[Feature Request]: Enhance Metrics Tab UI with Virtual Servers and Top 5 Performance Tables (#657)

* enhance metrics tab ui

Signed-off-by: IrushaBasukala <[email protected]>

* fixing error

Signed-off-by: IrushaBasukala <[email protected]>

* fixing build issues

Signed-off-by: IrushaBasukala <[email protected]>

* Fix many issues, rebase, still cannot properly add server

Signed-off-by: Mihai Criveti <[email protected]>

* Fix lint issues, rebae

Signed-off-by: Mihai Criveti <[email protected]>

* Fix lint issues, rebased

Signed-off-by: Mihai Criveti <[email protected]>

* Fixing the success rate

Signed-off-by: Mihai Criveti <[email protected]>

* Fixing the success rate

Signed-off-by: Mihai Criveti <[email protected]>

* Fix metrics display and calculations for tools, resources, and prompts

- Fixed SQLite boolean comparisons (using == 1 instead of .is_(True))
- Fixed Float type casting in SQLAlchemy queries
- Added division by zero protection
- Fixed frontend JavaScript to handle camelCase field names from API
- Added prompt metrics recording (partial - needs more work for auto-recording)
- Eliminated code duplication with shared metrics_common.py utility
- Fixed all pylint and flake8 issues

The metrics now correctly display success rates and execution counts for all entity types.

Signed-off-by: Mihai Criveti <[email protected]>

* Fixing lint-web

Signed-off-by: Mihai Criveti <[email protected]>

* Improve test coverage

Signed-off-by: Mihai Criveti <[email protected]>

---------

Signed-off-by: IrushaBasukala <[email protected]>
Signed-off-by: Mihai Criveti <[email protected]>
Co-authored-by: Mihai Criveti <[email protected]>

Author: IrushaBasukala

.pylintrc

@@ -575,7 +575,7 @@ contextmanager-decorators=contextlib.contextmanager
 # List of members which are set dynamically and missed by pylint inference
 # system, and so shouldn't trigger E1101 when accessed. Python regular
 # expressions are accepted.
-generated-members=
+generated-members=sqlalchemy.func.*
 
 # Tells whether to warn about missing members when the owner of the attribute
 # is inferred to be None.

MANIFEST.in

@@ -13,6 +13,7 @@ include Containerfile.lite
 include __init__
 include alembic.ini
 include tox.ini
+include alembic/README
 
 # 2️⃣  Top-level config, examples and helper scripts
 include *.py

mcpgateway/admin.py

@@ -3975,118 +3975,81 @@ async def admin_delete_root(uri: str, request: Request, user: str = Depends(requ
 MetricsDict = Dict[str, Union[ToolMetrics, ResourceMetrics, ServerMetrics, PromptMetrics]]
 
 
-@admin_router.get("/metrics", response_model=MetricsDict)
-async def admin_get_metrics(
+# @admin_router.get("/metrics", response_model=MetricsDict)
+# async def admin_get_metrics(
+#     db: Session = Depends(get_db),
+#     user: str = Depends(require_auth),
+# ) -> MetricsDict:
+#     """
+#     Retrieve aggregate metrics for all entity types via the admin UI.
+
+#     This endpoint collects and returns usage metrics for tools, resources, servers,
+#     and prompts. The metrics are retrieved by calling the aggregate_metrics method
+#     on each respective service, which compiles statistics about usage patterns,
+#     success rates, and other relevant metrics for administrative monitoring
+#     and analysis purposes.
+
+#     Args:
+#         db (Session): Database session dependency.
+#         user (str): Authenticated user dependency.
+
+#     Returns:
+#         MetricsDict: A dictionary containing the aggregated metrics for tools,
+#         resources, servers, and prompts. Each value is a Pydantic model instance
+#         specific to the entity type.
+#     """
+#     logger.debug(f"User {user} requested aggregate metrics")
+#     tool_metrics = await tool_service.aggregate_metrics(db)
+#     resource_metrics = await resource_service.aggregate_metrics(db)
+#     server_metrics = await server_service.aggregate_metrics(db)
+#     prompt_metrics = await prompt_service.aggregate_metrics(db)
+
+#     # Return actual Pydantic model instances
+#     return {
+#         "tools": tool_metrics,
+#         "resources": resource_metrics,
+#         "servers": server_metrics,
+#         "prompts": prompt_metrics,
+#     }
+
+
+@admin_router.get("/metrics")
+async def get_aggregated_metrics(
     db: Session = Depends(get_db),
-    user: str = Depends(require_auth),
-) -> MetricsDict:
-    """
-    Retrieve aggregate metrics for all entity types via the admin UI.
+    _user: str = Depends(require_auth),
+) -> Dict[str, Any]:
+    """Retrieve aggregated metrics and top performers for all entity types.
 
-    This endpoint collects and returns usage metrics for tools, resources, servers,
-    and prompts. The metrics are retrieved by calling the aggregate_metrics method
-    on each respective service, which compiles statistics about usage patterns,
-    success rates, and other relevant metrics for administrative monitoring
-    and analysis purposes.
+    This endpoint collects usage metrics and top-performing entities for tools,
+    resources, prompts, and servers by calling the respective service methods.
+    The results are compiled into a dictionary for administrative monitoring.
 
     Args:
-        db (Session): Database session dependency.
-        user (str): Authenticated user dependency.
+        db (Session): Database session dependency for querying metrics.
 
     Returns:
-        MetricsDict: A dictionary containing the aggregated metrics for tools,
-        resources, servers, and prompts. Each value is a Pydantic model instance
-        specific to the entity type.
-
-    Examples:
-        >>> import asyncio
-        >>> from unittest.mock import AsyncMock, MagicMock
-        >>> from mcpgateway.schemas import ToolMetrics, ResourceMetrics, ServerMetrics, PromptMetrics
-        >>>
-        >>> mock_db = MagicMock()
-        >>> mock_user = "test_user"
-        >>>
-        >>> mock_tool_metrics = ToolMetrics(
-        ...     total_executions=10,
-        ...     successful_executions=9,
-        ...     failed_executions=1,
-        ...     failure_rate=0.1,
-        ...     min_response_time=0.05,
-        ...     max_response_time=1.0,
-        ...     avg_response_time=0.3,
-        ...     last_execution_time=None
-        ... )
-        >>> mock_resource_metrics = ResourceMetrics(
-        ...     total_executions=5,
-        ...     successful_executions=5,
-        ...     failed_executions=0,
-        ...     failure_rate=0.0,
-        ...     min_response_time=0.1,
-        ...     max_response_time=0.5,
-        ...     avg_response_time=0.2,
-        ...     last_execution_time=None
-        ... )
-        >>> mock_server_metrics = ServerMetrics(
-        ...     total_executions=7,
-        ...     successful_executions=7,
-        ...     failed_executions=0,
-        ...     failure_rate=0.0,
-        ...     min_response_time=0.2,
-        ...     max_response_time=0.7,
-        ...     avg_response_time=0.4,
-        ...     last_execution_time=None
-        ... )
-        >>> mock_prompt_metrics = PromptMetrics(
-        ...     total_executions=3,
-        ...     successful_executions=3,
-        ...     failed_executions=0,
-        ...     failure_rate=0.0,
-        ...     min_response_time=0.15,
-        ...     max_response_time=0.6,
-        ...     avg_response_time=0.35,
-        ...     last_execution_time=None
-        ... )
-        >>>
-        >>> original_aggregate_metrics_tool = tool_service.aggregate_metrics
-        >>> original_aggregate_metrics_resource = resource_service.aggregate_metrics
-        >>> original_aggregate_metrics_server = server_service.aggregate_metrics
-        >>> original_aggregate_metrics_prompt = prompt_service.aggregate_metrics
-        >>>
-        >>> tool_service.aggregate_metrics = AsyncMock(return_value=mock_tool_metrics)
-        >>> resource_service.aggregate_metrics = AsyncMock(return_value=mock_resource_metrics)
-        >>> server_service.aggregate_metrics = AsyncMock(return_value=mock_server_metrics)
-        >>> prompt_service.aggregate_metrics = AsyncMock(return_value=mock_prompt_metrics)
-        >>>
-        >>> async def test_admin_get_metrics():
-        ...     result = await admin_get_metrics(mock_db, mock_user)
-        ...     return (
-        ...         isinstance(result, dict) and
-        ...         result.get("tools") == mock_tool_metrics and
-        ...         result.get("resources") == mock_resource_metrics and
-        ...         result.get("servers") == mock_server_metrics and
-        ...         result.get("prompts") == mock_prompt_metrics
-        ...     )
-        >>>
-        >>> import asyncio; asyncio.run(test_admin_get_metrics())
-        True
-        >>>
-        >>> tool_service.aggregate_metrics = original_aggregate_metrics_tool
-        >>> resource_service.aggregate_metrics = original_aggregate_metrics_resource
-        >>> server_service.aggregate_metrics = original_aggregate_metrics_server
-        >>> prompt_service.aggregate_metrics = original_aggregate_metrics_prompt
+        Dict[str, Any]: A dictionary containing aggregated metrics and top performers
+            for tools, resources, prompts, and servers. The structure includes:
+            - 'tools': Metrics for tools.
+            - 'resources': Metrics for resources.
+            - 'prompts': Metrics for prompts.
+            - 'servers': Metrics for servers.
+            - 'topPerformers': A nested dictionary with top 5 tools, resources, prompts,
+              and servers.
     """
-    logger.debug(f"User {user} requested aggregate metrics")
-    tool_metrics = await tool_service.aggregate_metrics(db)
-    resource_metrics = await resource_service.aggregate_metrics(db)
-    server_metrics = await server_service.aggregate_metrics(db)
-    prompt_metrics = await prompt_service.aggregate_metrics(db)
-
-    return {
-        "tools": tool_metrics,
-        "resources": resource_metrics,
-        "servers": server_metrics,
-        "prompts": prompt_metrics,
+    metrics = {
+        "tools": await tool_service.aggregate_metrics(db),
+        "resources": await resource_service.aggregate_metrics(db),
+        "prompts": await prompt_service.aggregate_metrics(db),
+        "servers": await server_service.aggregate_metrics(db),
+        "topPerformers": {
+            "tools": await tool_service.get_top_tools(db, limit=5),
+            "resources": await resource_service.get_top_resources(db, limit=5),
+            "prompts": await prompt_service.get_top_prompts(db, limit=5),
+            "servers": await server_service.get_top_servers(db, limit=5),
+        },
     }
+    return metrics
 
