dashboard: add failover trigger panel

Add panel to "Cluster overview" section to dashboard templates.
Update consists of panel with following metrics:
- tnt_cartridge_failover_trigger_total

Part of #178

Author: DifferentialOrange

CHANGELOG.md

@@ -10,6 +10,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Dashboard title customization
 - Panels for transaction operations
 - Panels with net statistics per thread
+- Panel with failover trigger count
 
 ### Changed
 - Replace LuaJit deprecated metrics with new ones

dashboard/panels/cluster.libsonnet

@@ -424,7 +424,7 @@ local prometheus = grafana.prometheus;
     title=title,
     description=description,
     datasource=datasource,
-    panel_width=12,
+    panel_width=8,
     max=1,
     min=0,
   ).addValueMapping(
@@ -450,6 +450,35 @@ local prometheus = grafana.prometheus;
       .selectField('value').addConverter('last')
   ),
 
+  failovers_per_second(
+    title='Failovers triggered',
+    description=|||
+      Displays the count of failover triggers in a replicaset.
+      Graph shows average per second.
+
+      Panel works with `metrics >= 0.15.0`.
+    |||,
+    datasource_type=null,
+    datasource=null,
+    policy=null,
+    measurement=null,
+    job=null,
+    alias=null,
+  ):: common.default_graph(
+    title=title,
+    description=description,
+    datasource=datasource,
+    labelY1='failovers per second',
+    panel_width=8,
+  ).addTarget(common.default_rps_target(
+    datasource_type,
+    'tnt_cartridge_failover_trigger_total',
+    job,
+    policy,
+    measurement,
+    alias,
+  )),
+
   read_only_status(
     title='Tarantool instance status',
     description=|||
@@ -469,7 +498,7 @@ local prometheus = grafana.prometheus;
     title=title,
     description=description,
     datasource=datasource,
-    panel_width=12,
+    panel_width=8,
     max=1,
     min=0,
   ).addValueMapping(

dashboard/section.libsonnet

@@ -51,6 +51,14 @@ local tdg_tuples = import 'dashboard/panels/tdg/tuples.libsonnet';
       alias=alias,
     ),
 
+    cluster.failovers_per_second(
+      datasource_type=datasource_type,
+      datasource=datasource,
+      policy=policy,
+      measurement=measurement,
+      alias=alias,
+    ),
+
     cluster.read_only_status(
       datasource_type=datasource_type,
       datasource=datasource,
@@ -143,6 +151,13 @@ local tdg_tuples = import 'dashboard/panels/tdg/tuples.libsonnet';
       alias=alias,
     ),
 
+    cluster.failovers_per_second(
+      datasource_type=datasource_type,
+      datasource=datasource,
+      job=job,
+      alias=alias,
+    ),
+
     cluster.read_only_status(
       datasource_type=datasource_type,
       datasource=datasource,

supported_metrics.md

@@ -120,7 +120,7 @@ Based on [tarantool/metrics 0.16.0](https://github.com/tarantool/metrics/release
 - [x] **tnt_vinyl_scheduler_dump_total**: see *Tarantool vinyl statistics/Vinyl scheduler dump count rate* panel ([#133](https://github.com/tarantool/grafana-dashboard/issues/133))
 - [x] **tnt_cartridge_issues**: see *Cluster overview/Cartridge warning issues*, *Cluster overview/Cartridge critical issues* panels ([#55](https://github.com/tarantool/grafana-dashboard/pull/55))
 - **tnt_cartridge_cluster_issues**: unsupported (decided not to support: superseded by **tnt_cartridge_issues**)
-- [ ] **tnt_cartridge_failover_trigger_total**: ([#178](https://github.com/tarantool/grafana-dashboard/issues/178))
+- [x] **tnt_cartridge_failover_trigger_total**: see *Cluster overview/Failovers triggered* panel ([#178](https://github.com/tarantool/grafana-dashboard/issues/178))
 - [ ] **tnt_synchro_queue_owner**: ([#178](https://github.com/tarantool/grafana-dashboard/issues/178))
 - [ ] **tnt_synchro_queue_term**: ([#178](https://github.com/tarantool/grafana-dashboard/issues/178))
 - [ ] **tnt_synchro_queue_len**: ([#178](https://github.com/tarantool/grafana-dashboard/issues/178))