Merge pull request #5499 from VladLazar/arroyo-tests

Run arroyo test suite in ducktape

tests/docker/Dockerfile

@@ -163,6 +163,13 @@ RUN apt update && \
       file && \
     rm -rf /var/lib/apt/lists/*
 
+# Clone and install the arroyo client library in order to run its test suite.
+RUN mkdir /root/external_test_suites && \
+    git -C /root/external_test_suites clone --depth=1 https://github.com/getsentry/arroyo.git && \
+    cd /root/external_test_suites/arroyo && \
+    git reset --hard 2631cf1406b0cb5bc05c8a37e8f9f5a40fcf31d4 && \
+    python3 -m pip install --force --no-cache-dir -e /root/external_test_suites/arroyo
+
 RUN mkdir -p /opt/scripts && \
     curl https://raw.githubusercontent.com/redpanda-data/seastar/2a9504b3238cba4150be59353bf8d0b3a01fe39c/scripts/addr2line.py -o /opt/scripts/addr2line.py && \
     curl https://raw.githubusercontent.com/redpanda-data/seastar/2a9504b3238cba4150be59353bf8d0b3a01fe39c/scripts/seastar-addr2line -o /opt/scripts/seastar-addr2line && \

tests/rptest/tests/compatibility/arroyo_test.py

@@ -0,0 +1,61 @@
+# Copyright 2022 Redpanda Data, Inc.
+#
+# Use of this software is governed by the Business Source License
+# included in the file licenses/BSL.md
+#
+# As of the Change Date specified in that file, in accordance with
+# the Business Source License, use of this software will be governed
+# by the Apache License, Version 2.0
+
+import os
+import subprocess
+
+from rptest.services.cluster import cluster
+from rptest.tests.prealloc_nodes import PreallocNodesTest
+from ducktape.cluster.remoteaccount import RemoteCommandError
+
+
+class ArroyoTest(PreallocNodesTest):
+    """
+    Run the arroyo test suite against a redpanda cluster in
+    a ducktape environment.
+
+    The test suite lives here under tests/ in https://github.com/getsentry/arroyo.
+    """
+    TEST_SUITE_PATH = "/root/external_test_suites/arroyo"
+
+    def __init__(self, ctx, *args, **kwargs):
+        super().__init__(test_context=ctx,
+                         node_prealloc_count=1,
+                         *args,
+                         **kwargs)
+
+    def _find_failed_tests(self, pytest_output: list[str]):
+        return [line for line in pytest_output if "FAILED" in line]
+
+    @cluster(num_nodes=4)
+    def test_arroyo_test_suite(self):
+        test_node = self.preallocated_nodes[0]
+
+        try:
+            env_preamble = f"DEFAULT_BROKERS={self.redpanda.brokers()}"
+            capture = test_node.account.ssh_capture(
+                f"{env_preamble} "
+                f"python3 -m pytest {ArroyoTest.TEST_SUITE_PATH} "
+                "-k KafkaStreamsTestCase -rf",
+                combine_stderr=True)
+
+            pytest_output = list(capture)
+            for log_line in pytest_output:
+                self.logger.info(log_line)
+
+            failure_reports = self._find_failed_tests(pytest_output)
+            if len(failure_reports) > 0:
+                assert False, "Arroyo test failures occured. Please check the log file"
+        except RemoteCommandError as err:
+            if err.exit_status == 2:
+                assert False, "Arroyo test suite was interrupted"
+            elif err.exit_status == 3:
+                assert False, "Internal error durring execution of Arroyo test suite"
+            elif err.exit_status == 4:
+                assert False, "Pytest command line invocation error"

tests/setup.py

@@ -17,7 +17,7 @@
         'prometheus-client==0.9.0', 'pyyaml==5.3.1', 'kafka-python==2.0.2',
         'crc32c==2.2', 'confluent-kafka==1.7.0', 'zstandard==0.15.2',
         'xxhash==2.0.2', 'protobuf==3.19.3', 'fastavro==1.4.9',
-        'psutil==5.9.0', 'numpy==1.22.3', 'pygal==3.0',
+        'psutil==5.9.0', 'numpy==1.22.3', 'pygal==3.0', 'pytest==7.1.2',
         'kafkatest@git+https://github.com/apache/kafka.git@058589b03db686803b33052d574ce887fb5cfbd1#egg=kafkatest&subdirectory=tests'
     ],
     scripts=[],