Merge pull request #2 from sundy-li/databend

improve(databend): use latest databend version and correct some numbers
timescale · Jul 13, 2022 · 8181fdc · 8181fdc
2 parents 794becb + 3664141
commit 8181fdc
Show file tree

Hide file tree

Showing 2 changed files with 59 additions and 71 deletions.
diff --git a/databend/benchmark.sh b/databend/benchmark.sh
@@ -1,8 +1,8 @@
 #!/bin/bash
 
 mkdir databend && cd databend
-curl -LJO 'https://github.com/datafuselabs/databend/releases/download/v0.7.113-nightly/databend-v0.7.113-nightly-x86_64-unknown-linux-musl.tar.gz'
-tar xzvf 'databend-v0.7.113-nightly-x86_64-unknown-linux-musl.tar.gz'
+curl -LJO 'https://github.com/datafuselabs/databend/releases/download/v0.7.127-nightly/databend-v0.7.127-nightly-x86_64-unknown-linux-musl.tar.gz'
+tar xzvf 'databend-v0.7.127-nightly-x86_64-unknown-linux-musl.tar.gz'
 
 echo 'dir = "metadata/_logs"
 admin_api_address = "127.0.0.1:8101"
@@ -68,38 +68,26 @@ curl https://clickhouse.com/ | sh
 sudo ./clickhouse install
 
 # Load the data
+# Docs: https://databend.rs/doc/learn/analyze-hits-dataset-with-databend
 
-curl 'http://default@localhost:8124/' --data-binary @create.sql
+curl 'http://default@localhost:8125/' --data-binary @create.sql
 
-wget --continue 'https://datasets.clickhouse.com/hits_compatible/hits.csv.gz'
-gzip -d hits.csv.gz
+wget --continue 'https://datasets.clickhouse.com/hits_compatible/hits.tsv.gz'
+gzip -d hits.tsv.gz
 
-# Note: if I run
+# Note:
 # clickhouse-client --time --query "INSERT INTO hits FORMAT TSV" < hits.tsv
-# it panics:
-# ERROR common_tracing::panic_hook: panicked at 'called `Result::unwrap()` on an `Err` value: SendError
+# can work but it's a bit slower than streaming load
 
-# Note: if I run
-# curl -XPUT 'http://root:@127.0.0.1:8000/v1/streaming_load' -H 'insert_sql: insert into hits format CSV' -H 'skip_header: 0' -H 'field_delimiter: ,' -H 'record_delimiter: \n' -F 'upload=@"./hits.csv"'
-# curl: (55) Send failure: Broken pipe
+time curl -XPUT 'http://root:@127.0.0.1:8081/v1/streaming_load' -H 'insert_sql: insert into hits format TSV' -H 'skip_header: 0' -H 'field_delimiter: \t' -H 'record_delimiter: \n' -F 'upload=@"./hits.tsv"'
 
-# This is not entirely correct, but starts to work:
-# curl -XPUT 'http://root:@127.0.0.1:8000/v1/streaming_load' -H 'insert_sql: insert into hits format TSV' -H 'skip_header: 0' -H 'field_delimiter: \t' -H 'record_delimiter: \n' -F 'upload=@"./hits.tsv"'
-# and fails after 7 minutes 38 seconds without loading any data:
-# Code: 4000, displayText = invalid data (Expected to have terminated string literal.) (while in processor thread 5).
-# the diagnostics is terrible.
+# {"id":"b9e20026-7eb2-4f09-a6b3-3ab79c4cb1fd","state":"SUCCESS","stats":{"rows":99997497,"bytes":81443407174},"error":null}
+# real    8m17.103s
 
-head -n 90000000 hits.tsv > hits90m.tsv
-time curl -XPUT 'http://root:@127.0.0.1:8000/v1/streaming_load' -H 'insert_sql: insert into hits format TSV' -H 'skip_header: 0' -H 'field_delimiter: \t' -H 'record_delimiter: \n' -F 'upload=@"./hits90m.tsv"'
+wc -l hits.tsv                                  1
+# 99997497 hits.tsv
 
-# {"id":"08f59e6c-2924-483e-bb96-cbcb458588f5","state":"SUCCESS","stats":{"rows":90000000,"bytes":73152552024},"error":null}
-# real    7m15.312s
-
-du -bcs _data
-# 38714978944
-
-# It does not support ClickHouse protocol well (it hangs on some queries if they are too long).
+du -bcs benddata
+# 15376458490
 
 ./run.sh 2>&1 | tee log.txt
-
-# Note: divide every number by 0.9 as only 90% of the data was loaded successfully.
diff --git a/databend/results/c6a.4xlarge.json b/databend/results/c6a.4xlarge.json
@@ -3,56 +3,56 @@
     "date": "2022-07-01",
     "machine": "c6a.4xlarge, 500gb gp2",
     "cluster_size": 1,
-    "comment": "Only 90% of data successfully loaded.",
+    "comment": "",
 
     "tags": ["Rust", "column-oriented", "ClickHouse derivative"],
 
-    "load_time": 484,
-    "data_size": 43016643271,
+    "load_time": 497,
+    "data_size": 15376458490,
 
     "result": [
-[0.010087, 0.002961, 0.003271],
-[0.127964, 0.080012, 0.075741],
-[0.162388, 0.143967, 0.144762],
-[0.252904, 0.217471, 0.217369],
-[34.281026, 34.844158, 34.526942],
-[25.290307, 25.793068, 25.620563],
-[0.112484, 0.093867, 0.090891],
-[0.086604, 0.07796, 0.076448],
-[20.723203, 20.7483, 20.354869],
-[20.81994, 20.72446, 20.696573],
-[1.964378, 1.93559, 1.893824],
-[1.846866, 1.789111, 1.763664],
-[4.468158, 4.407959, 4.438036],
-[19.947276, 19.8859, 19.853514],
-[5.478573, 5.474461, 5.460604],
-[5.509521, 5.513413, 5.363123],
-[15.430359, 15.5406, 15.461211],
-[14.905998, 15.029721, 15.019642],
-[31.069663, 30.811763, 30.737336],
-[0.281067, 0.220021, 0.217741],
-[8.89374, 4.12692, 4.131689],
-[10.38448, 4.603694, 4.571757],
-[19.980572, 8.836322, 8.892694],
-[59.786474, 52.452881, 39.941988],
-[2.804019, 0.994794, 0.958224],
-[0.765299, 0.730434, 0.723964],
-[2.784648, 0.94665, 0.936684],
-[8.905027, 5.418438, 5.386109],
-[12.187652, 12.230066, 12.164123],
-[3.35748, 3.395991, 3.319434],
-[4.309389, 3.854977, 3.772506],
-[9.958201, 7.027432, 6.888253],
-[50.200569, 50.535126, 50.283066],
-[24.469412, 21.222713, 21.010188],
-[26.115852, 23.93507, 24.835342],
-[7.511517, 7.296179, 7.324549],
-[2.156784, 1.298258, 1.278441],
-[2.155447, 1.314499, 1.331237],
-[2.007053, 1.181676, 1.155612],
+[0.013902, 0.004422, 0.004122],
+[0.101888, 0.104271, 0.073037],
+[0.168212, 0.141071, 0.142650],
+[0.243747, 0.203377, 0.204867],
+[33.140331, 34.229988, 33.679454],
+[26.319933, 27.272582, 25.767886],
+[0.110273, 0.090036, 0.092146],
+[0.084274, 0.076594, 0.074711],
+[19.987245, 20.475889, 20.303076],
+[20.374940, 20.468011, 20.483538],
+[1.963584, 1.907614, 1.869584],
+[1.894976, 1.876273, 1.879085],
+[4.593503, 4.507381, 4.462135],
+[19.678799, 19.799023, 19.694602],
+[5.575033, 5.484197, 5.419428],
+[6.095004, 6.054998, 6.095125],
+[14.886180, 15.040093, 15.089891],
+[14.785431, 14.593298, 14.753010],
+[29.590730, 29.849316, 29.661235],
+[0.299483, 0.224117, 0.214859],
+[9.179294, 4.006400, 3.993079],
+[10.670255, 4.430619, 4.398151],
+[20.355242, 8.688529, 8.802367],
+[60.704942, 55.980810, 56.102004],
+[2.940611, 0.889858, 0.897108],
+[0.744909, 0.703389, 0.700635],
+[2.906223, 0.887556, 0.895110],
+[9.252901, 5.417411, 5.492932],
+[12.345430, 12.032540, 11.790489],
+[3.465527, 3.389873, 3.372103],
+[4.263470, 3.611936, 3.615036],
+[9.795959, 6.932302, 6.912054],
+[48.752064, 49.549056, 49.086483],
+[26.008925, 22.072019, 21.553101],
+[27.521008, 25.469400, 25.627924],
+[7.262501, 7.209355, 7.242376],
+[1.880355, 1.138028, 1.171569],
+[1.881614, 1.146638, 1.164525],
+[1.750486, 0.956581, 1.048284],
 [null, null, null],
-[0.485363, 0.420291, 0.416819],
-[0.372131, 0.322068, 0.323578],
+[0.406606, 0.351536, 0.367830],
+[0.314160, 0.290318, 0.271116],
 [null, null, null]
 ]
 }