Fix 'ceph_disk_occupation' query expressions

Need to address changes in 'ceph_disk_occupation' metric labels. What is the change in 'ceph_disk_occupation' metric? 'ceph_disk_occupation' result no longer has 'exported_instance' label, instead it has 'instance' label. What is the issue we are facing because of it? We are hitting 'PrometheusRuleFailures' due to this new label changes in our alerts / rules, where this metric is used. Second issue is that we are not seeing any results for some of the query expressions. What is the solution? Update the query expressions, change 'exported_instance' to 'instance'. Any 'label_replace' action which changes 'exported_instance' label to 'instance' label is no longer required (as the 'instance' label is directly available now) Signed-off-by: Arun Kumar Mohan <amohan@redhat.com>
red-hat-storage · Dec 13, 2024 · a81e357 · a81e357
1 parent 1d6235c
commit a81e357
Show file tree

Hide file tree

Showing 3 changed files with 16 additions and 40 deletions.
diff --git a/controllers/storagecluster/prometheus/localcephrules.yaml b/controllers/storagecluster/prometheus/localcephrules.yaml
@@ -11,10 +11,10 @@ spec:
   - name: ceph.rules
     rules:
     - expr: |
-        kube_node_status_condition{condition="Ready",job="kube-state-metrics",status="true"} * on (node) group_right() max by (node, namespace) (label_replace(ceph_disk_occupation{job="rook-ceph-mgr"},"node","$1","exported_instance","(.*)"))
+        kube_node_status_condition{condition="Ready",job="kube-state-metrics",status="true"} * on (node) group_right() max by (node, namespace) (label_replace(ceph_disk_occupation{job="rook-ceph-mgr"},"node","$1","instance","(.*)"))
       record: cluster:ceph_node_down:join_kube
     - expr: |
-        avg by (namespace) (topk by (ceph_daemon, namespace) (1, label_replace(label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "instance", "$1", "exported_instance", "(.*)"), "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_right(ceph_daemon, namespace) topk by (instance, device, namespace) (1,(irate(node_disk_read_time_seconds_total[1m]) + irate(node_disk_write_time_seconds_total[1m]) / (clamp_min(irate(node_disk_reads_completed_total[1m]), 1) + irate(node_disk_writes_completed_total[1m])))))
+        avg by (namespace) (topk by (ceph_daemon, namespace) (1, label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_right(ceph_daemon, namespace) topk by (instance, device, namespace) (1,(irate(node_disk_read_time_seconds_total[1m]) + irate(node_disk_write_time_seconds_total[1m]) / (clamp_min(irate(node_disk_reads_completed_total[1m]), 1) + irate(node_disk_writes_completed_total[1m])))))
       record: cluster:ceph_disk_latency:join_ceph_node_disk_irate1m
   - name: telemeter.rules
     rules:
@@ -171,7 +171,7 @@ spec:
         storage_type: ceph
         runbook_url: https://github.com/openshift/runbooks/blob/master/alerts/openshift-container-storage-operator/CephOSDDiskNotResponding.md
       expr: |
-        label_replace((ceph_osd_in == 1 and ceph_osd_up == 0),"disk","$1","ceph_daemon","osd.(.*)") + on(ceph_daemon, namespace, managedBy) group_left(host, device) label_replace(ceph_disk_occupation{job=~"rook-ceph-mgr|rook-ceph-mgr-external"},"host","$1","exported_instance","(.*)")
+        label_replace((ceph_osd_in == 1 and ceph_osd_up == 0),"disk","$1","ceph_daemon","osd.(.*)") + on(ceph_daemon, namespace, managedBy) group_left(host, device) label_replace(ceph_disk_occupation{job=~"rook-ceph-mgr|rook-ceph-mgr-external"},"host","$1","instance","(.*)")
       for: 15m
       labels:
         severity: critical
@@ -183,7 +183,7 @@ spec:
         storage_type: ceph
         runbook_url: https://github.com/openshift/runbooks/blob/master/alerts/openshift-container-storage-operator/CephOSDDiskUnavailable.md
       expr: |
-        label_replace((ceph_osd_in == 0 and ceph_osd_up == 0),"disk","$1","ceph_daemon","osd.(.*)") + on(ceph_daemon, namespace, managedBy) group_left(host, device) label_replace(ceph_disk_occupation{job=~"rook-ceph-mgr|rook-ceph-mgr-external"},"host","$1","exported_instance","(.*)")
+        label_replace((ceph_osd_in == 0 and ceph_osd_up == 0),"disk","$1","ceph_daemon","osd.(.*)") + on(ceph_daemon, namespace, managedBy) group_left(host, device) label_replace(ceph_disk_occupation{job=~"rook-ceph-mgr|rook-ceph-mgr-external"},"host","$1","instance","(.*)")
       for: 1m
       labels:
         severity: critical

diff --git a/metrics/deploy/prometheus-ocs-rules-external.yaml b/metrics/deploy/prometheus-ocs-rules-external.yaml
@@ -10,17 +10,11 @@ spec:
   groups:
   - name: ocs_performance.rules
     rules:
-    - expr: "sum by (namespace, managedBy) (\n    topk by (ceph_daemon) (1, label_replace(label_replace(ceph_disk_occupation{job=\"rook-ceph-mgr\"},
-        \"instance\", \"$1\", \"exported_instance\", \"(.*)\"), \"device\", \"$1\",
-        \"device\", \"/dev/(.*)\")) \n  * on(instance, device) group_left topk by
-        (instance,device) \n    (1,\n      (\n        rate(node_disk_read_time_seconds_total[1m])
-        / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1))\n      )\n    )\n)\n"
+    - expr: |
+        sum by (namespace, managedBy) (topk by (ceph_daemon) (1, label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_left topk by (instance,device) (1, (rate(node_disk_read_time_seconds_total[1m]) / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1)))))
       record: cluster:ceph_disk_latency_read:join_ceph_node_disk_rate1m
-    - expr: "sum by (namespace, managedBy) (\n    topk by (ceph_daemon) (1, label_replace(label_replace(ceph_disk_occupation{job=\"rook-ceph-mgr\"},
-        \"instance\", \"$1\", \"exported_instance\", \"(.*)\"), \"device\", \"$1\",
-        \"device\", \"/dev/(.*)\")) \n  * on(instance, device) group_left topk by
-        (instance,device) \n    (1,\n      (\n        rate(node_disk_write_time_seconds_total[1m])
-        / (clamp_min(rate(node_disk_writes_completed_total[1m]), 1))\n      )\n    )\n)\n"
+    - expr: |
+        sum by (namespace, managedBy) (topk by (ceph_daemon) (1, label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_left topk by (instance,device) (1, (rate(node_disk_write_time_seconds_total[1m]) / (clamp_min(rate(node_disk_writes_completed_total[1m]), 1)))))
       record: cluster:ceph_disk_latency_write:join_ceph_node_disk_rate1m
   - name: ODF_standardized_metrics.rules
     rules:
@@ -54,14 +48,8 @@ spec:
         system_type: OCS
         system_vendor: Red Hat
       record: odf_system_throughput_total_bytes
-    - expr: "sum by (namespace, managedBy, job, service)\n(\n  topk by (ceph_daemon)
-        (1, label_replace(label_replace(ceph_disk_occupation{job=\"rook-ceph-mgr\"},
-        \"instance\", \"$1\", \"exported_instance\", \"(.*)\"), \"device\", \"$1\",
-        \"device\", \"/dev/(.*)\")) \n  * on(instance, device) group_left() topk by
-        (instance,device) \n  (1,\n    (\n      (  \n          rate(node_disk_read_time_seconds_total[1m])
-        / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1))\n      ) +\n      (\n
-        \         rate(node_disk_write_time_seconds_total[1m]) / (clamp_min(rate(node_disk_writes_completed_total[1m]),
-        1))\n      )\n    )\n  )\n)\n"
+    - expr: |
+        sum by (namespace, managedBy, job, service) (topk by (ceph_daemon) (1, label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_left() topk by (instance,device) (1, ((rate(node_disk_read_time_seconds_total[1m]) / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1))) + (rate(node_disk_write_time_seconds_total[1m]) / (clamp_min(rate(node_disk_writes_completed_total[1m]), 1))))))
       labels:
         system_type: OCS
         system_vendor: Red Hat

