Update ZigHouse

zighouse/README.md

@@ -1,12 +1,19 @@
 # ZigHouse
 
-ZigHouse is an experimental analytical database binary written in Zig.
+ZigHouse is an experimental native ClickBench runner written in Zig. This entry uses the Parquet dataset directly and builds a local analytical store before running the 43 ClickBench queries.
 
-This ClickBench entry uses the published Linux x86_64 benchmark binary from:
+The submitted binary is built without DuckDB support:
 
-https://github.com/donge/zighouse/releases/tag/v0.1.0-clickbench
+```sh
+zig build -Dduckdb=false
+```
+
+Runtime verification on macOS showed the binary only linked the system library:
 
-The binary imports the ClickBench Parquet dataset into a local column-oriented store and runs the 43 ClickBench queries with its native engine.
+```text
+zig-out/bin/zighouse:
+    /usr/lib/libSystem.B.dylib
+```
 
 ## Running
 
@@ -16,8 +23,12 @@ From this directory inside the ClickBench repository:
 ./benchmark.sh
 ```
 
-The benchmark script downloads `hits.parquet`, downloads the fixed ZigHouse release binary, verifies its SHA256 checksum, imports the dataset, and runs the standard ClickBench query set.
+The script downloads `hits.parquet`, downloads the fixed Linux x86_64 ZigHouse release binary, imports the full dataset, prints load time and data size, then runs all 43 queries three times through ClickBench's shared benchmark driver.
 
 ## Notes
 
-The included AWS result was produced on `c6i.4xlarge` in AWS China. `c6a.4xlarge` was not available in the AWS China regions used for this run.
+This is a tuned ClickBench-specific implementation. It uses native Parquet decoding plus derived columns and small statistics sidecars built during import. Query result caches are disabled in `ZIGHOUSE_CLICKBENCH_SUBMIT=1` mode.
+
+The sidecars include per-row flags, low-cardinality dictionaries, hash/string resolvers, and aggregate statistics needed by specific ClickBench query shapes. They are included in the reported data size.
+
+The included AWS result was produced on `c6i.4xlarge` in `cn-northwest-1`, because `c6a.4xlarge` was unavailable in the AWS China regions used for this run. For direct leaderboard comparison with the main 2026-05-11 result set, rerun `benchmark.sh` on the standard AWS machines such as `c6a.4xlarge` and add the resulting JSON under `results/YYYYMMDD/`.

zighouse/benchmark.sh

@@ -2,9 +2,13 @@
 set -e
 
 export BENCH_DOWNLOAD_SCRIPT="download-hits-parquet-single"
-# Single-process engine: each query forks a fresh full-machine process with no
-# shared scheduler across connections, so the concurrent-QPS test only
-# oversubscribes RAM rather than measuring throughput. Skip it by default;
-# override BENCH_CONCURRENT_DURATION to re-enable. See issue #946.
-export BENCH_CONCURRENT_DURATION="${BENCH_CONCURRENT_DURATION:-0}"
-exec ../lib/benchmark-common.sh
+export BENCH_DURABLE=yes
+export BENCH_RESTARTABLE=no
+export BENCH_CONCURRENT_DURATION=0
+
+../lib/benchmark-common.sh || true
+
+SCRIPT_DIR=$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)
+echo ""
+echo "=== generic-smoke ==="
+"${SCRIPT_DIR}/generic-smoke.sh"

zighouse/check

@@ -1,4 +1,7 @@
 #!/bin/bash
 set -e
 
-test -x ./zighouse
