Add operation metrics with logging integration in metrics UI

app/__init__.py

@@ -16,6 +16,7 @@ from flask_wtf.csrf import CSRFError
 from werkzeug.middleware.proxy_fix import ProxyFix
 
 from .access_logging import AccessLoggingService
+from .operation_metrics import OperationMetricsCollector, classify_endpoint
 from .compression import GzipMiddleware
 from .acl import AclService
 from .bucket_policies import BucketPolicyStore
@@ -187,6 +188,15 @@ def create_app(
     app.extensions["notifications"] = notification_service
     app.extensions["access_logging"] = access_logging_service
 
+    operation_metrics_collector = None
+    if app.config.get("OPERATION_METRICS_ENABLED", False):
+        operation_metrics_collector = OperationMetricsCollector(
+            storage_root,
+            interval_minutes=app.config.get("OPERATION_METRICS_INTERVAL_MINUTES", 5),
+            retention_hours=app.config.get("OPERATION_METRICS_RETENTION_HOURS", 24),
+        )
+    app.extensions["operation_metrics"] = operation_metrics_collector
+
     @app.errorhandler(500)
     def internal_error(error):
         return render_template('500.html'), 500
@@ -356,6 +366,7 @@ def _configure_logging(app: Flask) -> None:
     def _log_request_start() -> None:
         g.request_id = uuid.uuid4().hex
         g.request_started_at = time.perf_counter()
+        g.request_bytes_in = request.content_length or 0
         app.logger.info(
             "Request started",
             extra={"path": request.path, "method": request.method, "remote_addr": request.remote_addr},
@@ -377,4 +388,21 @@ def _configure_logging(app: Flask) -> None:
             },
         )
         response.headers["X-Request-Duration-ms"] = f"{duration_ms:.2f}"
+
+        operation_metrics = app.extensions.get("operation_metrics")
+        if operation_metrics:
+            bytes_in = getattr(g, "request_bytes_in", 0)
+            bytes_out = response.content_length or 0
+            error_code = getattr(g, "s3_error_code", None)
+            endpoint_type = classify_endpoint(request.path)
+            operation_metrics.record_request(
+                method=request.method,
+                endpoint_type=endpoint_type,
+                status_code=response.status_code,
+                latency_ms=duration_ms,
+                bytes_in=bytes_in,
+                bytes_out=bytes_out,
+                error_code=error_code,
+            )
+
         return response

app/config.py

@@ -87,6 +87,9 @@ class AppConfig:
     metrics_history_enabled: bool
     metrics_history_retention_hours: int
     metrics_history_interval_minutes: int
+    operation_metrics_enabled: bool
+    operation_metrics_interval_minutes: int
+    operation_metrics_retention_hours: int
 
     @classmethod
     def from_env(cls, overrides: Optional[Dict[str, Any]] = None) -> "AppConfig":
@@ -186,6 +189,9 @@ class AppConfig:
         metrics_history_enabled = str(_get("METRICS_HISTORY_ENABLED", "0")).lower() in {"1", "true", "yes", "on"}
         metrics_history_retention_hours = int(_get("METRICS_HISTORY_RETENTION_HOURS", 24))
         metrics_history_interval_minutes = int(_get("METRICS_HISTORY_INTERVAL_MINUTES", 5))
+        operation_metrics_enabled = str(_get("OPERATION_METRICS_ENABLED", "0")).lower() in {"1", "true", "yes", "on"}
+        operation_metrics_interval_minutes = int(_get("OPERATION_METRICS_INTERVAL_MINUTES", 5))
+        operation_metrics_retention_hours = int(_get("OPERATION_METRICS_RETENTION_HOURS", 24))
 
         return cls(storage_root=storage_root,
                    max_upload_size=max_upload_size,
@@ -227,7 +233,10 @@ class AppConfig:
                    lifecycle_interval_seconds=lifecycle_interval_seconds,
                    metrics_history_enabled=metrics_history_enabled,
                    metrics_history_retention_hours=metrics_history_retention_hours,
-                   metrics_history_interval_minutes=metrics_history_interval_minutes)
+                   metrics_history_interval_minutes=metrics_history_interval_minutes,
+                   operation_metrics_enabled=operation_metrics_enabled,
+                   operation_metrics_interval_minutes=operation_metrics_interval_minutes,
+                   operation_metrics_retention_hours=operation_metrics_retention_hours)
 
     def validate_and_report(self) -> list[str]:
         """Validate configuration and return a list of warnings/issues.
@@ -359,4 +368,7 @@ class AppConfig:
             "METRICS_HISTORY_ENABLED": self.metrics_history_enabled,
             "METRICS_HISTORY_RETENTION_HOURS": self.metrics_history_retention_hours,
             "METRICS_HISTORY_INTERVAL_MINUTES": self.metrics_history_interval_minutes,
+            "OPERATION_METRICS_ENABLED": self.operation_metrics_enabled,
+            "OPERATION_METRICS_INTERVAL_MINUTES": self.operation_metrics_interval_minutes,
+            "OPERATION_METRICS_RETENTION_HOURS": self.operation_metrics_retention_hours,
         }

