Stability testing: workload_log (#13804)

Author: naspirato

ydb/tests/stability/library/__main__.py

@@ -39,6 +39,7 @@ def __init__(self, ssh_username, cluster_path, ydbd_path, ydbd_next_path=None):
             self._unpack_resource('simple_queue'),
             self._unpack_resource('olap_workload'),
             self._unpack_resource('statistics_workload'),
+            self._unpack_resource('ydb_cli'),
         )
 
         self.kikimr_cluster = ExternalKiKiMRCluster(
@@ -193,7 +194,7 @@ def cleanup(self, mode='all'):
             if mode in ['all', 'dumps']:
                 node.ssh_command('sudo rm -rf /coredumps/*', raise_on_error=False)
             if mode in ['all', 'logs']:
-                node.ssh_command('sudo rm -rf /Berkanavt/kikimr_31003/logs/*', raise_on_error=False)
+                node.ssh_command('sudo rm -rf /Berkanavt/kikimr_31*/logs/*', raise_on_error=False)
                 node.ssh_command('sudo rm -rf /Berkanavt/kikimr/logs/*', raise_on_error=False)
                 node.ssh_command('sudo rm -rf /Berkanavt/nemesis/log/*', raise_on_error=False)
             if mode == 'all':
@@ -291,6 +292,7 @@ def parse_args():
             "start_workload_simple_queue_row",
             "start_workload_simple_queue_column",
             "start_workload_olap_workload",
+            "start_workload_log",
             "stop_workloads",
             "perform_checks",
         ],
@@ -341,6 +343,49 @@ def main():
                     'screen -d -m bash -c "while true; do /Berkanavt/nemesis/bin/simple_queue --database /Root/db1 --mode row; done"',
                     raise_on_error=True
                 )
+        if action == "start_workload_log":
+            for node_id, node in enumerate(stability_cluster.kikimr_cluster.nodes.values()):
+                if node_id == 1:
+                    node.ssh_command([
+                        '/Berkanavt/nemesis/bin/ydb_cli',
+                        '--endpoint', f'grpc://localhost:{node.grpc_port}',
+                        '--database', '/Root/db1',
+                        'workload', 'log', 'clean'
+                        ],
+                        raise_on_error=True
+                    )
+                    node.ssh_command([
+                        '/Berkanavt/nemesis/bin/ydb_cli',
+                        '--endpoint', f'grpc://localhost:{node.grpc_port}',
+                        '--database', '/Root/db1',
+                        'workload', 'log', 'init',
+                        '--len', '1000',
+                        '--int-cols', '20',
+                        '--key-cols', '20',
+                        '--min-partitions', '100',
+                        '--partition-size', '10',
+                        '--auto-partition', '0',
+                        '--ttl', '3600'
+                        ],
+                        raise_on_error=True
+                    )
+            for node_id, node in enumerate(stability_cluster.kikimr_cluster.nodes.values()):
+                node.ssh_command([
+                    'screen -s workload_log -d -m bash -c "while true; do',
+                    '/Berkanavt/nemesis/bin/ydb_cli',
+                    '--endpoint', f'grpc://localhost:{node.grpc_port}',
+                    '--database', '/Root/db1',
+                    'workload', 'log', 'run', 'bulk_upsert',
+                    '--len', '1000',
+                    '--int-cols', '20',
+                    '--key-cols', '20',
+                    '--threads', '2000',
+                    '--timestamp_deviation', '180',
+                    '--seconds', '86400',
+                    '; done"'
+                    ],
+                    raise_on_error=True
+                )
         if action == "start_workload_simple_queue_column":
             for node_id, node in enumerate(stability_cluster.kikimr_cluster.nodes.values()):
                 node.ssh_command(

ydb/tests/stability/library/how_to.md

@@ -20,6 +20,9 @@
     - `start_workload_simple_queue_row`
     - `start_workload_simple_queue_column`
     - `start_workload_olap_workload`
+    
+    not included in  `start_all_workloads`:
+    - `start_workload_log`
 
     ```
     ./library start_all_workloads --cluster_path=<path_to_cluster.yaml> --ydbd_path=<repo_root>/ydb/apps/ydbd/ydbd
@@ -48,7 +51,7 @@
         ``yq '.hosts[].name' <path_to_cluster.yaml> > ~/hosts.txt``
     2) check status
         ```
-        parallel-ssh -h ~/hosts.txt -i '
+        parallel-ssh -h ~/hosts.txt -i -x "-o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no" '
         if systemctl is-active --quiet nemesis; then
             echo "nemesis: Active"
         else
@@ -69,6 +72,11 @@
         else
             echo "simple_queue_column: Stopped"
         fi
+        if ps aux | grep "/Berkanavt/nemesis/bin/ydb_cli" | grep -v grep > /dev/null; then
+            echo "log workload: Running"
+        else
+            echo "log workload: Stopped"
+        fi
         '
         ```
 8) check cluster stability

ydb/tests/stability/library/ya.make

@@ -9,6 +9,7 @@ DATA(
 )
 
 DEPENDS(
+    ydb/apps/ydb
     ydb/tools/cfg/bin
     ydb/tests/tools/nemesis/driver
 )
@@ -19,9 +20,11 @@ BUNDLE(
     ydb/tests/stress/statistics_workload NAME statistics_workload
     ydb/tools/cfg/bin NAME cfg
     ydb/tests/tools/nemesis/driver NAME nemesis
+    ydb/apps/ydb NAME ydb_cli
 )
 
 RESOURCE(
+    ydb_cli ydb_cli
     simple_queue simple_queue
     olap_workload olap_workload
     statistics_workload statistics_workload