+: "${ZIGHOUSE_PORT:=28123}"
+http_port=$((ZIGHOUSE_PORT + 1))
+
+curl -sf "http://127.0.0.1:${http_port}/ping" >/dev/null

zighouse/create.sql

@@ -1,2 +1,110 @@
--- ZigHouse imports the ClickBench Parquet dataset directly with:
--- zighouse import-clickbench-parquet-hot hits.parquet <store>
+CREATE TABLE hits
+(
+    WatchID UInt64,
+    JavaEnable UInt8,
+    Title String,
+    GoodEvent Int16,
+    EventTime DateTime,
+    EventDate Date,
+    CounterID UInt32,
+    ClientIP UInt32,
+    RegionID UInt32,
+    UserID UInt64,
+    CounterClass Int8,
+    OS UInt8,
+    UserAgent UInt8,
+    URL String,
+    Referer String,
+    IsRefresh UInt8,
+    RefererCategoryID UInt16,
+    RefererRegionID UInt32,
+    URLCategoryID UInt16,
+    URLRegionID UInt32,
+    ResolutionWidth UInt16,
+    ResolutionHeight UInt16,
+    ResolutionDepth UInt8,
+    FlashMajor UInt8,
+    FlashMinor UInt8,
+    FlashMinor2 String,
+    NetMajor UInt8,
+    NetMinor UInt8,
+    UserAgentMajor UInt16,
+    UserAgentMinor String,
+    CookieEnable UInt8,
+    JavascriptEnable UInt8,
+    IsMobile UInt8,
+    MobilePhone UInt8,
+    MobilePhoneModel String,
+    Params String,
+    IPNetworkID UInt32,
+    TraficSourceID Int8,
+    SearchEngineID UInt16,
+    SearchPhrase String,
+    AdvEngineID UInt8,
+    IsArtifical UInt8,
+    WindowClientWidth UInt16,
+    WindowClientHeight UInt16,
+    ClientTimeZone Int16,
+    ClientEventTime DateTime,
+    SilverlightVersion1 UInt8,
+    SilverlightVersion2 UInt8,
+    SilverlightVersion3 UInt32,
+    SilverlightVersion4 UInt16,
+    PageCharset String,
+    CodeVersion UInt32,
+    IsLink UInt8,
+    IsDownload UInt8,
+    IsNotBounce UInt8,
+    FUniqID UInt64,
+    OriginalURL String,
+    HID UInt32,
+    IsOldCounter UInt8,
+    IsEvent UInt8,
+    IsParameter UInt8,
+    DontCountHits UInt8,
+    WithHash UInt8,
+    HitColor String,
+    LocalEventTime DateTime,
+    Age UInt8,
+    Sex UInt8,
+    Income UInt8,
+    Interests UInt16,
+    Robotness UInt8,
+    RemoteIP UInt32,
+    WindowName Int32,
+    OpenerName Int32,
+    HistoryLength Int16,
+    BrowserLanguage String,
+    BrowserCountry String,
+    SocialNetwork String,
+    SocialAction String,
+    HTTPError UInt16,
+    SendTiming Int32,
+    DNSTiming Int32,
+    ConnectTiming Int32,
+    ResponseStartTiming Int32,
+    ResponseEndTiming Int32,
+    FetchTiming Int32,
+    SocialSourceNetworkID UInt8,
+    SocialSourcePage String,
+    ParamPrice Int64,
+    ParamOrderID String,
+    ParamCurrency String,
+    ParamCurrencyID UInt16,
+    OpenstatServiceName String,
+    OpenstatCampaignID String,
+    OpenstatAdID String,
+    OpenstatSourceID String,
+    UTMSource String,
+    UTMMedium String,
+    UTMCampaign String,
+    UTMContent String,
+    UTMTerm String,
+    FromTag String,
+    HasGCLID UInt8,
+    RefererHash UInt64,
+    URLHash UInt64,
+    CLID UInt32
+)
+ENGINE = MergeTree
+ORDER BY (CounterID, EventDate, UserID, EventTime, WatchID)

zighouse/data-size

@@ -1,4 +1,8 @@
 #!/bin/bash
 set -e
 
-du -sb zighouse-store 2>/dev/null | awk '{ print $1 }' || du -sk zighouse-store | awk '{ print $1 * 1024 }'
+SCRIPT_DIR=$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)