app/operation_metrics.py

@@ -0,0 +1,271 @@
+from __future__ import annotations
+
+import json
+import logging
+import threading
+import time
+from dataclasses import dataclass, field
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class OperationStats:
+    count: int = 0
+    success_count: int = 0
+    error_count: int = 0
+    latency_sum_ms: float = 0.0
+    latency_min_ms: float = float("inf")
+    latency_max_ms: float = 0.0
+    bytes_in: int = 0
+    bytes_out: int = 0
+
+    def record(self, latency_ms: float, success: bool, bytes_in: int = 0, bytes_out: int = 0) -> None:
+        self.count += 1
+        if success:
+            self.success_count += 1
+        else:
+            self.error_count += 1
+        self.latency_sum_ms += latency_ms
+        if latency_ms < self.latency_min_ms:
+            self.latency_min_ms = latency_ms
+        if latency_ms > self.latency_max_ms:
+            self.latency_max_ms = latency_ms
+        self.bytes_in += bytes_in
+        self.bytes_out += bytes_out
+
+    def to_dict(self) -> Dict[str, Any]:
+        avg_latency = self.latency_sum_ms / self.count if self.count > 0 else 0.0
+        min_latency = self.latency_min_ms if self.latency_min_ms != float("inf") else 0.0
+        return {
+            "count": self.count,
+            "success_count": self.success_count,
+            "error_count": self.error_count,
+            "latency_avg_ms": round(avg_latency, 2),
+            "latency_min_ms": round(min_latency, 2),
+            "latency_max_ms": round(self.latency_max_ms, 2),
+            "bytes_in": self.bytes_in,
+            "bytes_out": self.bytes_out,
+        }
+
+    def merge(self, other: "OperationStats") -> None:
+        self.count += other.count
+        self.success_count += other.success_count
+        self.error_count += other.error_count
+        self.latency_sum_ms += other.latency_sum_ms
+        if other.latency_min_ms < self.latency_min_ms:
+            self.latency_min_ms = other.latency_min_ms
+        if other.latency_max_ms > self.latency_max_ms:
+            self.latency_max_ms = other.latency_max_ms
+        self.bytes_in += other.bytes_in
+        self.bytes_out += other.bytes_out
+
+
+@dataclass
+class MetricsSnapshot:
+    timestamp: datetime
+    window_seconds: int
+    by_method: Dict[str, Dict[str, Any]]
+    by_endpoint: Dict[str, Dict[str, Any]]
+    by_status_class: Dict[str, int]
+    error_codes: Dict[str, int]
+    totals: Dict[str, Any]
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "timestamp": self.timestamp.isoformat(),
+            "window_seconds": self.window_seconds,
+            "by_method": self.by_method,
+            "by_endpoint": self.by_endpoint,
+            "by_status_class": self.by_status_class,
+            "error_codes": self.error_codes,
+            "totals": self.totals,
+        }
+
+    @classmethod
+    def from_dict(cls, data: Dict[str, Any]) -> "MetricsSnapshot":
+        return cls(
+            timestamp=datetime.fromisoformat(data["timestamp"]),
+            window_seconds=data.get("window_seconds", 300),
+            by_method=data.get("by_method", {}),
+            by_endpoint=data.get("by_endpoint", {}),
+            by_status_class=data.get("by_status_class", {}),
+            error_codes=data.get("error_codes", {}),
+            totals=data.get("totals", {}),
+        )
+
+
+class OperationMetricsCollector:
+    def __init__(
+        self,
+        storage_root: Path,
+        interval_minutes: int = 5,
+        retention_hours: int = 24,
+    ):
+        self.storage_root = storage_root
+        self.interval_seconds = interval_minutes * 60
+        self.retention_hours = retention_hours
+        self._lock = threading.Lock()
+        self._by_method: Dict[str, OperationStats] = {}
+        self._by_endpoint: Dict[str, OperationStats] = {}
+        self._by_status_class: Dict[str, int] = {}
+        self._error_codes: Dict[str, int] = {}
+        self._totals = OperationStats()
+        self._window_start = time.time()
+        self._shutdown = threading.Event()
+        self._snapshots: List[MetricsSnapshot] = []
+
+        self._load_history()
+
+        self._snapshot_thread = threading.Thread(
+            target=self._snapshot_loop, name="operation-metrics-snapshot", daemon=True
+        )
+        self._snapshot_thread.start()
+
+    def _config_path(self) -> Path:
+        return self.storage_root / ".myfsio.sys" / "config" / "operation_metrics.json"
+
+    def _load_history(self) -> None:
+        config_path = self._config_path()
+        if not config_path.exists():
+            return
+        try:
+            data = json.loads(config_path.read_text(encoding="utf-8"))
+            snapshots_data = data.get("snapshots", [])
+            self._snapshots = [MetricsSnapshot.from_dict(s) for s in snapshots_data]
+            self._prune_old_snapshots()
