Flexible Engine - Advanced guide


Updated on :

Step 7 – Configuring the monitoring (CES)

In this step, you will use a Cloud Eye Service (CES) to see how the ECS, AS, and ELB services respond to overloading the created infrastructure. CES is a service that oversees resources on the Flexible Engine. It provides several alarms, reporting modes and sends notifications quickly, so you can ensure that the services are working.

In this step, you will have access to monitoring to explore the monitoring dashboard, create new monitoring panels and create alerts that send you notifications.

 

7.1.    View metrics by object

To access CES, choose the Cloud Eye service from the console.

 

 

The Dashboard displays the number of metrics available for the services deployed on your infrastructure. For each type of object deployed (in our case: AS, ECS, ELB, EVS, RDS and VPC), several metrics are available by default.

For example, to see the metrics linked to our ELB, click Cloud Service Monitoring then on ELB. You can choose to display metrics on a predefined time window, from  1H (Last 1 hour) to last 30 days, or select the time interval (Select Time Range) precisely.

 

7.2.    Creating a custom view

To create a new view of the monitoring panel, return to the Dashboard>Monitoring Panel and click Create, and give it the name: advanced.

Click on Add Graph that appears at the right of the page to add a metric of monitoring.

Fill in the following information :

  • Title: AS_Instance_Qty
  • Monitoring Metric:
    • Service: Auto Scaling
    • Dimension: AS Group
    • Monitored Object: Select as-group-advanced
    • Metric: Instance Qty

 

Do the same for CPU usage of ECS ecs-web :

  • Title: ECS_CPU_Usage
  • Monitoring Metric:
    • Service: Elastic Cloud Server
    • Dimension: ECS
    • Monitored Object: Select: ecs-web
    • Metric: CPU Usage

 

  • Title: ECS_CPU_Usage 2
  • Monitoring Metric:
    • Service: Elastic Cloud Server
    • Dimension: ECS
    • Monitored Object: Select: The ECS created by the AS service
    • Metric: CPU Usage

 

Do the same for packets entering the ECS ecs-web:

  • Title: ECS_Incoming_Rate
  • Monitoring Metric:
    • Service: Elastic Cloud Server
    • Dimension: ECS
    • Monitored Object: Select: ecs-web
    • Metric: Outband Incoming Rate

 

Do the same for packets entering the ELB:

  • Title: ELB_Incoming_Rate
  • Monitoring Metric:
    • Service: Elastic Load Balancer
    • Dimension: Load Balance
    • Monitored Object: Select: elb-advanced
    • Metric: Incoming Packets

 

At the end, you get a metric table equivalent to the one shown below:

 

 

7.3.    Creating an alarm with notification

We will now create an alarm to be notified when a metric is exceeded. We will take the case of  a CPU value higher than 10% for an ECS.

 

Click Alarm, then click the Add Alarm Rule button, and enter the following information:

  • Name: alarm-advanced
  • Description: Alarm related to CPU use of an ECS
  • Service: Choose Elastic Cloud Server
  • Dimension: ECS
  • Monitoring Scope: Specific ressources,  select Ecsweb
  • Method: Create manually
  • