fix(group_common_events): filter expected raft_topology error messages

sdcm/nemesis.py

@@ -82,6 +82,7 @@
     ignore_no_space_errors,
     ignore_scrub_invalid_errors,
     ignore_stream_mutation_fragments_errors,
+    ignore_raft_topology_cmd_failing,
     ignore_ycsb_connection_refused,
     decorate_with_context,
     ignore_reactor_stall_errors,
@@ -3853,7 +3854,7 @@ def decommission_post_action():
             for expected_start_failed_context in self.target_node.raft.get_severity_change_filters_scylla_start_failed(
                     terminate_pattern.timeout):
                 stack.enter_context(expected_start_failed_context)
-            with ignore_stream_mutation_fragments_errors():
+            with ignore_stream_mutation_fragments_errors(), ignore_raft_topology_cmd_failing():
                 ParallelObject(objects=[trigger, watcher], timeout=full_operations_timeout).call_objects()
             if new_node := decommission_post_action():
                 new_node.wait_node_fully_start()
@@ -3932,21 +3933,22 @@ def disrupt_decommission_streaming_err(self):
                 "This nemesis logic is not compatible with K8S approach "
                 "for handling Scylla member's decommissioning.")
 
-        with ignore_stream_mutation_fragments_errors():
+        with ignore_stream_mutation_fragments_errors(), ignore_raft_topology_cmd_failing():
             self.start_and_interrupt_decommission_streaming()
 
     def disrupt_rebuild_streaming_err(self):
         """
         Stop rebuild in middle to trigger some streaming fails, then rebuild the data on the node.
         """
-        with ignore_stream_mutation_fragments_errors():
+        with ignore_stream_mutation_fragments_errors(), ignore_raft_topology_cmd_failing():
             self.start_and_interrupt_rebuild_streaming()
 
     def disrupt_repair_streaming_err(self):
         """
         Stop repair in middle to trigger some streaming fails, then rebuild the data on the node.
         """
-        self.start_and_interrupt_repair_streaming()
+        with ignore_raft_topology_cmd_failing():
+            self.start_and_interrupt_repair_streaming()
 
     def _corrupt_data_file(self):
         """Randomly corrupt data file by dd"""

sdcm/sct_events/group_common_events.py

@@ -298,6 +298,42 @@ def ignore_stream_mutation_fragments_errors():
         yield
 
 
+@contextmanager
+def ignore_raft_topology_cmd_failing():
+    with ExitStack() as stack:
+        stack.enter_context(EventsSeverityChangerFilter(
+            new_severity=Severity.WARNING,
+            event_class=DatabaseLogEvent,
+            regex=r".*raft_topology - raft_topology_cmd failed with: seastar::abort_requested_exception \(abort requested\)",
+            extra_time_to_expiration=30
+        ))
+        stack.enter_context(EventsSeverityChangerFilter(
+            new_severity=Severity.WARNING,
+            event_class=DatabaseLogEvent,
+            regex=r".*raft_topology - raft_topology_cmd failed with: raft::request_aborted \(Request is aborted by a caller\)",
+            extra_time_to_expiration=30
+        ))
+        stack.enter_context(EventsSeverityChangerFilter(
+            new_severity=Severity.WARNING,
+            event_class=DatabaseLogEvent,
+            regex=r".*raft_topology - send_raft_topology_cmd\(stream_ranges\) failed with exception \(node state is decommissioning\)",
+            extra_time_to_expiration=30
+        ))
+        stack.enter_context(EventsSeverityChangerFilter(
+            new_severity=Severity.WARNING,
+            event_class=DatabaseLogEvent,
+            regex=r".*raft_topology - send_raft_topology_cmd\(stream_ranges\) failed with exception \(node state is rebuilding\)",
+            extra_time_to_expiration=30
+        ))
+        stack.enter_context(EventsSeverityChangerFilter(
+            new_severity=Severity.WARNING,
+            event_class=DatabaseLogEvent,
+            regex=r".*raft_topology - send_raft_topology_cmd\(stream_ranges\) failed with exception \(node state is replacing\)",
+            extra_time_to_expiration=30
+        ))
+        yield
+
+
 def decorate_with_context(context_list: list[Callable | ContextManager] | Callable | ContextManager):
     """
     helper to decorate a function to run with a list of callables that return context managers

sdcm/utils/raft/common.py

@@ -8,7 +8,7 @@
 from sdcm.cluster import BaseNode, BaseScyllaCluster, BaseMonitorSet
 from sdcm.wait import wait_for
 from sdcm.sct_events.group_common_events import decorate_with_context, \
-    ignore_stream_mutation_fragments_errors, ignore_ycsb_connection_refused
+    ignore_stream_mutation_fragments_errors, ignore_ycsb_connection_refused, ignore_raft_topology_cmd_failing
 from sdcm.utils.adaptive_timeouts import Operations, adaptive_timeout
 from sdcm.utils.common import ParallelObject
 
@@ -117,7 +117,7 @@ def run_bootstrap_and_abort_with_action(self, terminate_pattern, abort_action: C
 
         wait_operations_timeout = (self.SUCCESS_BOOTSTRAP_TIMEOUT + self.INSTANCE_START_TIMEOUT
                                    + terminate_pattern.timeout + abort_action_timeout)
-        with ignore_stream_mutation_fragments_errors(), contextlib.ExitStack() as stack:
+        with ignore_stream_mutation_fragments_errors(), ignore_raft_topology_cmd_failing(), contextlib.ExitStack() as stack:
             for expected_start_failed_context in self.verification_node.raft.get_severity_change_filters_scylla_start_failed(
                     terminate_pattern.timeout):
                 stack.enter_context(expected_start_failed_context)
@@ -142,7 +142,9 @@ def clean_and_restart_bootstrap_after_abort(self):
                 self.verification_node.raft.clean_group0_garbage(raise_exception=True)
                 LOGGER.debug("Clean old scylla data and restart scylla service")
                 self.bootstrap_node.clean_scylla_data()
-            with adaptive_timeout(operation=Operations.NEW_NODE, node=self.verification_node, timeout=3600) as bootstrap_timeout:
+            with ignore_raft_topology_cmd_failing(), \
+                    adaptive_timeout(operation=Operations.NEW_NODE, node=self.verification_node, timeout=3600) as bootstrap_timeout:
+
                 self.bootstrap_node.start_scylla_server(verify_up_timeout=bootstrap_timeout, verify_down=True)
                 self.bootstrap_node.start_scylla_jmx()
             self.db_cluster.check_nodes_up_and_normal(