#11 Add backup metrics and alaram

.gitignore

@@ -5,7 +5,8 @@ target
 .idea
 .vscode
 *.iml
-
+.settings/org.eclipse.m2e.core.prefs
+.settings/org.eclipse.core.resources.prefs
 # CDK asset staging directory
 .cdk.staging
 cdk.out
@@ -26,4 +27,4 @@ pom.xml.versionsBackup
 *.bak
 *.orig
 *.old
-*.md.html
+*.md.html

README.md

@@ -35,14 +35,25 @@ In this repository we define the dashboard by code using [AWS CDK](https://aws.a
 You can use this as a starting point for your own dashboards. In contrast to changing the dashboard using the AWS Console that allows you for example to version your dashboards in Git, add a review process or simply copy the same dashboard for multiple Exasol deployments. For that you need to go through the following steps:
 
 * Checkout (or fork) this repository
-* Adapt the dashboard to your needs (modify `src/main/java/com/exasol/cloudwatchexampledashboard/CloudwatchDashboardExamplesStack.java`)
+* Adapt the dashboard to your needs (modify [`src/main/java/com/exasol/cloudwatchexampledashboard/CloudwatchDashboardExamplesStack.java`](src/main/java/com/exasol/cloudwatchexampledashboard/CloudwatchDashboardExamplesStack.java))
 * [Install the AWS CDK](https://docs.aws.amazon.com/cdk/latest/guide/getting_started.html#getting_started_install)
 * Deploy your dashboard using
     ```shell
      cdk deploy --parameters deploymentName=<YOUR DEPLOYMENT NAME>
     ```
   (Don't forget to replace `<YOUR DEPLOYMENT NAME>` with the value you configured during the CloudWatch-adapter setup)
 
+## Injecting Metrics Values for Testing
+
+To test your dashboard and alarms you can manually inject metrics values by executing commands like these:
+
+```sh
+aws cloudwatch put-metric-data --namespace "Exasol" --dimensions "Cluster Name=MAIN,Deployment=<YOUR DEPLOYMENT NAME>" --unit Seconds --value 30 --metric-name "BACKUP_DURATION"
+aws cloudwatch put-metric-data --namespace "Exasol" --dimensions "Cluster Name=MAIN,Deployment=<YOUR DEPLOYMENT NAME>" --unit Count --value 1 --metric-name "BACKUP_START"      
+aws cloudwatch put-metric-data --namespace "Exasol" --dimensions "Cluster Name=MAIN,Deployment=<YOUR DEPLOYMENT NAME>" --unit Count --value 1 --metric-name "BACKUP_END"  
+aws cloudwatch put-metric-data --namespace "Exasol" --dimensions "Cluster Name=MAIN,Deployment=<YOUR DEPLOYMENT NAME>" --unit Count --value 1 --metric-name "BACKUP_BACKUP_ABORTED"
+```
+
 ## Additional Information
 
 * [Changelog](doc/changes/changelog.md)

src/main/java/com/exasol/cloudwatchexampledashboard/CloudwatchDashboardExamplesStack.java

@@ -10,12 +10,13 @@
  * This stack creates an CloudWatch dashboard with graphs for the metrics reported by the exasol cloudwatch-adapter.
  */
 public class CloudwatchDashboardExamplesStack extends Stack {
+    private static final String NAMESPACE = "Exasol";
     private final CfnParameter deploymentName;
     private final Map<String, String> dimensions;
 
     /**
      * Create a new instance of {@link CloudwatchDashboardExamplesStack}.
-     * 
+     *
      * @param scope parent scope
      * @param id    stack id
      */
@@ -34,14 +35,19 @@ public CloudwatchDashboardExamplesStack(final Construct scope, final String id,
         super(scope, id, props);
         this.deploymentName = CfnParameter.Builder.create(this, "deploymentName").type("String")
                 .description("Deployment name matching the one configured in the cloud watch adapter.").build();
-        this.dimensions = Map.of("Cluster Name", "MASTER", "Deployment", this.deploymentName.getValueAsString());
+        this.dimensions = Map.of("Cluster Name", "MAIN", "Deployment", this.deploymentName.getValueAsString());
 
         final Dashboard dashboard = Dashboard.Builder.create(this, "Exasol Dashboard").build();
 
