fix: add presto-trino demo to ci pipeline (#14477)

ci/scripts/gen-integration-test-yaml.py

@@ -37,8 +37,8 @@
     'starrocks-sink': ['json'],
     'deltalake-sink': ['json'],
     'pinot-sink': ['json'],
+    'presto-trino': ['json'],
     'client-library': ['none'],
-
 }
 
 def gen_pipeline_steps():

integration_tests/presto-trino/README.md

@@ -4,6 +4,7 @@
 
 1. Start the cluster with `docker compose up -d` command.
 The command will start a RisingWave cluster together with a integrated trino and presto instance.
+
 2. Connect the RisingWave frontend via the psql client. Create and insert data into the RisingWave table.
 ```shell
 psql -h localhost -p 4566 -d dev -U root
@@ -16,6 +17,7 @@ INSERT 0 1
 dev=> flush;
 FLUSH
 ```
+
 3. Query RisingWave from presto or trino
 ```shell
 # Start trino client

integration_tests/presto-trino/create_source.sql

@@ -0,0 +1,26 @@
+CREATE TABLE IF NOT EXISTS data_types (
+    id BIGINT PRIMARY KEY,
+    varchar_column VARCHAR,
+    text_column TEXT,
+    integer_column INTEGER,
+    smallint_column SMALLINT,
+    bigint_column BIGINT,
+    -- decimal_column DECIMAL, prestodb cannot handle postgres's decimal when the precision is unspecified, as the precision range exceeds its maximum precision.
+    real_column REAL,
+    double_column DOUBLE PRECISION,
+    boolean_column BOOLEAN,
+    date_column DATE,
+    time_column TIME,
+    timestamp_column TIMESTAMP,
+    timestamptz_column TIMESTAMPTZ,
+    jsonb_column JSONB,
+    bytea_column BYTEA
+);
+
+INSERT INTO data_types (id, varchar_column, text_column, integer_column, smallint_column, bigint_column, real_column, double_column, boolean_column, date_column, time_column, timestamp_column, timestamptz_column, jsonb_column, bytea_column)
+VALUES
+    (1, 'Varchar value 1', 'Text value 1', 123, 456, 789, 12.34, 90.12, TRUE, '2023-05-22', '12:34:56', '2023-05-22 12:34:56', '2023-05-22T12:34:56Z', '{"key": "value"}', E'\\xDEADBEEF'),
+    (2, 'Varchar value 2', 'Text value 2', 234, 567, 890, 23.45, 01.23, FALSE, '2023-05-23', '23:45:01', '2023-05-23 23:45:01', '2023-05-23T23:45:01Z', '{"key": "value2"}', E'\\xFEEDBEEF'),
+    (3, 'Varchar value 3', 'Text value 3', 345, 678, 901, 34.56, 12.34, TRUE, '2023-05-24', '12:34:56', '2023-05-24 12:34:56', '2023-05-24T12:34:56Z', '{"key": "value3"}', E'\\xCAFEBABE'),
+    (4, 'Varchar value 4', 'Text value 4', 456, 789, 012, 45.67, 23.45, FALSE, '2023-05-25', '23:45:01', '2023-05-25 23:45:01', '2023-05-25T23:45:01Z', '{"key": "value4"}', E'\\xBABEC0DE'),
+    (5, 'Varchar value 5', 'Text value 5', 567, 890, 123, 56.78, 34.56, TRUE, '2023-05-26', '12:34:56', '2023-05-26 12:34:56', '2023-05-26T12:34:56Z', '{"key": "value5"}', E'\\xDEADBABE');

integration_tests/presto-trino/data_check

@@ -0,0 +1 @@
+data_types

integration_tests/presto-trino/docker-compose.yml

@@ -33,14 +33,16 @@ services:
     profiles: [ "client" ]
     entrypoint: [ "trino", "--server", "trino:8080", "--catalog", "risingwave", "--schema", "public" ]
   presto:
-    image: ahanaio/prestodb-sandbox
+    image: prestodb/presto:0.284
     container_name: presto
     volumes:
       - ./etc/risingwave.properties:/opt/presto-server/etc/catalog/risingwave.properties
+      - ./etc/config.properties:/opt/presto-server/etc/config.properties
+      - ./etc/jvm.config:/opt/presto-server/etc/jvm.config
     ports:
       - "8080:8080"
   presto-client:
-    image: ahanaio/prestodb-sandbox
+    image: prestodb/presto:0.284
     profiles: [ "client" ]
     entrypoint: [ "presto-cli", "--server", "presto:8080", "--catalog", "risingwave", "--schema", "public" ]
 volumes:

integration_tests/presto-trino/etc/config.properties

@@ -0,0 +1,7 @@
+coordinator=true
+node-scheduler.include-coordinator=true
+http-server.http.port=8080
+query.max-memory=5GB
+query.max-memory-per-node=1GB
+discovery-server.enabled=true
+discovery.uri=http://0.0.0.0:8080

integration_tests/presto-trino/etc/jvm.config

@@ -0,0 +1,9 @@
+-server
+-Xmx16G
+-XX:+UseG1GC
+-XX:G1HeapRegionSize=32M
+-XX:+UseGCOverheadLimit
+-XX:+ExplicitGCInvokesConcurrent
+-XX:+HeapDumpOnOutOfMemoryError
+-XX:+ExitOnOutOfMemoryError
+-Djdk.attach.allowAttachSelf=true

integration_tests/presto-trino/sink_check.py

@@ -0,0 +1,16 @@
+import subprocess
+import sys
+
+failed_cases = []
+for client_image in ['presto-client', 'trino-client']:
+    output = subprocess.check_output(
+        ["docker", "compose", "run", client_image,
+            "--execute", "select * from data_types"],
+    )
+    rows_cnt = len(output.splitlines())
+    if rows_cnt < 1:
+        failed_cases.append(client_image)
+
+if len(failed_cases) != 0:
+    print(f"Data check failed for case {failed_cases}")
+    sys.exit(1)