From fd5b77cadf0de374cf3ad01a8a526f4246b77bbd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Beno=C3=AEt=20Knecht?= Date: Mon, 28 Oct 2019 16:25:51 +0100 Subject: [PATCH 1/2] jsonnetfile.lock.json: Update node-mixin dependency Upgrade to prometheus/node_exporter@d574b4b, which includes a better metric for memory saturation. See prometheus/node_exporter#1524. --- jsonnetfile.lock.json | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/jsonnetfile.lock.json b/jsonnetfile.lock.json index b64313fe..0df0f590 100644 --- a/jsonnetfile.lock.json +++ b/jsonnetfile.lock.json @@ -97,7 +97,7 @@ "subdir": "docs/node-mixin" } }, - "version": "9f49fff79ef85fcebb69289622150e6d5346528b" + "version": "d574b4b41b72966a2ace4d6d81195081dfca8301" }, { "name": "promgrafonnet", From 20eb5b312aaff6cce2208001c4ca99b3c9e5ff70 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Beno=C3=AEt=20Knecht?= Date: Mon, 28 Oct 2019 16:35:12 +0100 Subject: [PATCH 2/2] manifests: Regenerate with new node-mixin dependency Generate manifests for fd5b77c. --- manifests/grafana-dashboardDefinitions.yaml | 10 +++++----- manifests/prometheus-rules.yaml | 8 ++------ 2 files changed, 7 insertions(+), 11 deletions(-) diff --git a/manifests/grafana-dashboardDefinitions.yaml b/manifests/grafana-dashboardDefinitions.yaml index 41841404..a9c7baba 100644 --- a/manifests/grafana-dashboardDefinitions.yaml +++ b/manifests/grafana-dashboardDefinitions.yaml @@ -20553,7 +20553,7 @@ items: "steppedLine": false, "targets": [ { - "expr": "instance:node_memory_swap_io_pages:rate1m{job=\"node-exporter\"}", + "expr": "instance:node_vmstat_pgmajfault:rate1m{job=\"node-exporter\"}", "format": "time_series", "intervalFactor": 2, "legendFormat": "{{instance}}", @@ -20566,7 +20566,7 @@ items: ], "timeFrom": null, "timeShift": null, - "title": "Memory Saturation (Swapped Pages)", + "title": "Memory Saturation (Major Page Faults)", "tooltip": { "shared": false, "sort": 0, @@ -21501,10 +21501,10 @@ items: "steppedLine": false, "targets": [ { - "expr": "instance:node_memory_swap_io_pages:rate1m{job=\"node-exporter\", instance=\"$instance\"}", + "expr": "instance:node_vmstat_pgmajfault:rate1m{job=\"node-exporter\", instance=\"$instance\"}", "format": "time_series", "intervalFactor": 2, - "legendFormat": "Swap IO", + "legendFormat": "Major page faults", "legendLink": null, "step": 10 } @@ -21514,7 +21514,7 @@ items: ], "timeFrom": null, "timeShift": null, - "title": "Memory Saturation (pages swapped per second)", + "title": "Memory Saturation (Major Page Faults)", "tooltip": { "shared": false, "sort": 0, diff --git a/manifests/prometheus-rules.yaml b/manifests/prometheus-rules.yaml index 88a02050..2b9e6d26 100644 --- a/manifests/prometheus-rules.yaml +++ b/manifests/prometheus-rules.yaml @@ -37,12 +37,8 @@ spec: ) record: instance:node_memory_utilisation:ratio - expr: | - ( - rate(node_vmstat_pgpgin{job="node-exporter"}[1m]) - + - rate(node_vmstat_pgpgout{job="node-exporter"}[1m]) - ) - record: instance:node_memory_swap_io_pages:rate1m + rate(node_vmstat_pgmajfault{job="node-exporter"}[1m]) + record: instance:node_vmstat_pgmajfault:rate1m - expr: | rate(node_disk_io_time_seconds_total{job="node-exporter", device=~"nvme.+|rbd.+|sd.+|vd.+|xvd.+|dm-.+"}[1m]) record: instance_device:node_disk_io_time_seconds:rate1m