+        except (json.JSONDecodeError, OSError, KeyError) as e:
+            logger.warning(f"Failed to load operation metrics history: {e}")
+
+    def _save_history(self) -> None:
+        config_path = self._config_path()
+        config_path.parent.mkdir(parents=True, exist_ok=True)
+        try:
+            data = {"snapshots": [s.to_dict() for s in self._snapshots]}
+            config_path.write_text(json.dumps(data, indent=2), encoding="utf-8")
+        except OSError as e:
+            logger.warning(f"Failed to save operation metrics history: {e}")
+
+    def _prune_old_snapshots(self) -> None:
+        if not self._snapshots:
+            return
+        cutoff = datetime.now(timezone.utc).timestamp() - (self.retention_hours * 3600)
+        self._snapshots = [
+            s for s in self._snapshots if s.timestamp.timestamp() > cutoff
+        ]
+
+    def _snapshot_loop(self) -> None:
+        while not self._shutdown.is_set():
+            self._shutdown.wait(timeout=self.interval_seconds)
+            if not self._shutdown.is_set():
+                self._take_snapshot()
+
+    def _take_snapshot(self) -> None:
+        with self._lock:
+            now = datetime.now(timezone.utc)
+            window_seconds = int(time.time() - self._window_start)
+
+            snapshot = MetricsSnapshot(
+                timestamp=now,
+                window_seconds=window_seconds,
+                by_method={k: v.to_dict() for k, v in self._by_method.items()},
+                by_endpoint={k: v.to_dict() for k, v in self._by_endpoint.items()},
+                by_status_class=dict(self._by_status_class),
+                error_codes=dict(self._error_codes),
+                totals=self._totals.to_dict(),
+            )
+
+            self._snapshots.append(snapshot)
+            self._prune_old_snapshots()
+            self._save_history()
+
+            self._by_method.clear()
+            self._by_endpoint.clear()
+            self._by_status_class.clear()
+            self._error_codes.clear()
+            self._totals = OperationStats()
+            self._window_start = time.time()
+
+    def record_request(
+        self,
+        method: str,
+        endpoint_type: str,
+        status_code: int,
+        latency_ms: float,
+        bytes_in: int = 0,
+        bytes_out: int = 0,
+        error_code: Optional[str] = None,
+    ) -> None:
+        success = 200 <= status_code < 400
+        status_class = f"{status_code // 100}xx"
+
+        with self._lock:
+            if method not in self._by_method:
+                self._by_method[method] = OperationStats()
+            self._by_method[method].record(latency_ms, success, bytes_in, bytes_out)
+
+            if endpoint_type not in self._by_endpoint:
+                self._by_endpoint[endpoint_type] = OperationStats()
+            self._by_endpoint[endpoint_type].record(latency_ms, success, bytes_in, bytes_out)
+
+            self._by_status_class[status_class] = self._by_status_class.get(status_class, 0) + 1
+
+            if error_code:
+                self._error_codes[error_code] = self._error_codes.get(error_code, 0) + 1
+
+            self._totals.record(latency_ms, success, bytes_in, bytes_out)
+
+    def get_current_stats(self) -> Dict[str, Any]:
+        with self._lock:
+            window_seconds = int(time.time() - self._window_start)
+            return {
+                "timestamp": datetime.now(timezone.utc).isoformat(),
+                "window_seconds": window_seconds,
+                "by_method": {k: v.to_dict() for k, v in self._by_method.items()},
+                "by_endpoint": {k: v.to_dict() for k, v in self._by_endpoint.items()},
+                "by_status_class": dict(self._by_status_class),
+                "error_codes": dict(self._error_codes),
+                "totals": self._totals.to_dict(),
+            }
+
+    def get_history(self, hours: Optional[int] = None) -> List[Dict[str, Any]]:
+        with self._lock:
+            snapshots = list(self._snapshots)
+
+        if hours:
+            cutoff = datetime.now(timezone.utc).timestamp() - (hours * 3600)
+            snapshots = [s for s in snapshots if s.timestamp.timestamp() > cutoff]
+
+        return [s.to_dict() for s in snapshots]
+
+    def shutdown(self) -> None:
+        self._shutdown.set()
+        self._take_snapshot()
+        self._snapshot_thread.join(timeout=5.0)
+
+
+def classify_endpoint(path: str) -> str:
+    if not path or path == "/":
+        return "service"
+
+    path = path.rstrip("/")
+
+    if path.startswith("/ui"):
+        return "ui"
+
+    if path.startswith("/kms"):
+        return "kms"
+
+    if path.startswith("/myfsio"):
+        return "service"
+
+    parts = path.lstrip("/").split("/")
+    if len(parts) == 0:
+        return "service"
+    elif len(parts) == 1:
+        return "bucket"
+    else:
+        return "object"