diff --git a/metrics/deploy/prometheus-ocs-rules.yaml b/metrics/deploy/prometheus-ocs-rules.yaml
@@ -10,17 +10,11 @@ spec:
   groups:
   - name: ocs_performance.rules
     rules:
-    - expr: "sum by (namespace, managedBy) (\n    topk by (ceph_daemon) (1, label_replace(label_replace(ceph_disk_occupation{job=\"rook-ceph-mgr\"},
-        \"instance\", \"$1\", \"exported_instance\", \"(.*)\"), \"device\", \"$1\",
-        \"device\", \"/dev/(.*)\")) \n  * on(instance, device) group_left topk by
-        (instance,device) \n    (1,\n      (\n        rate(node_disk_read_time_seconds_total[1m])
-        / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1))\n      )\n    )\n)\n"
+    - expr: |
+        sum by (namespace, managedBy) (topk by (ceph_daemon) (1, label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_left topk by (instance,device) (1, (rate(node_disk_read_time_seconds_total[1m]) / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1)))))
       record: cluster:ceph_disk_latency_read:join_ceph_node_disk_rate1m
-    - expr: "sum by (namespace, managedBy) (\n    topk by (ceph_daemon) (1, label_replace(label_replace(ceph_disk_occupation{job=\"rook-ceph-mgr\"},
-        \"instance\", \"$1\", \"exported_instance\", \"(.*)\"), \"device\", \"$1\",
-        \"device\", \"/dev/(.*)\")) \n  * on(instance, device) group_left topk by
-        (instance,device) \n    (1,\n      (\n        rate(node_disk_write_time_seconds_total[1m])
-        / (clamp_min(rate(node_disk_writes_completed_total[1m]), 1))\n      )\n    )\n)\n"
+    - expr: |
+        sum by (namespace, managedBy) (topk by (ceph_daemon) (1, label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_left topk by (instance,device) (1, (rate(node_disk_write_time_seconds_total[1m]) / (clamp_min(rate(node_disk_writes_completed_total[1m]), 1)))))
       record: cluster:ceph_disk_latency_write:join_ceph_node_disk_rate1m
   - name: ODF_standardized_metrics.rules
     rules:
@@ -54,14 +48,8 @@ spec:
         system_type: OCS
         system_vendor: Red Hat
       record: odf_system_throughput_total_bytes
-    - expr: "sum by (namespace, managedBy, job, service)\n(\n  topk by (ceph_daemon)
-        (1, label_replace(label_replace(ceph_disk_occupation{job=\"rook-ceph-mgr\"},
-        \"instance\", \"$1\", \"exported_instance\", \"(.*)\"), \"device\", \"$1\",
-        \"device\", \"/dev/(.*)\")) \n  * on(instance, device) group_left() topk by
-        (instance,device) \n  (1,\n    (\n      (  \n          rate(node_disk_read_time_seconds_total[1m])
-        / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1))\n      ) +\n      (\n
-        \         rate(node_disk_write_time_seconds_total[1m]) / (clamp_min(rate(node_disk_writes_completed_total[1m]),
-        1))\n      )\n    )\n  )\n)\n"
+    - expr: |
+        sum by (namespace, managedBy, job, service) (topk by (ceph_daemon) (1, label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_left() topk by (instance,device) (1, ((rate(node_disk_read_time_seconds_total[1m]) / (clamp_min(rate(node_disk_reads_completed_total[1m]), 1))) + (rate(node_disk_write_time_seconds_total[1m]) / (clamp_min(rate(node_disk_writes_completed_total[1m]), 1))))))
       labels:
         system_type: OCS
         system_vendor: Red Hat