 
 @admin_router.post("/metrics/reset", response_model=Dict[str, object])

mcpgateway/main.py

@@ -29,6 +29,7 @@
 import asyncio
 from contextlib import asynccontextmanager
 import json
+import time
 from typing import Any, AsyncIterator, Dict, List, Optional, Union
 from urllib.parse import urlparse, urlunparse
 
@@ -52,7 +53,7 @@
 from fastapi.staticfiles import StaticFiles
 from fastapi.templating import Jinja2Templates
 from pydantic import ValidationError
-from sqlalchemy import text
+from sqlalchemy import select, text
 from sqlalchemy.exc import IntegrityError
 from sqlalchemy.orm import Session
 from starlette.middleware.base import BaseHTTPMiddleware
@@ -64,7 +65,8 @@
 from mcpgateway.bootstrap_db import main as bootstrap_db
 from mcpgateway.cache import ResourceCache, SessionRegistry
 from mcpgateway.config import jsonpath_modifier, settings
-from mcpgateway.db import refresh_slugs_on_startup, SessionLocal
+from mcpgateway.db import Prompt as DbPrompt
+from mcpgateway.db import PromptMetric, refresh_slugs_on_startup, SessionLocal
 from mcpgateway.handlers.sampling import SamplingHandler
 from mcpgateway.models import (
     InitializeRequest,
@@ -1780,17 +1782,49 @@ async def get_prompt(
 
     Returns:
         Rendered prompt or metadata.
+
+    Raises:
+        Exception: Re-raised if not a handled exception type.
     """
     logger.debug(f"User: {user} requested prompt: {name} with args={args}")
+    start_time = time.monotonic()
+    success = False
+    error_message = None
+    result = None
+
     try:
         PromptExecuteArgs(args=args)
-        return await prompt_service.get_prompt(db, name, args)
+        result = await prompt_service.get_prompt(db, name, args)
+        success = True
+        logger.debug(f"Prompt execution successful for '{name}'")
     except Exception as ex:
+        error_message = str(ex)
         logger.error(f"Could not retrieve prompt {name}: {ex}")
         if isinstance(ex, (ValueError, PromptError)):
-            return JSONResponse(content={"message": "Prompt execution arguments contains HTML tags that may cause security issues"}, status_code=422)
-        if isinstance(ex, PluginViolationError):
-            return JSONResponse(content={"message": "Prompt execution arguments contains HTML tags that may cause security issues", "details": ex.message}, status_code=422)
+            result = JSONResponse(content={"message": "Prompt execution arguments contains HTML tags that may cause security issues"}, status_code=422)
+        elif isinstance(ex, PluginViolationError):
+            result = JSONResponse(content={"message": "Prompt execution arguments contains HTML tags that may cause security issues", "details": ex.message}, status_code=422)
+        else:
+            raise
+
+    # Record metrics (moved outside try/except/finally to ensure it runs)
+    end_time = time.monotonic()
+    response_time = end_time - start_time
+
+    # Get the prompt from database to get its ID
+    prompt = db.execute(select(DbPrompt).where(DbPrompt.name == name)).scalar_one_or_none()
+
+    if prompt:
+        metric = PromptMetric(
+            prompt_id=prompt.id,
+            response_time=response_time,
+            is_success=success,
+            error_message=error_message,
+        )
+        db.add(metric)
+        db.commit()
+
+    return result
 
