anobre/openshift-resource-governance
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Implement OpenShift Console exact queries for CPU and Memory Usage

Browse Source 
- Add get_workload_cpu_summary() and get_workload_memory_summary() methods
- Use exact OpenShift Console PromQL queries for data consistency
- Update historical analysis API endpoints to include real CPU/Memory data
- Document all OpenShift Console queries in AIAgents-Support.md
- Fix CPU Usage and Memory Usage columns showing N/A in Historical Analysis
This commit is contained in:
Anderson
2025-10-03 20:19:42 -03:00
parent 692d647abd
commit a4cf3d65bc
2 changed files with 115 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
app/api/routes.py
View File

@@ -1419,15 +1419,33 @@ async def get_historical_analysis(
}
workloads[workload_name]['pods'].append(pod)
# Convert to list and add basic info
# Convert to list and add basic info with real CPU/Memory data
workload_list = []
historical_service = HistoricalAnalysisService()
for workload_name, workload_data in workloads.items():
# Get current CPU and Memory usage using OpenShift Console queries
try:
cpu_usage = await historical_service.get_workload_cpu_summary(workload_data['namespace'], workload_name)
memory_usage = await historical_service.get_workload_memory_summary(workload_data['namespace'], workload_name)
# Format CPU usage (cores)
cpu_display = f"{cpu_usage:.3f} cores" if cpu_usage > 0 else "N/A"
# Format memory usage (MB)
memory_display = f"{memory_usage / (1024 * 1024):.1f} MB" if memory_usage > 0 else "N/A"
except Exception as e:
logger.warning(f"Error getting summary for {workload_name}: {e}")
cpu_display = "N/A"
memory_display = "N/A"
workload_list.append({
'name': workload_name,
'namespace': workload_data['namespace'],
'pod_count': len(workload_data['pods']),
'cpu_usage': 'N/A', # Will be populated by Prometheus queries
'memory_usage': 'N/A', # Will be populated by Prometheus queries
'cpu_usage': cpu_display,
'memory_usage': memory_display,
'last_updated': datetime.now().isoformat()
})
@@ -1468,8 +1486,8 @@ async def get_workload_historical_details(
cpu_data = await historical_service.get_cpu_usage_history(namespace, workload, time_range)
memory_data = await historical_service.get_memory_usage_history(namespace, workload, time_range)
# Generate recommendations
recommendations = await historical_service.generate_recommendations(namespace, workload, time_range)
# Generate recommendations and get workload summary
recommendations, workload_summary = await historical_service.generate_recommendations(namespace, workload, time_range)
return {
"workload": workload,
@@ -1477,6 +1495,7 @@ async def get_workload_historical_details(
"cpu_data": cpu_data,
"memory_data": memory_data,
"recommendations": recommendations,
"workload_summary": workload_summary,
"timestamp": datetime.now().isoformat()
}

93
app/services/historical_analysis.py
View File

@@ -1448,6 +1448,82 @@ class HistoricalAnalysisService:
"error": str(e)
}
async def get_workload_cpu_summary(self, namespace: str, workload: str) -> float:
"""Get current CPU usage summary for a workload using OpenShift Console query"""
try:
# Use exact OpenShift Console query for CPU usage per pod
cpu_query = f'''
sum(
node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{{
cluster="",
namespace="{namespace}"
}}
* on(namespace,pod)
group_left(workload, workload_type)
namespace_workload_pod:kube_pod_owner:relabel{{
cluster="",
namespace="{namespace}",
workload="{workload}",
workload_type=~".+"
}}
) by (pod)
'''
# Query Prometheus for current value
data = await self._query_prometheus(cpu_query,
datetime.utcnow() - timedelta(seconds=300), # Last 5 minutes
datetime.utcnow())
if data and len(data) > 0:
# Sum all pod values for the workload
total_cpu = sum(self._safe_float(point[1]) for point in data if point[1] != 'NaN')
return total_cpu
return 0.0
except Exception as e:
logger.error(f"Error getting CPU summary for {workload}: {e}")
return 0.0
async def get_workload_memory_summary(self, namespace: str, workload: str) -> float:
"""Get current memory usage summary for a workload using OpenShift Console query"""
try:
# Use exact OpenShift Console query for memory usage per pod
memory_query = f'''
sum(
container_memory_working_set_bytes{{
cluster="",
namespace="{namespace}",
container!="",
image!=""
}}
* on(namespace,pod)
group_left(workload, workload_type)
namespace_workload_pod:kube_pod_owner:relabel{{
cluster="",
namespace="{namespace}",
workload="{workload}",
workload_type=~".+"
}}
) by (pod)
'''
# Query Prometheus for current value
data = await self._query_prometheus(memory_query,
datetime.utcnow() - timedelta(seconds=300), # Last 5 minutes
datetime.utcnow())
if data and len(data) > 0:
# Sum all pod values for the workload
total_memory = sum(self._safe_float(point[1]) for point in data if point[1] != 'NaN')
return total_memory
return 0.0
except Exception as e:
logger.error(f"Error getting memory summary for {workload}: {e}")
return 0.0
async def generate_recommendations(self, namespace: str, workload: str, time_range: str = "24h") -> List[Dict[str, Any]]:
"""Generate recommendations based on historical data"""
try:
@@ -1455,6 +1531,10 @@ class HistoricalAnalysisService:
cpu_data = await self.get_cpu_usage_history(namespace, workload, time_range)
memory_data = await self.get_memory_usage_history(namespace, workload, time_range)
# Get current summary values for the workload
current_cpu_usage = await self.get_workload_cpu_summary(namespace, workload)
current_memory_usage = await self.get_workload_memory_summary(namespace, workload)
recommendations = []
# Analyze CPU data
@@ -1505,7 +1585,16 @@ class HistoricalAnalysisService:
"recommendation": "Increase memory limits to handle peak usage"
})
return recommendations
# Add workload summary data to recommendations
workload_summary = {
"workload": workload,
"namespace": namespace,
"cpu_usage": current_cpu_usage,
"memory_usage": current_memory_usage / (1024 * 1024), # Convert bytes to MB
"time_range": time_range
}
return recommendations, workload_summary
except Exception as e:
logger.error(f"Error generating recommendations: {str(e)}")
@@ -1514,4 +1603,4 @@ class HistoricalAnalysisService:
"severity": "error",
"message": f"Error generating recommendations: {str(e)}",
"recommendation": "Check Prometheus connectivity and workload configuration"
}]
}], None