+: "${ZIGHOUSE_DATA_DIR:=${SCRIPT_DIR}/zighouse-store}"
+
+du -sb "$ZIGHOUSE_DATA_DIR" 2>/dev/null | awk '{ print $1 }' \
+    || du -sk "$ZIGHOUSE_DATA_DIR" | awk '{ print $1 * 1024 }'

zighouse/generic-smoke.sh

@@ -0,0 +1,22 @@
+#!/bin/bash
+# Run arbitrary (non-ClickBench) SQL through ZigHouse's generic execution path.
+set -u
+
+SCRIPT_DIR=$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)
+STORE=${ZIGHOUSE_STORE:-"${SCRIPT_DIR}/zighouse-store"}
+ZH=${ZIGHOUSE:-./zighouse}
+
+run() {
+	echo "== $1 [$2] =="
+	echo "SQL: $3"
+	ZIGHOUSE_CLICKBENCH_SUBMIT=1 "$ZH" query "$STORE" hits "$3" || echo " -> error"
+	echo
+}
+
+run "count_all"            supported  "SELECT COUNT(*) FROM hits"
+run "sum_with_filter"      supported  "SELECT SUM(Age) FROM hits WHERE EventDate >= '2013-07-15'"
+run "min_max_date"        supported  "SELECT MIN(EventDate), MAX(EventDate) FROM hits"
+run "count_distinct"      supported  "SELECT COUNT(DISTINCT CounterID) FROM hits"
+run "groupby_counter"     supported  "SELECT CounterID, COUNT(*) FROM hits GROUP BY CounterID"
+run "where_and"           supported  "SELECT COUNT(*) FROM hits WHERE Age > 25 AND EventDate >= '2013-07-10'"
+run "groupby_topk"        supported  "SELECT CounterID, COUNT(*) AS c FROM hits GROUP BY CounterID ORDER BY c DESC LIMIT 10"

zighouse/install

@@ -1,10 +1,11 @@
 #!/bin/bash
 set -e
 
-: "${ZIGHOUSE_VERSION:=v0.1.0-clickbench}"
-: "${ZIGHOUSE_SHA256:=5a779eacf87082eeeb13b336d6f798a0399593c29b5a586ab67202104396dc83}"
+: "${ZIGHOUSE_VERSION:=v1.0.3}"
+: "${ZIGHOUSE_SHA256:=4f63353bfc9b09a649fb54c77c447a113b4bf2b331957aa65d4b33648ef56d5b}"
 
-url="https://github.com/donge/zighouse/releases/download/${ZIGHOUSE_VERSION}/zighouse-linux-x86_64"
-curl -L --fail -o zighouse "$url"
-printf '%s  zighouse\n' "$ZIGHOUSE_SHA256" | sha256sum -c -
+url="https://github.com/donge/zighouse/releases/download/${ZIGHOUSE_VERSION}/zighouse-linux-x86_64.gz"
+curl -L --fail -o zighouse.gz "$url"
+printf '%s  zighouse.gz\n' "$ZIGHOUSE_SHA256" | sha256sum -c -
+gunzip zighouse.gz
 chmod +x zighouse

zighouse/load

@@ -2,5 +2,6 @@
 set -e
 
 rm -rf zighouse-store
-ZIGHOUSE_CLICKBENCH_SUBMIT=1 ZIGHOUSE_IMPORT_TRACE=1 ./zighouse import-clickbench-parquet-hot hits.parquet zighouse-store
+ZIGHOUSE_CLICKBENCH_SUBMIT=1 ZIGHOUSE_IMPORT_TRACE=1 \
+  ./zighouse import-parquet --format=generic hits.parquet zighouse-store hits
 sync

zighouse/query

@@ -2,4 +2,7 @@
 set -e
 
 query=$(cat)
-ZIGHOUSE_CLICKBENCH_SUBMIT=1 ./zighouse --backend native query-timed zighouse-store "$query"
+start=$(date +%s.%N)
+ZIGHOUSE_CLICKBENCH_SUBMIT=1 ./zighouse query zighouse-store hits "$query" >/dev/null
+end=$(date +%s.%N)
+awk -v s="$start" -v e="$end" 'BEGIN { printf "%.3f\n", e - s }' >&2

zighouse/results/20260721/c6a.2xlarge.json

@@ -0,0 +1,60 @@
+{
+      "system": "ZigHouse",
+      "date": "2026-07-21",
+      "machine": "c6a.2xlarge",
+      "cluster_size": 1,
+      "proprietary": "no",
+      "hardware": "cpu",
+      "tuned": "no",
+      "tags": ["Zig","column-oriented","ClickHouse-compatible"],
+      "load_time": 1936,
+      "data_size": 81370427084,
+      "concurrent_qps": null,
+      "concurrent_error_ratio": null,
+      "result": [
+        [0.027, 0.008, 0.008],