app/s3_api.py

@@ -88,6 +88,7 @@ def _xml_response(element: Element, status: int = 200) -> Response:
 
 
 def _error_response(code: str, message: str, status: int) -> Response:
+    g.s3_error_code = code
     error = Element("Error")
     SubElement(error, "Code").text = code
     SubElement(error, "Message").text = message

app/ui.py

@@ -141,6 +141,10 @@ def _acl() -> AclService:
     return current_app.extensions["acl"]
 
 
+def _operation_metrics():
+    return current_app.extensions.get("operation_metrics")
+
+
 def _format_bytes(num: int) -> str:
     step = 1024
     units = ["B", "KB", "MB", "GB", "TB", "PB"]
@@ -2196,6 +2200,7 @@ def metrics_dashboard():
             "uptime_days": uptime_days,
         },
         metrics_history_enabled=current_app.config.get("METRICS_HISTORY_ENABLED", False),
+        operation_metrics_enabled=current_app.config.get("OPERATION_METRICS_ENABLED", False),
     )
 
 
@@ -2329,6 +2334,52 @@ def metrics_settings():
     })
 
 
+@ui_bp.get("/metrics/operations")
+def metrics_operations():
+    principal = _current_principal()
+
+    try:
+        _iam().authorize(principal, None, "iam:list_users")
+    except IamError:
+        return jsonify({"error": "Access denied"}), 403
+
+    collector = _operation_metrics()
+    if not collector:
+        return jsonify({
+            "enabled": False,
+            "stats": None,
+        })
+
+    return jsonify({
+        "enabled": True,
+        "stats": collector.get_current_stats(),
+    })
+
+
+@ui_bp.get("/metrics/operations/history")
+def metrics_operations_history():
+    principal = _current_principal()
+
+    try:
+        _iam().authorize(principal, None, "iam:list_users")
+    except IamError:
+        return jsonify({"error": "Access denied"}), 403
+
+    collector = _operation_metrics()
+    if not collector:
+        return jsonify({
+            "enabled": False,
+            "history": [],
+        })
+
+    hours = request.args.get("hours", type=int)
+    return jsonify({
+        "enabled": True,
+        "history": collector.get_history(hours),
+        "interval_minutes": current_app.config.get("OPERATION_METRICS_INTERVAL_MINUTES", 5),
+    })
+
+
 @ui_bp.route("/buckets/<bucket_name>/lifecycle", methods=["GET", "POST", "DELETE"])
 def bucket_lifecycle(bucket_name: str):
     principal = _current_principal()

templates/metrics.html

@@ -268,6 +268,121 @@
   </div>
 </div>
 
+{% if operation_metrics_enabled %}
+<div class="row g-4 mt-2">
+  <div class="col-12">
+    <div class="card shadow-sm border-0">
+      <div class="card-header bg-transparent border-0 pt-4 px-4 d-flex justify-content-between align-items-center">
+        <h5 class="card-title mb-0 fw-semibold">API Operations</h5>
+        <div class="d-flex align-items-center gap-3">
+          <span class="small text-muted" id="opStatus">Loading...</span>
+          <button class="btn btn-outline-secondary btn-sm" id="resetOpMetricsBtn" title="Reset current window">
+            <svg xmlns="http://www.w3.org/2000/svg" width="14" height="14" fill="currentColor" class="bi bi-arrow-counterclockwise" viewBox="0 0 16 16">
+              <path fill-rule="evenodd" d="M8 3a5 5 0 1 1-4.546 2.914.5.5 0 0 0-.908-.417A6 6 0 1 0 8 2v1z"/>
+              <path d="M8 4.466V.534a.25.25 0 0 0-.41-.192L5.23 2.308a.25.25 0 0 0 0 .384l2.36 1.966A.25.25 0 0 0 8 4.466z"/>
+            </svg>
+          </button>
+        </div>
+      </div>
+      <div class="card-body p-4">
+        <div class="row g-3 mb-4">
+          <div class="col-6 col-md-4 col-lg-2">
+            <div class="text-center p-3 bg-light rounded h-100">
+              <h4 class="fw-bold mb-1" id="opTotalRequests">0</h4>
+              <small class="text-muted">Requests</small>
+            </div>
+          </div>
+          <div class="col-6 col-md-4 col-lg-2">
+            <div class="text-center p-3 bg-light rounded h-100">
+              <h4 class="fw-bold mb-1 text-success" id="opSuccessRate">0%</h4>
+              <small class="text-muted">Success</small>
+            </div>
+          </div>
+          <div class="col-6 col-md-4 col-lg-2">
+            <div class="text-center p-3 bg-light rounded h-100">
+              <h4 class="fw-bold mb-1 text-danger" id="opErrorCount">0</h4>
+              <small class="text-muted">Errors</small>
+            </div>
+          </div>
+          <div class="col-6 col-md-4 col-lg-2">
+            <div class="text-center p-3 bg-light rounded h-100">
+              <h4 class="fw-bold mb-1 text-info" id="opAvgLatency">0ms</h4>
+              <small class="text-muted">Latency</small>
+            </div>
+          </div>
+          <div class="col-6 col-md-4 col-lg-2">
+            <div class="text-center p-3 bg-light rounded h-100">
+              <h4 class="fw-bold mb-1 text-primary" id="opBytesIn">0 B</h4>
+              <small class="text-muted">Bytes In</small>
+            </div>
+          </div>
+          <div class="col-6 col-md-4 col-lg-2">
+            <div class="text-center p-3 bg-light rounded h-100">
+              <h4 class="fw-bold mb-1 text-secondary" id="opBytesOut">0 B</h4>
+              <small class="text-muted">Bytes Out</small>
+            </div>
+          </div>
+        </div>
+        <div class="row g-4">
+          <div class="col-lg-6">
+            <div class="bg-light rounded p-3">
+              <h6 class="text-muted small fw-bold text-uppercase mb-3">Requests by Method</h6>
+              <div style="height: 220px; display: flex; align-items: center; justify-content: center;">
+                <canvas id="methodChart"></canvas>
+              </div>
+            </div>
+          </div>
+          <div class="col-lg-6">
+            <div class="bg-light rounded p-3">
+              <h6 class="text-muted small fw-bold text-uppercase mb-3">Requests by Status</h6>
+              <div style="height: 220px;">
+                <canvas id="statusChart"></canvas>
+              </div>
+            </div>
+          </div>
+        </div>
+        <div class="row g-4 mt-1">
+          <div class="col-lg-6">
+            <div class="bg-light rounded p-3">
+              <h6 class="text-muted small fw-bold text-uppercase mb-3">Requests by Endpoint</h6>
+              <div style="height: 180px;">
+                <canvas id="endpointChart"></canvas>
+              </div>
+            </div>
+          </div>
+          <div class="col-lg-6">
+            <div class="bg-light rounded p-3 h-100 d-flex flex-column">
+              <div class="d-flex justify-content-between align-items-start mb-3">
+                <h6 class="text-muted small fw-bold text-uppercase mb-0">S3 Error Codes</h6>
+                <span class="badge bg-secondary-subtle text-secondary" style="font-size: 0.65rem;" title="Tracks S3 API errors like NoSuchKey, AccessDenied, etc.">API Only</span>
+              </div>
+              <div class="flex-grow-1 d-flex flex-column" style="min-height: 150px;">
+                <div class="d-flex border-bottom pb-2 mb-2" style="font-size: 0.75rem;">
+                  <div class="text-muted fw-semibold" style="flex: 1;">Code</div>
+                  <div class="text-muted fw-semibold text-end" style="width: 60px;">Count</div>
+                  <div class="text-muted fw-semibold text-end" style="width: 100px;">Distribution</div>
+                </div>
+                <div id="errorCodesContainer" class="flex-grow-1" style="overflow-y: auto;">
+                  <div id="errorCodesBody">
+                    <div class="text-muted small text-center py-4">
+                      <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" fill="currentColor" class="bi bi-check-circle mb-2 text-success" viewBox="0 0 16 16">
+                        <path d="M8 15A7 7 0 1 1 8 1a7 7 0 0 1 0 14zm0 1A8 8 0 1 0 8 0a8 8 0 0 0 0 16z"/>
+                        <path d="M10.97 4.97a.235.235 0 0 0-.02.022L7.477 9.417 5.384 7.323a.75.75 0 0 0-1.06 1.06L6.97 11.03a.75.75 0 0 0 1.079-.02l3.992-4.99a.75.75 0 0 0-1.071-1.05z"/>
+                      </svg>
+                      <div>No S3 API errors</div>
+                    </div>
+                  </div>
+                </div>
+              </div>
+            </div>
+          </div>
+        </div>
+      </div>
+    </div>
+  </div>
+</div>
+{% endif %}
+
 {% if metrics_history_enabled %}
 <div class="row g-4 mt-2">
   <div class="col-12">
@@ -307,7 +422,7 @@
 {% endblock %}
 
