Mei_Sheng_Textiles/nomad_mcp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

✨ Enhance log analysis with advanced filtering and search capabilities

Browse Source 
Add comprehensive log analysis features to improve troubleshooting workflows:
- Time-based filtering (8pm-6am shifts) with HH:MM format support
- Multi-level log filtering (ERROR, WARNING, EMERGENCY, etc.)
- Full-text search across log content with case-insensitive matching
- Proper line break formatting for readable output
- Line count limiting for large log files

New REST API endpoints:
- /api/logs/errors/{job_id} - Get only error/warning logs
- /api/logs/search/{job_id} - Search logs for specific terms
- Enhanced /api/logs/job/{job_id} with filtering parameters

New MCP tools:
- get_error_logs - Streamlined error analysis
- search_job_logs - Pattern-based log searching
- Enhanced get_job_logs with all filtering options

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
Nicolas Koehl
2025-06-02 17:42:35 +07:00
parent 3aa1d90b89
commit dc2fe4c425
3 changed files with 489 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files

293
mcp_server.py
View File

@ -9,6 +9,7 @@ import json
import logging
import os
import sys
import requests
from typing import Any, Dict, List, Optional
from mcp.server import NotificationOptions, Server
@ -172,7 +173,7 @@ async def handle_list_tools() -> List[types.Tool]:
),
types.Tool(
name="get_job_logs",
description="Get logs for a Nomad job",
description="Get logs for a Nomad job with advanced filtering options",
inputSchema={
"type": "object",
"properties": {
@ -184,11 +185,103 @@ async def handle_list_tools() -> List[types.Tool]:
"type": "string",
"description": "Nomad namespace",
"default": "development"
},
"log_type": {
"type": "string",
"description": "Type of logs: stdout or stderr",
"enum": ["stdout", "stderr"],
"default": "stderr"
},
"start_time": {
"type": "string",
"description": "Start time filter (HH:MM format, e.g., '20:00' for 8 PM)"
},
"end_time": {
"type": "string",
"description": "End time filter (HH:MM format, e.g., '06:00' for 6 AM)"
},
"log_level": {
"type": "string",
"description": "Filter by log level: ERROR, WARNING, INFO, DEBUG, EMERGENCY, CRITICAL"
},
"search": {
"type": "string",
"description": "Search term to filter logs"
},
"lines": {
"type": "integer",
"description": "Number of recent log lines to return"
},
"limit": {
"type": "integer",
"description": "Number of allocations to check",
"default": 1
}
},
"required": ["job_id"]
}
),
types.Tool(
name="get_error_logs",
description="Get only error and warning logs for a Nomad job, useful for troubleshooting",
inputSchema={
"type": "object",
"properties": {
"job_id": {
"type": "string",
"description": "ID of the job to get error logs for"
},
"namespace": {
"type": "string",
"description": "Nomad namespace",
"default": "development"
},
"start_time": {
"type": "string",
"description": "Start time filter (HH:MM format, e.g., '20:00' for 8 PM)"
},
"end_time": {
"type": "string",
"description": "End time filter (HH:MM format, e.g., '06:00' for 6 AM)"
}
},
"required": ["job_id"]
}
),
types.Tool(
name="search_job_logs",
description="Search Nomad job logs for specific terms or patterns",
inputSchema={
"type": "object",
"properties": {
"job_id": {
"type": "string",
"description": "ID of the job to search logs for"
},
"search_term": {
"type": "string",
"description": "Term or pattern to search for in logs"
},
"namespace": {
"type": "string",
"description": "Nomad namespace",
"default": "development"
},
"log_type": {
"type": "string",
"description": "Type of logs: stdout or stderr",
"enum": ["stdout", "stderr"],
"default": "stderr"
},
"lines": {
"type": "integer",
"description": "Number of matching lines to return",
"default": 50
}
},
"required": ["job_id", "search_term"]
}
),
types.Tool(
name="submit_job_file",
description="Submit a Nomad job from HCL or JSON file content",
@ -431,56 +524,174 @@ async def handle_call_tool(name: str, arguments: Dict[str, Any]) -> List[types.T
text="Error: job_id is required"
)]
# Get allocations
allocations = nomad_service.get_allocations(job_id)
if not allocations:
# Use the enhanced REST API endpoint
import requests
base_url = os.getenv("BASE_URL", "http://localhost:8000")
# Build query parameters
params = {
"namespace": arguments.get("namespace", namespace),
"log_type": arguments.get("log_type", "stderr"),
"limit": arguments.get("limit", 1),
"plain_text": True,
"formatted": True
}
# Add optional filters
if arguments.get("start_time"):
params["start_time"] = arguments["start_time"]
if arguments.get("end_time"):
params["end_time"] = arguments["end_time"]
if arguments.get("log_level"):
params["log_level"] = arguments["log_level"]
if arguments.get("search"):
params["search"] = arguments["search"]
if arguments.get("lines"):
params["lines"] = arguments["lines"]
try:
response = requests.get(
f"{base_url}/api/logs/job/{job_id}",
params=params,
timeout=30
)
if response.status_code == 200:
logs_text = response.text
result = {
"success": True,
"job_id": job_id,
"namespace": params["namespace"],
"message": f"Retrieved logs for job {job_id}",
"logs": logs_text,
"filters_applied": {k: v for k, v in params.items() if k not in ["namespace", "plain_text", "formatted"]}
}
else:
result = {
"success": False,
"job_id": job_id,
"message": f"Failed to get logs: {response.status_code} - {response.text}",
"logs": None
}
except Exception as e:
result = {
"success": False,
"job_id": job_id,
"message": f"No allocations found for job {job_id}",
"message": f"Error getting logs: {str(e)}",
"logs": None
}
else:
# Get latest allocation
sorted_allocations = sorted(
allocations,
key=lambda a: a.get("CreateTime", 0),
reverse=True
)
latest_alloc = sorted_allocations[0]
alloc_id = latest_alloc.get("ID")
# Get task name
task_name = None
if "TaskStates" in latest_alloc:
task_states = latest_alloc["TaskStates"]
if task_states:
task_name = next(iter(task_states.keys()))
if not task_name:
task_name = "app" # Default task name
# Get logs
stdout_logs = nomad_service.get_allocation_logs(alloc_id, task_name, "stdout")
stderr_logs = nomad_service.get_allocation_logs(alloc_id, task_name, "stderr")
result = {
"success": True,
"job_id": job_id,
"allocation_id": alloc_id,
"task_name": task_name,
"message": f"Retrieved logs for job {job_id}",
"logs": {
"stdout": stdout_logs,
"stderr": stderr_logs
}
}
return [types.TextContent(
type="text",
text=json.dumps(result, indent=2)
text=json.dumps(result, indent=2) if not result.get("success") else result["logs"]
)]
elif name == "get_error_logs":
job_id = arguments.get("job_id")
if not job_id:
return [types.TextContent(
type="text",
text="Error: job_id is required"
)]
# Use the error logs endpoint
import requests
base_url = os.getenv("BASE_URL", "http://localhost:8000")
params = {
"namespace": arguments.get("namespace", namespace),
"plain_text": True
}
if arguments.get("start_time"):
params["start_time"] = arguments["start_time"]
if arguments.get("end_time"):
params["end_time"] = arguments["end_time"]
try:
response = requests.get(
f"{base_url}/api/logs/errors/{job_id}",
params=params,
timeout=30
)
if response.status_code == 200:
logs_text = response.text
result = {
"success": True,
"job_id": job_id,
"message": f"Retrieved error logs for job {job_id}",
"logs": logs_text
}
return [types.TextContent(
type="text",
text=logs_text
)]
else:
return [types.TextContent(
type="text",
text=f"Error: Failed to get error logs: {response.status_code} - {response.text}"
)]
except Exception as e:
return [types.TextContent(
type="text",
text=f"Error getting error logs: {str(e)}"
)]
elif name == "search_job_logs":
job_id = arguments.get("job_id")
search_term = arguments.get("search_term")
if not job_id or not search_term:
return [types.TextContent(
type="text",
text="Error: job_id and search_term are required"
)]
# Use the search logs endpoint
import requests
base_url = os.getenv("BASE_URL", "http://localhost:8000")
params = {
"q": search_term,
"namespace": arguments.get("namespace", namespace),
"log_type": arguments.get("log_type", "stderr"),
"lines": arguments.get("lines", 50),
"plain_text": True
}
try:
response = requests.get(
f"{base_url}/api/logs/search/{job_id}",
params=params,
timeout=30
)
if response.status_code == 200:
logs_text = response.text
return [types.TextContent(
type="text",
text=logs_text
)]
else:
return [types.TextContent(
type="text",
text=f"Error: Failed to search logs: {response.status_code} - {response.text}"
)]
except Exception as e:
return [types.TextContent(
type="text",
text=f"Error searching logs: {str(e)}"
)]
elif name == "submit_job_file":
file_content = arguments.get("file_content")
file_type = arguments.get("file_type", "json")