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

Update Prometheus queries to use OpenShift-specific metrics

Browse Source 
- Use node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate for CPU usage
- Use container_memory_working_set_bytes with kubelet job for memory usage
- Use kube_pod_container_resource_requests/limits with kube-state-metrics job
- Add workload-specific filtering to match OpenShift dashboard behavior
- This should resolve the 'insufficient data' issue by using the same metrics as OpenShift
This commit is contained in:
Anderson
2025-09-30 20:42:59 -03:00
parent 0068db5a9e
commit 3445f58a11
1 changed files with 82 additions and 58 deletions
Download Patch File Download Diff File Expand all files Collapse all files

140
app/api/routes.py
View File

@@ -499,25 +499,25 @@ async def get_workload_historical_metrics(
try: try:
prometheus_client = PrometheusClient() prometheus_client = PrometheusClient()
# Get current usage (latest values) # Get current usage using OpenShift-specific metrics
cpu_usage_query = f'rate(container_cpu_usage_seconds_total{{namespace="{namespace}",pod=~"{workload}-.*"}}[5m])' cpu_usage_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_type=~".+"}}) by (workload, workload_type)'
memory_usage_query = f'container_memory_working_set_bytes{{namespace="{namespace}",pod=~"{workload}-.*"}}' memory_usage_query = f'sum(container_memory_working_set_bytes{{job="kubelet", metrics_path="/metrics/cadvisor", cluster="", namespace="{namespace}", container!="", image!=""}} * on(namespace,pod) group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{{cluster="", namespace="{namespace}", workload_type=~".+"}}) by (workload, workload_type)'
cpu_usage_data = await prometheus_client.query(cpu_usage_query) cpu_usage_data = await prometheus_client.query(cpu_usage_query)
memory_usage_data = await prometheus_client.query(memory_usage_query) memory_usage_data = await prometheus_client.query(memory_usage_query)
# Get resource requests and limits # Get resource requests and limits using OpenShift-specific metrics
cpu_requests_query = f'kube_pod_container_resource_requests{{namespace="{namespace}",pod=~"{workload}-.*",resource="cpu"}}' cpu_requests_query = f'sum(kube_pod_container_resource_requests{{job="kube-state-metrics", cluster="", namespace="{namespace}", resource="cpu"}} * on(namespace,pod) group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{{cluster="", namespace="{namespace}", workload_type=~".+"}}) by (workload, workload_type)'
memory_requests_query = f'kube_pod_container_resource_requests{{namespace="{namespace}",pod=~"{workload}-.*",resource="memory"}}' memory_requests_query = f'sum(kube_pod_container_resource_requests{{job="kube-state-metrics", cluster="", namespace="{namespace}", resource="memory"}} * on(namespace,pod) group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{{cluster="", namespace="{namespace}", workload_type=~".+"}}) by (workload, workload_type)'
cpu_requests_data = await prometheus_client.query(cpu_requests_query) cpu_requests_data = await prometheus_client.query(cpu_requests_query)
memory_requests_data = await prometheus_client.query(memory_requests_query) memory_requests_data = await prometheus_client.query(memory_requests_query)
cpu_limits_query = f'kube_pod_container_resource_limits{{namespace="{namespace}",pod=~"{workload}-.*",resource="cpu"}}' cpu_limits_query = f'sum(kube_pod_container_resource_limits{{job="kube-state-metrics", cluster="", namespace="{namespace}", resource="cpu"}} * on(namespace,pod) group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{{cluster="", namespace="{namespace}", workload_type=~".+"}}) by (workload, workload_type)'
memory_limits_query = f'kube_pod_container_resource_limits{{namespace="{namespace}",pod=~"{workload}-.*",resource="memory"}}' memory_limits_query = f'sum(kube_pod_container_resource_limits{{job="kube-state-metrics", cluster="", namespace="{namespace}", resource="memory"}} * on(namespace,pod) group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{{cluster="", namespace="{namespace}", workload_type=~".+"}}) by (workload, workload_type)'
cpu_limits_data = await prometheus_client.query(cpu_limits_query) cpu_limits_data = await prometheus_client.query(cpu_limits_query)
memory_limits_data = await prometheus_client.query(memory_limits_query) memory_limits_data = await prometheus_client.query(memory_limits_query)
# Get cluster total resources # Get cluster total resources
cluster_cpu_query = 'sum(kube_node_status_allocatable{resource="cpu"})' cluster_cpu_query = 'sum(kube_node_status_allocatable{resource="cpu"})'
@@ -526,46 +526,70 @@ async def get_workload_historical_metrics(
cluster_cpu_data = await prometheus_client.query(cluster_cpu_query) cluster_cpu_data = await prometheus_client.query(cluster_cpu_query)