 {% block extra_scripts %}
-{% if metrics_history_enabled %}
+{% if metrics_history_enabled or operation_metrics_enabled %}
 <script src="https://cdn.jsdelivr.net/npm/chart.js@4.4.1/dist/chart.umd.min.js"></script>
 {% endif %}
 <script>
@@ -413,6 +528,237 @@
   startPolling();
 })();
 
+{% if operation_metrics_enabled %}
+(function() {
+  var methodChart = null;
+  var statusChart = null;
+  var endpointChart = null;
+  var opStatus = document.getElementById('opStatus');
+  var opTimer = null;
+  var methodColors = {
+    'GET': '#0d6efd',
+    'PUT': '#198754',
+    'POST': '#ffc107',
+    'DELETE': '#dc3545',
+    'HEAD': '#6c757d',
+    'OPTIONS': '#0dcaf0'
+  };
+  var statusColors = {
+    '2xx': '#198754',
+    '3xx': '#0dcaf0',
+    '4xx': '#ffc107',
+    '5xx': '#dc3545'
+  };
+  var endpointColors = {
+    'object': '#0d6efd',
+    'bucket': '#198754',
+    'ui': '#6c757d',
+    'service': '#0dcaf0',
+    'kms': '#ffc107'
+  };
+
+  function formatBytes(bytes) {
+    if (bytes === 0) return '0 B';
+    var k = 1024;
+    var sizes = ['B', 'KB', 'MB', 'GB', 'TB'];
+    var i = Math.floor(Math.log(bytes) / Math.log(k));
+    return parseFloat((bytes / Math.pow(k, i)).toFixed(1)) + ' ' + sizes[i];
+  }
+
+  function initOpCharts() {
+    var methodCtx = document.getElementById('methodChart');
+    var statusCtx = document.getElementById('statusChart');
+    var endpointCtx = document.getElementById('endpointChart');
+
+    if (methodCtx) {
+      methodChart = new Chart(methodCtx, {
+        type: 'doughnut',
+        data: {
+          labels: [],
+          datasets: [{
+            data: [],
+            backgroundColor: []
+          }]
+        },
+        options: {
+          responsive: true,
+          maintainAspectRatio: false,
+          animation: false,
+          plugins: {
+            legend: { position: 'right', labels: { boxWidth: 12, font: { size: 11 } } }
+          }
+        }
+      });
+    }
+
+    if (statusCtx) {
+      statusChart = new Chart(statusCtx, {
+        type: 'bar',
+        data: {
+          labels: [],
+          datasets: [{
+            data: [],
+            backgroundColor: []
+          }]
+        },
+        options: {
+          responsive: true,
+          maintainAspectRatio: false,
+          animation: false,
+          plugins: { legend: { display: false } },
+          scales: {
+            y: { beginAtZero: true, ticks: { stepSize: 1 } }
+          }
+        }
+      });
+    }
+
+    if (endpointCtx) {
+      endpointChart = new Chart(endpointCtx, {
+        type: 'bar',
+        data: {
+          labels: [],
+          datasets: [{
+            data: [],
+            backgroundColor: []
+          }]
+        },
+        options: {
+          responsive: true,
+          maintainAspectRatio: false,
+          indexAxis: 'y',
+          animation: false,
+          plugins: { legend: { display: false } },
+          scales: {
+            x: { beginAtZero: true, ticks: { stepSize: 1 } }
+          }
+        }
+      });
+    }
+  }
+
+  function updateOpMetrics() {
+    if (document.hidden) return;
+    fetch('/ui/metrics/operations')
+      .then(function(r) { return r.json(); })
+      .then(function(data) {
+        if (!data.enabled || !data.stats) {
+          if (opStatus) opStatus.textContent = 'Operation metrics not available';
+          return;
+        }
+        var stats = data.stats;
+        var totals = stats.totals || {};
+
+        var totalEl = document.getElementById('opTotalRequests');
+        var successEl = document.getElementById('opSuccessRate');
+        var errorEl = document.getElementById('opErrorCount');
+        var latencyEl = document.getElementById('opAvgLatency');
+        var bytesInEl = document.getElementById('opBytesIn');
+        var bytesOutEl = document.getElementById('opBytesOut');
+
+        if (totalEl) totalEl.textContent = totals.count || 0;
+        if (successEl) {
+          var rate = totals.count > 0 ? ((totals.success_count / totals.count) * 100).toFixed(1) : 0;
+          successEl.textContent = rate + '%';
+        }
+        if (errorEl) errorEl.textContent = totals.error_count || 0;
+        if (latencyEl) latencyEl.textContent = (totals.latency_avg_ms || 0).toFixed(1) + 'ms';
+        if (bytesInEl) bytesInEl.textContent = formatBytes(totals.bytes_in || 0);
+        if (bytesOutEl) bytesOutEl.textContent = formatBytes(totals.bytes_out || 0);
+
+        if (methodChart && stats.by_method) {
