=("${t}")
-            t=$((t * 2))
-        done
-    else
-        t=1
-        while [ "${t}" -le "${cpu_count}" ]; do
-            points+=("${t}")
-            t=$((t + 2))
-        done
-    fi
-
-    printf "%s\n" "${points[*]}"
-}