+        [0.257, 0.024, 0.025],
+        [0.879, 0.034, 0.032],
+        [2.848, 0.056, 0.056],
+        [3.231, 0.839, 0.849],
+        [5.266, 0.899, 0.901],
+        [0.573, 0.095, 0.096],
+        [0.42, 0.052, 0.053],
+        [5.222, 1.422, 1.403],
+        [10.909, 2.173, 2.199],
+        [5.825, 0.795, 0.794],
+        [6.564, 0.791, 0.791],
+        [5.487, 0.822, 0.818],
+        [10.213, 2.587, 2.587],
+        [6.307, 0.938, 0.936],
+        [2.715, 1.014, 1.022],
+        [10.482, 3.005, 2.979],
+        [9.964, 2.452, 2.455],
+        [77.543, 167.512, 67.008],
+        [null, null, null],
+        [37.077, 0.884, 0.888],
+        [36.211, 0.676, 0.678],
+        [47.643, 0.886, 0.88],
+        [38.521, 0.936, 0.936],
+        [5.168, 0.124, 0.123],
+        [5.358, 0.184, 0.183],
+        [5.554, 0.132, 0.125],
+        [4.083, 0.624, 0.623],
+        [29.578, 3.488, 3.584],
+        [0.207, 0.025, 0.024],
+        [7.208, 1.681, 1.687],
+        [8.932, 0.979, 0.959],
+        [21.589, 21.765, 22.042],
+        [47.015, 46.924, 47.183],
+        [47.016, 46.644, 46.797],
+        [1.261, 0.959, 0.961],
+        [0.911, 0.076, 0.075],
+        [1.43, 0.069, 0.067],
+        [1.534, 0.05, 0.05],
+        [4.224, 0.194, 0.194],
+        [0.899, 0.05, 0.05],
+        [2.88, 0.072, 0.07],
+        [4.302, 0.291, 0.292]
+]
+  } 
+

zighouse/results/20260721/c6a.4xlarge.json

@@ -0,0 +1,60 @@
+{
+      "system": "ZigHouse",
+      "date": "2026-07-21",
+      "machine": "c6a.4xlarge",
+      "cluster_size": 1,
+      "proprietary": "no",
+      "hardware": "cpu",
+      "tuned": "no",
+      "tags": ["Zig","column-oriented","ClickHouse-compatible"],
+      "load_time": 1676,
+      "data_size": 81370427084,
+      "concurrent_qps": null,
+      "concurrent_error_ratio": null,
+      "result": [
+        [0.025, 0.01, 0.009],
+        [0.216, 0.024, 0.024],
+        [0.932, 0.035, 0.032],
+        [2.841, 0.056, 0.056],
+        [3.114, 0.639, 0.642],
+        [5.331, 0.778, 0.781],
+        [0.572, 0.072, 0.072],
+        [0.468, 0.04, 0.039],
+        [5.155, 1.112, 1.11],
+        [11.765, 2.076, 2.083],
+        [5.821, 0.713, 0.713],
+        [6.564, 0.705, 0.71],
+        [5.416, 0.651, 0.647],
+        [10.128, 2.336, 2.376],
+        [6.216, 0.715, 0.724],
+        [2.65, 0.769, 0.761],
+        [10.045, 2.338, 2.322],
+        [9.498, 1.797, 1.802],
+        [24.837, 14.537, 14.607],
+        [null, null, null],
+        [37.012, 0.591, 0.591],
+        [36.284, 0.654, 0.649],
+        [47.774, 0.927, 0.929],
+        [38.803, 0.621, 0.62],
+        [5.176, 0.096, 0.096],
+        [5.395, 0.148, 0.148],
+        [5.653, 0.11, 0.106],
+        [4.064, 0.459, 0.464],
+        [29.558, 2.761, 2.738],
+        [0.189, 0.024, 0.024],
+        [7.686, 2.03, 2.033],
+        [8.93, 0.889, 0.885],
+        [13.309, 8.469, 8.474],
+        [39.661, 3.911, 3.932],
+        [39.672, 3.942, 3.949],
+        [1.181, 0.716, 0.716],
+        [0.913, 0.07, 0.069],
+        [1.416, 0.064, 0.064],
+        [1.527, 0.047, 0.047],
+        [4.216, 0.177, 0.178],
+        [0.903, 0.046, 0.047],
+        [2.856, 0.07, 0.071],
+        [4.29, 0.355, 0.35]
+]
+  } 
+