+          var methods = Object.keys(stats.by_method);
+          var methodData = methods.map(function(m) { return stats.by_method[m].count; });
+          var methodBg = methods.map(function(m) { return methodColors[m] || '#6c757d'; });
+          methodChart.data.labels = methods;
+          methodChart.data.datasets[0].data = methodData;
+          methodChart.data.datasets[0].backgroundColor = methodBg;
+          methodChart.update('none');
+        }
+
+        if (statusChart && stats.by_status_class) {
+          var statuses = Object.keys(stats.by_status_class).sort();
+          var statusData = statuses.map(function(s) { return stats.by_status_class[s]; });
+          var statusBg = statuses.map(function(s) { return statusColors[s] || '#6c757d'; });
+          statusChart.data.labels = statuses;
+          statusChart.data.datasets[0].data = statusData;
+          statusChart.data.datasets[0].backgroundColor = statusBg;
+          statusChart.update('none');
+        }
+
+        if (endpointChart && stats.by_endpoint) {
+          var endpoints = Object.keys(stats.by_endpoint);
+          var endpointData = endpoints.map(function(e) { return stats.by_endpoint[e].count; });
+          var endpointBg = endpoints.map(function(e) { return endpointColors[e] || '#6c757d'; });
+          endpointChart.data.labels = endpoints;
+          endpointChart.data.datasets[0].data = endpointData;
+          endpointChart.data.datasets[0].backgroundColor = endpointBg;
+          endpointChart.update('none');
+        }
+
+        var errorBody = document.getElementById('errorCodesBody');
+        if (errorBody && stats.error_codes) {
+          var errorCodes = Object.entries(stats.error_codes);
+          errorCodes.sort(function(a, b) { return b[1] - a[1]; });
+          var totalErrors = errorCodes.reduce(function(sum, e) { return sum + e[1]; }, 0);
+          errorCodes = errorCodes.slice(0, 10);
+          if (errorCodes.length === 0) {
+            errorBody.innerHTML = '<div class="text-muted small text-center py-4">' +
+              '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" fill="currentColor" class="bi bi-check-circle mb-2 text-success" viewBox="0 0 16 16">' +
+              '<path d="M8 15A7 7 0 1 1 8 1a7 7 0 0 1 0 14zm0 1A8 8 0 1 0 8 0a8 8 0 0 0 0 16z"/>' +
+              '<path d="M10.97 4.97a.235.235 0 0 0-.02.022L7.477 9.417 5.384 7.323a.75.75 0 0 0-1.06 1.06L6.97 11.03a.75.75 0 0 0 1.079-.02l3.992-4.99a.75.75 0 0 0-1.071-1.05z"/>' +
+              '</svg><div>No S3 API errors</div></div>';
+          } else {
+            errorBody.innerHTML = errorCodes.map(function(e) {
+              var pct = totalErrors > 0 ? ((e[1] / totalErrors) * 100).toFixed(0) : 0;
+              return '<div class="d-flex align-items-center py-1" style="font-size: 0.8rem;">' +
+                '<div style="flex: 1;"><code class="text-danger">' + e[0] + '</code></div>' +
+                '<div class="text-end fw-semibold" style="width: 60px;">' + e[1] + '</div>' +
+                '<div style="width: 100px; padding-left: 10px;"><div class="progress" style="height: 6px;"><div class="progress-bar bg-danger" style="width: ' + pct + '%"></div></div></div>' +
+                '</div>';
+            }).join('');
+          }
+        }
+
+        var windowMins = Math.floor(stats.window_seconds / 60);
+        var windowSecs = stats.window_seconds % 60;
+        var windowStr = windowMins > 0 ? windowMins + 'm ' + windowSecs + 's' : windowSecs + 's';
+        if (opStatus) opStatus.textContent = 'Window: ' + windowStr + ' | ' + new Date().toLocaleTimeString();
+      })
+      .catch(function(err) {
+        console.error('Operation metrics fetch error:', err);
+        if (opStatus) opStatus.textContent = 'Failed to load';
+      });
+  }
+
+  function startOpPolling() {
+    if (opTimer) clearInterval(opTimer);
+    opTimer = setInterval(updateOpMetrics, 5000);
+  }
+
+  var resetBtn = document.getElementById('resetOpMetricsBtn');
+  if (resetBtn) {
+    resetBtn.addEventListener('click', function() {
+      updateOpMetrics();
+    });
+  }
+
+  document.addEventListener('visibilitychange', function() {
+    if (document.hidden) {
+      if (opTimer) clearInterval(opTimer);
+      opTimer = null;
+    } else {
+      updateOpMetrics();
+      startOpPolling();
+    }
+  });
+
+  initOpCharts();
+  updateOpMetrics();
+  startOpPolling();
+})();
+{% endif %}
+
 {% if metrics_history_enabled %}
 (function() {
   var cpuChart = null;