         final Metric queriesMetric = getExasolMetricBuilder().metricName("QUERIES")
                 .label("Parallel running queries (5 min MAX)").statistic("Maximum").build();
         final Metric usersMetric = getExasolMetricBuilder().metricName("USERS").label("Users (5 min MAX)")
                 .statistic("Maximum").build();
+
+        final Metric eventBackupStart = eventMetric("BACKUP_START", "Backup started");
+        final Metric eventBackupEnd = eventMetric("BACKUP_END", "Backup finished successfully");
+        final Metric eventBackupAborted = eventMetric("BACKUP_ABORTED", "Backup failed or aborted");
+
         dashboard.addWidgets(//
                 cpuWidget(), //
                 tempDbRamWidget(), //
@@ -50,9 +56,12 @@ public CloudwatchDashboardExamplesStack(final Construct scope, final String id,
                 usageWidget(queriesMetric, usersMetric), //
                 recommendedDbRamSizeWidget(), //
                 currentDbSizeWidget(), //
-                currentQueriesAndUsersWidget(queriesMetric, usersMetric));
+                currentQueriesAndUsersWidget(queriesMetric, usersMetric), //
+                backupEvents(List.of(eventBackupStart, eventBackupEnd, eventBackupAborted)));
 
         addTempdbRamAlarm();
+        addBackupDidNotSucceedAlarms(eventBackupEnd);
+        addBackupFailedAlarms(eventBackupAborted);
     }
 
     private void addTempdbRamAlarm() {
@@ -108,7 +117,7 @@ private GraphWidget recommendedDbRamSizeWidget() {
 
     private SingleValueWidget currentQueriesAndUsersWidget(final Metric queriesMetric, final Metric usersMetric) {
         return SingleValueWidget.Builder.create().title("Current Queries and Users").setPeriodToTimeRange(false)
-                .metrics(List.of(queriesMetric, usersMetric)).width(6).build();
+                .metrics(List.of(queriesMetric, usersMetric)).width(6).height(6).build();
     }
 
     private SingleValueWidget currentDbSizeWidget() {
@@ -131,10 +140,36 @@ private SingleValueWidget currentDbSizeWidget() {
                 getExasolMetricBuilder().metricName("OBJECT_COUNT").label("Object count").period(Duration.hours(1))
                         .build(), //
                 getExasolMetricBuilder().metricName("NODES").label("Node count").period(Duration.days(1)).build() //
-        )).setPeriodToTimeRange(false).width(12).height(6).build();
+        )).setPeriodToTimeRange(false).width(12).height(8).build();
+    }
+
+    private GraphWidget backupEvents(final List<? extends IMetric> events) {
+        return GraphWidget.Builder.create().title("Backup Events").left(events)
+                .right(List.of(getExasolMetricBuilder().metricName("BACKUP_DURATION").label("Backup duration")
+                        .statistic("Avg").period(Duration.minutes(1)).build()))
+                .build();
+    }
+
+    private Metric eventMetric(final String metricName, final String label) {
+        return getExasolMetricBuilder().metricName(metricName).label(label).statistic("SampleCount")
+                .period(Duration.minutes(1)).build();
+    }
+
+    private void addBackupDidNotSucceedAlarms(final Metric eventBackupEnd) {
+        Alarm.Builder.create(this, "backup_not_suceeded").metric(eventBackupEnd).threshold(1)
+                .comparisonOperator(ComparisonOperator.LESS_THAN_THRESHOLD).evaluationPeriods(1).datapointsToAlarm(1)
+                .alarmName("Backup did not succeed").alarmDescription("Backup did not succeed for more than one minute")
+                .treatMissingData(TreatMissingData.BREACHING).actionsEnabled(false).build();
+    }
+
+    private void addBackupFailedAlarms(final Metric eventBackupAborted) {
+        Alarm.Builder.create(this, "backup_failed").metric(eventBackupAborted).threshold(0)
+                .comparisonOperator(ComparisonOperator.GREATER_THAN_THRESHOLD).evaluationPeriods(1).datapointsToAlarm(1)
+                .alarmName("Backup failed").alarmDescription("Backup failed within one minute")
+                .treatMissingData(TreatMissingData.NOT_BREACHING).actionsEnabled(false).build();
     }
 
     private Metric.Builder getExasolMetricBuilder() {
-        return Metric.Builder.create().namespace("Exasol").dimensions(this.dimensions);
+        return Metric.Builder.create().namespace(NAMESPACE).dimensionsMap(this.dimensions);
     }
 }