dashboard: clarify expected Prometheus setup

Some panels dealing with cluster overview (supported only for
Prometheus) expect to have each instance as a separate pull target
since they rely on Prometheus target `up` healthcheck. They won't
work properly when something like TCM is used to provide metrics for
the whole cluster. This patch adds a note to clarify this.

Part of #224

Author: DifferentialOrange

dashboard/panels/cluster.libsonnet

@@ -15,7 +15,7 @@ local prometheus = grafana.prometheus;
   health_overview_table(
     cfg,
     title='Cluster status overview',
-    description=|||
+    description=common.prometheus_metrics_pull_note(|||
       Overview of Tarantool instances observed by Prometheus job.
 
       If instance row is *red*, it means Prometheus can't reach
@@ -28,7 +28,7 @@ local prometheus = grafana.prometheus;
 
       If Prometheus job filter is not specified, displays running instances
       and ignores unreachable instances (we have no specific source to fetch)
-    |||,
+    |||),
   ):: tablePanel.new(
     title=title,
     description=description,
@@ -163,13 +163,13 @@ local prometheus = grafana.prometheus;
   health_overview_stat(
     cfg,
     title='',
-    description=|||
+    description=common.prometheus_metrics_pull_note(|||
       Count of running Tarantool instances observed by Prometheus job.
       If Prometheus can't reach URI specified in targets
       or ran into error, instance is not counted.
 
       Instance alias filtering is disabled here.
-    |||,
+    |||),
   ):: if cfg.type == variable.datasource_type.prometheus then
     overview_stat(
       cfg,

dashboard/panels/common.libsonnet

@@ -123,4 +123,10 @@ local influxdb_query_filters(filters) = std.join(' AND ', std.map(
 
   remove_field(obj, key)::
     { [item.key]: item.value for item in std.objectKeysValuesAll(obj) if item.key != key },
+
+  prometheus_metrics_pull_note(description)::
+    std.join('\n', [description, |||
+      This panel is designed to work with Prometheus client
+      set up to have each Tarantool instance as a separate pull target.
+    |||]),
 }

tests/Prometheus/dashboard_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "$prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "$prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,

tests/Prometheus/dashboard_custom_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "$prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "$prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,

tests/Prometheus/dashboard_static_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "Prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "Prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,

tests/Prometheus/dashboard_static_custom_title_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "Prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "Prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,

tests/Prometheus/dashboard_static_with_instance_variable_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "Prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "Prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,

tests/Prometheus/dashboard_tdg_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "$prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "$prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,

tests/Prometheus/dashboard_tdg_static_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "Prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "Prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,

tests/Prometheus/dashboard_tdg_static_custom_title_compiled.json

@@ -69,7 +69,7 @@
             {
                "columns": [ ],
                "datasource": "Prometheus",
-               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n",
+               "description": "Overview of Tarantool instances observed by Prometheus job.\n\nIf instance row is *red*, it means Prometheus can't reach\nURI specified in targets or ran into error.\nIf instance row is *green*, it means instance is up and running and\nPrometheus is successfully extracting metrics from it.\n\"Uptime\" column shows time since instant start.\n\nInstance alias filtering is disabled here.\n\nIf Prometheus job filter is not specified, displays running instances\nand ignores unreachable instances (we have no specific source to fetch)\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 8,
                   "w": 12,
@@ -144,7 +144,7 @@
             },
             {
                "datasource": "Prometheus",
-               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n",
+               "description": "Count of running Tarantool instances observed by Prometheus job.\nIf Prometheus can't reach URI specified in targets\nor ran into error, instance is not counted.\n\nInstance alias filtering is disabled here.\n\nThis panel is designed to work with Prometheus client\nset up to have each Tarantool instance as a separate pull target.\n",
                "gridPos": {
                   "h": 3,
                   "w": 6,