tests/test_operation_metrics.py

@@ -0,0 +1,297 @@
+import threading
+import time
+from pathlib import Path
+
+import pytest
+
+from app.operation_metrics import (
+    OperationMetricsCollector,
+    OperationStats,
+    classify_endpoint,
+)
+
+
+class TestOperationStats:
+    def test_initial_state(self):
+        stats = OperationStats()
+        assert stats.count == 0
+        assert stats.success_count == 0
+        assert stats.error_count == 0
+        assert stats.latency_sum_ms == 0.0
+        assert stats.bytes_in == 0
+        assert stats.bytes_out == 0
+
+    def test_record_success(self):
+        stats = OperationStats()
+        stats.record(latency_ms=50.0, success=True, bytes_in=100, bytes_out=200)
+
+        assert stats.count == 1
+        assert stats.success_count == 1
+        assert stats.error_count == 0
+        assert stats.latency_sum_ms == 50.0
+        assert stats.latency_min_ms == 50.0
+        assert stats.latency_max_ms == 50.0
+        assert stats.bytes_in == 100
+        assert stats.bytes_out == 200
+
+    def test_record_error(self):
+        stats = OperationStats()
+        stats.record(latency_ms=100.0, success=False, bytes_in=50, bytes_out=0)
+
+        assert stats.count == 1
+        assert stats.success_count == 0
+        assert stats.error_count == 1
+
+    def test_latency_min_max(self):
+        stats = OperationStats()
+        stats.record(latency_ms=50.0, success=True)
+        stats.record(latency_ms=10.0, success=True)
+        stats.record(latency_ms=100.0, success=True)
+
+        assert stats.latency_min_ms == 10.0
+        assert stats.latency_max_ms == 100.0
+        assert stats.latency_sum_ms == 160.0
+
+    def test_to_dict(self):
+        stats = OperationStats()
+        stats.record(latency_ms=50.0, success=True, bytes_in=100, bytes_out=200)
+        stats.record(latency_ms=100.0, success=False, bytes_in=50, bytes_out=0)
+
+        result = stats.to_dict()
+        assert result["count"] == 2
+        assert result["success_count"] == 1
+        assert result["error_count"] == 1
+        assert result["latency_avg_ms"] == 75.0
+        assert result["latency_min_ms"] == 50.0
+        assert result["latency_max_ms"] == 100.0
+        assert result["bytes_in"] == 150
+        assert result["bytes_out"] == 200
+
+    def test_to_dict_empty(self):
+        stats = OperationStats()
+        result = stats.to_dict()
+        assert result["count"] == 0
+        assert result["latency_avg_ms"] == 0.0
+        assert result["latency_min_ms"] == 0.0
+
+    def test_merge(self):
+        stats1 = OperationStats()
+        stats1.record(latency_ms=50.0, success=True, bytes_in=100, bytes_out=200)
+
+        stats2 = OperationStats()
+        stats2.record(latency_ms=10.0, success=True, bytes_in=50, bytes_out=100)
+        stats2.record(latency_ms=100.0, success=False, bytes_in=25, bytes_out=50)
+
+        stats1.merge(stats2)
+
+        assert stats1.count == 3
+        assert stats1.success_count == 2
+        assert stats1.error_count == 1
+        assert stats1.latency_min_ms == 10.0
+        assert stats1.latency_max_ms == 100.0
+        assert stats1.bytes_in == 175
+        assert stats1.bytes_out == 350
+
+
+class TestClassifyEndpoint:
+    def test_root_path(self):
+        assert classify_endpoint("/") == "service"
+        assert classify_endpoint("") == "service"
+
+    def test_ui_paths(self):
+        assert classify_endpoint("/ui") == "ui"
+        assert classify_endpoint("/ui/buckets") == "ui"
+        assert classify_endpoint("/ui/metrics") == "ui"
+
+    def test_kms_paths(self):
+        assert classify_endpoint("/kms") == "kms"
+        assert classify_endpoint("/kms/keys") == "kms"
+
+    def test_service_paths(self):
+        assert classify_endpoint("/myfsio/health") == "service"
+
+    def test_bucket_paths(self):
+        assert classify_endpoint("/mybucket") == "bucket"
+        assert classify_endpoint("/mybucket/") == "bucket"
+
+    def test_object_paths(self):
+        assert classify_endpoint("/mybucket/mykey") == "object"
+        assert classify_endpoint("/mybucket/folder/nested/key.txt") == "object"
+
+
+class TestOperationMetricsCollector:
+    def test_record_and_get_stats(self, tmp_path: Path):
+        collector = OperationMetricsCollector(
+            storage_root=tmp_path,
+            interval_minutes=60,
+            retention_hours=24,
+        )
+
+        try:
+            collector.record_request(
+                method="GET",
+                endpoint_type="bucket",
+                status_code=200,
+                latency_ms=50.0,
+                bytes_in=0,
+                bytes_out=1000,
+            )
+
+            collector.record_request(
+                method="PUT",
+                endpoint_type="object",
+                status_code=201,
+                latency_ms=100.0,
+                bytes_in=500,
+                bytes_out=0,
+            )
+
+            collector.record_request(
+                method="GET",
+                endpoint_type="object",
+                status_code=404,
+                latency_ms=25.0,
+                bytes_in=0,
+                bytes_out=0,
+                error_code="NoSuchKey",