 
 @prompt_router.get("/{name}")
@@ -1810,9 +1844,41 @@ async def get_prompt_no_args(
 
     Returns:
         The prompt template information
+
+    Raises:
+        Exception: Re-raised from prompt service.
     """
     logger.debug(f"User: {user} requested prompt: {name} with no arguments")
-    return await prompt_service.get_prompt(db, name, {})
+    start_time = time.monotonic()
+    success = False
+    error_message = None
+    result = None
+
+    try:
+        result = await prompt_service.get_prompt(db, name, {})
+        success = True
+    except Exception as ex:
+        error_message = str(ex)
+        raise
+
+    # Record metrics
+    end_time = time.monotonic()
+    response_time = end_time - start_time
+
+    # Get the prompt from database to get its ID
+    prompt = db.execute(select(DbPrompt).where(DbPrompt.name == name)).scalar_one_or_none()
+
+    if prompt:
+        metric = PromptMetric(
+            prompt_id=prompt.id,
+            response_time=response_time,
+            is_success=success,
+            error_message=error_message,
+        )
+        db.add(metric)
+        db.commit()
+
+    return result
 
 
 @prompt_router.put("/{name}", response_model=PromptRead)

mcpgateway/schemas.py

@@ -590,7 +590,7 @@ def prevent_manual_mcp_creation(cls, values: Dict[str, Any]) -> Dict[str, Any]:
         """
         integration_type = values.get("integration_type")
         if integration_type == "MCP":
-            raise ValueError("Cannot manually create MCP tools. Add MCP servers via the Gateways interface - " "tools will be auto-discovered and registered with integration_type='MCP'.")
+            raise ValueError("Cannot manually create MCP tools. Add MCP servers via the Gateways interface - tools will be auto-discovered and registered with integration_type='MCP'.")
         return values
 
 
@@ -2936,3 +2936,26 @@ class TagInfo(BaseModelWithConfigDict):
     name: str = Field(..., description="The tag name")
     stats: TagStats = Field(..., description="Statistics for this tag")
     entities: Optional[List[TaggedEntity]] = Field(default_factory=list, description="Entities that have this tag")
+
+
+class TopPerformer(BaseModelWithConfigDict):
+    """Schema for representing top-performing entities with performance metrics.
+
+    Used to encapsulate metrics for entities such as prompts, resources, servers, or tools,
+    including execution count, average response time, success rate, and last execution timestamp.
+
+    Attributes:
+        id (Union[str, int]): Unique identifier for the entity.
+        name (str): Name of the entity (e.g., prompt name, resource URI, server name, or tool name).
+        execution_count (int): Total number of executions for the entity.
+        avg_response_time (Optional[float]): Average response time in seconds, or None if no metrics.
+        success_rate (Optional[float]): Success rate percentage, or None if no metrics.
+        last_execution (Optional[datetime]): Timestamp of the last execution, or None if no metrics.
+    """
+
+    id: Union[str, int] = Field(..., description="Entity ID")
+    name: str = Field(..., description="Entity name")
+    execution_count: int = Field(..., description="Number of executions")
+    avg_response_time: Optional[float] = Field(None, description="Average response time in seconds")
+    success_rate: Optional[float] = Field(None, description="Success rate percentage")
+    last_execution: Optional[datetime] = Field(None, description="Timestamp of last execution")

mcpgateway/services/logging_service.py

@@ -123,7 +123,7 @@ async def initialize(self) -> None:
             try:
                 root_logger.addHandler(_get_file_handler())
                 if settings.log_rotation_enabled:
-                    logging.info(f"File logging enabled with rotation: {settings.log_folder or '.'}/{settings.log_file} " f"(max: {settings.log_max_size_mb}MB, backups: {settings.log_backup_count})")
+                    logging.info(f"File logging enabled with rotation: {settings.log_folder or '.'}/{settings.log_file} (max: {settings.log_max_size_mb}MB, backups: {settings.log_backup_count})")
                 else:
                     logging.info(f"File logging enabled (no rotation): {settings.log_folder or '.'}/{settings.log_file}")
             except Exception as e: