fix(cdk): add error handling to _read_memory() for graceful degradation

Co-Authored-By: bot_apk <apk@cognition.ai>

Author: devin-ai-integration[bot]

airbyte_cdk/utils/memory_monitor.py

@@ -68,32 +68,35 @@ def _read_memory(self) -> Optional[tuple[int, int]]:
         """Read current memory usage and limit from cgroup files.
 
         Returns a tuple of (usage_bytes, limit_bytes) or None if unavailable.
+        Best-effort: failures to read memory info never crash a sync.
         """
         if self._cgroup_version is None:
             return None
 
-        if self._cgroup_version == 2:
-            usage_path = _CGROUP_V2_CURRENT
-            limit_path = _CGROUP_V2_MAX
-        else:
-            usage_path = _CGROUP_V1_USAGE
-            limit_path = _CGROUP_V1_LIMIT
+        try:
+            if self._cgroup_version == 2:
+                usage_path = _CGROUP_V2_CURRENT
+                limit_path = _CGROUP_V2_MAX
+            else:
+                usage_path = _CGROUP_V1_USAGE
+                limit_path = _CGROUP_V1_LIMIT
 
-        usage_text = usage_path.read_text().strip()
-        limit_text = limit_path.read_text().strip()
+            limit_text = limit_path.read_text().strip()
+            # cgroup v2 memory.max can be the literal string "max" (unlimited)
+            if limit_text == "max":
+                return None
 
-        # cgroup v2 memory.max can be the literal string "max" (unlimited)
-        if limit_text == "max":
-            return None
+            usage_bytes = int(usage_path.read_text().strip())
+            limit_bytes = int(limit_text)
 
-        usage_bytes = int(usage_text)
-        limit_bytes = int(limit_text)
+            if limit_bytes <= 0:
+                return None
 
-        if limit_bytes <= 0:
+            return usage_bytes, limit_bytes
+        except (OSError, ValueError):
+            logger.debug("Failed to read cgroup memory files; skipping memory check.")
             return None
 
-        return usage_bytes, limit_bytes
-
     def check_memory_usage(self) -> None:
         """Check memory usage against thresholds.
 

unit_tests/utils/test_memory_monitor.py

@@ -264,6 +264,44 @@ def mock_read_text(self: Path) -> str:
             with pytest.raises(MemoryLimitExceeded):
                 monitor.check_memory_usage()
 
+    def test_malformed_cgroup_file_degrades_gracefully(self) -> None:
+        """Malformed cgroup files should not crash the sync."""
+
+        def mock_exists(self: Path) -> bool:
+            return self in (_CGROUP_V2_CURRENT, _CGROUP_V2_MAX)
+
+        def mock_read_text(self: Path) -> str:
+            return "not_a_number\n"
+
+        with patch.object(Path, "exists", mock_exists):
+            monitor = MemoryMonitor()
+
+        with patch.object(Path, "read_text", mock_read_text):
+            # Should not raise — degrades gracefully
+            monitor.check_memory_usage()
+
+        assert not monitor._warning_emitted
+        assert not monitor._critical_raised
+
+    def test_os_error_degrades_gracefully(self) -> None:
+        """OSError reading cgroup files should not crash the sync."""
+
+        def mock_exists(self: Path) -> bool:
+            return self in (_CGROUP_V2_CURRENT, _CGROUP_V2_MAX)
+
+        def mock_read_text(self: Path) -> str:
+            raise OSError("Permission denied")
+
+        with patch.object(Path, "exists", mock_exists):
+            monitor = MemoryMonitor()
+
+        with patch.object(Path, "read_text", mock_read_text):
+            # Should not raise — degrades gracefully
+            monitor.check_memory_usage()
+
+        assert not monitor._warning_emitted
+        assert not monitor._critical_raised
+
 
 class TestMemoryLimitExceeded:
     """Tests for the MemoryLimitExceeded exception."""