+            )
+
+            stats = collector.get_current_stats()
+
+            assert stats["totals"]["count"] == 3
+            assert stats["totals"]["success_count"] == 2
+            assert stats["totals"]["error_count"] == 1
+
+            assert "GET" in stats["by_method"]
+            assert stats["by_method"]["GET"]["count"] == 2
+            assert "PUT" in stats["by_method"]
+            assert stats["by_method"]["PUT"]["count"] == 1
+
+            assert "bucket" in stats["by_endpoint"]
+            assert "object" in stats["by_endpoint"]
+            assert stats["by_endpoint"]["object"]["count"] == 2
+
+            assert stats["by_status_class"]["2xx"] == 2
+            assert stats["by_status_class"]["4xx"] == 1
+
+            assert stats["error_codes"]["NoSuchKey"] == 1
+        finally:
+            collector.shutdown()
+
+    def test_thread_safety(self, tmp_path: Path):
+        collector = OperationMetricsCollector(
+            storage_root=tmp_path,
+            interval_minutes=60,
+            retention_hours=24,
+        )
+
+        try:
+            num_threads = 5
+            requests_per_thread = 100
+            threads = []
+
+            def record_requests():
+                for _ in range(requests_per_thread):
+                    collector.record_request(
+                        method="GET",
+                        endpoint_type="object",
+                        status_code=200,
+                        latency_ms=10.0,
+                    )
+
+            for _ in range(num_threads):
+                t = threading.Thread(target=record_requests)
+                threads.append(t)
+                t.start()
+
+            for t in threads:
+                t.join()
+
+            stats = collector.get_current_stats()
+            assert stats["totals"]["count"] == num_threads * requests_per_thread
+        finally:
+            collector.shutdown()
+
+    def test_status_class_categorization(self, tmp_path: Path):
+        collector = OperationMetricsCollector(
+            storage_root=tmp_path,
+            interval_minutes=60,
+            retention_hours=24,
+        )
+
+        try:
+            collector.record_request("GET", "object", 200, 10.0)
+            collector.record_request("GET", "object", 204, 10.0)
+            collector.record_request("GET", "object", 301, 10.0)
+            collector.record_request("GET", "object", 304, 10.0)
+            collector.record_request("GET", "object", 400, 10.0)
+            collector.record_request("GET", "object", 403, 10.0)
+            collector.record_request("GET", "object", 404, 10.0)
+            collector.record_request("GET", "object", 500, 10.0)
+            collector.record_request("GET", "object", 503, 10.0)
+
+            stats = collector.get_current_stats()
+            assert stats["by_status_class"]["2xx"] == 2
+            assert stats["by_status_class"]["3xx"] == 2
+            assert stats["by_status_class"]["4xx"] == 3
+            assert stats["by_status_class"]["5xx"] == 2
+        finally:
+            collector.shutdown()
+
+    def test_error_code_tracking(self, tmp_path: Path):
+        collector = OperationMetricsCollector(
+            storage_root=tmp_path,
+            interval_minutes=60,
+            retention_hours=24,
+        )
+
+        try:
+            collector.record_request("GET", "object", 404, 10.0, error_code="NoSuchKey")
+            collector.record_request("GET", "object", 404, 10.0, error_code="NoSuchKey")
+            collector.record_request("GET", "bucket", 403, 10.0, error_code="AccessDenied")
+            collector.record_request("PUT", "object", 500, 10.0, error_code="InternalError")
+
+            stats = collector.get_current_stats()
+            assert stats["error_codes"]["NoSuchKey"] == 2
+            assert stats["error_codes"]["AccessDenied"] == 1
+            assert stats["error_codes"]["InternalError"] == 1
+        finally:
+            collector.shutdown()
+
+    def test_history_persistence(self, tmp_path: Path):
+        collector = OperationMetricsCollector(
+            storage_root=tmp_path,
+            interval_minutes=60,
+            retention_hours=24,
+        )
+
+        try:
+            collector.record_request("GET", "object", 200, 10.0)
+            collector._take_snapshot()
+
+            history = collector.get_history()
+            assert len(history) == 1
+            assert history[0]["totals"]["count"] == 1
+
+            config_path = tmp_path / ".myfsio.sys" / "config" / "operation_metrics.json"
+            assert config_path.exists()
+        finally:
+            collector.shutdown()
+
+    def test_get_history_with_hours_filter(self, tmp_path: Path):
+        collector = OperationMetricsCollector(
+            storage_root=tmp_path,
+            interval_minutes=60,
+            retention_hours=24,
+        )
+
+        try:
+            collector.record_request("GET", "object", 200, 10.0)
+            collector._take_snapshot()
+
+            history_all = collector.get_history()
+            history_recent = collector.get_history(hours=1)
+
+            assert len(history_all) >= len(history_recent)
+        finally:
+            collector.shutdown()