cluster_memory_data = await prometheus_client.query(cluster_memory_query) cluster_memory_data = await prometheus_client.query(cluster_memory_query)
# Extract values # Extract values from OpenShift-specific queries
cpu_usage = 0 cpu_usage = 0
memory_usage = 0 memory_usage = 0
cpu_requests = 0 cpu_requests = 0
memory_requests = 0 memory_requests = 0
cpu_limits = 0 cpu_limits = 0
memory_limits = 0 memory_limits = 0
cluster_cpu_total = 0 cluster_cpu_total = 0
cluster_memory_total = 0 cluster_memory_total = 0
# Check if we got any data from Prometheus
prometheus_available = False
# Check if we got any data from Prometheus # Extract CPU usage from workload-specific query
prometheus_available = False if cpu_usage_data.get("status") == "success" and cpu_usage_data.get("data", {}).get("result"):
for result in cpu_usage_data["data"]["result"]:
if cpu_usage_data.get("status") == "success" and cpu_usage_data.get("data", {}).get("result"): if result.get("metric", {}).get("workload") == workload:
cpu_usage = float(cpu_usage_data["data"]["result"][0]["value"][1]) cpu_usage = float(result["value"][1])
break
if memory_usage_data.get("status") == "success" and memory_usage_data.get("data", {}).get("result"):
memory_usage = float(memory_usage_data["data"]["result"][0]["value"][1]) # Extract Memory usage from workload-specific query
if memory_usage_data.get("status") == "success" and memory_usage_data.get("data", {}).get("result"):
if cpu_requests_data.get("status") == "success" and cpu_requests_data.get("data", {}).get("result"): for result in memory_usage_data["data"]["result"]:
cpu_requests = float(cpu_requests_data["data"]["result"][0]["value"][1]) if result.get("metric", {}).get("workload") == workload:
memory_usage = float(result["value"][1])
if memory_requests_data.get("status") == "success" and memory_requests_data.get("data", {}).get("result"): break
memory_requests = float(memory_requests_data["data"]["result"][0]["value"][1])
# Extract CPU requests from workload-specific query
if cpu_limits_data.get("status") == "success" and cpu_limits_data.get("data", {}).get("result"): if cpu_requests_data.get("status") == "success" and cpu_requests_data.get("data", {}).get("result"):
cpu_limits = float(cpu_limits_data["data"]["result"][0]["value"][1]) for result in cpu_requests_data["data"]["result"]:
if result.get("metric", {}).get("workload") == workload:
if memory_limits_data.get("status") == "success" and memory_limits_data.get("data", {}).get("result"): cpu_requests = float(result["value"][1])
memory_limits = float(memory_limits_data["data"]["result"][0]["value"][1]) break
if cluster_cpu_data.get("status") == "success" and cluster_cpu_data.get("data", {}).get("result"): # Extract Memory requests from workload-specific query
cluster_cpu_total = float(cluster_cpu_data["data"]["result"][0]["value"][1]) if memory_requests_data.get("status") == "success" and memory_requests_data.get("data", {}).get("result"):
for result in memory_requests_data["data"]["result"]:
if cluster_memory_data.get("status") == "success" and cluster_memory_data.get("data", {}).get("result"): if result.get("metric", {}).get("workload") == workload:
cluster_memory_total = float(cluster_memory_data["data"]["result"][0]["value"][1]) memory_requests = float(result["value"][1])
break
# Check if Prometheus is available (any non-zero values)
if cluster_cpu_total > 0 or cluster_memory_total > 0: # Extract CPU limits from workload-specific query
prometheus_available = True if cpu_limits_data.get("status") == "success" and cpu_limits_data.get("data", {}).get("result"):
for result in cpu_limits_data["data"]["result"]:
if result.get("metric", {}).get("workload") == workload:
cpu_limits = float(result["value"][1])
break
# Extract Memory limits from workload-specific query
if memory_limits_data.get("status") == "success" and memory_limits_data.get("data", {}).get("result"):
for result in memory_limits_data["data"]["result"]:
if result.get("metric", {}).get("workload") == workload:
memory_limits = float(result["value"][1])
break
if cluster_cpu_data.get("status") == "success" and cluster_cpu_data.get("data", {}).get("result"):
cluster_cpu_total = float(cluster_cpu_data["data"]["result"][0]["value"][1])
if cluster_memory_data.get("status") == "success" and cluster_memory_data.get("data", {}).get("result"):
cluster_memory_total = float(cluster_memory_data["data"]["result"][0]["value"][1])
# Check if Prometheus is available (any non-zero values)
if cluster_cpu_total > 0 or cluster_memory_total > 0:
prometheus_available = True
# If Prometheus is not available, provide simulated data for demonstration # If Prometheus is not available, provide simulated data for demonstration
if not prometheus_available: if not prometheus_available: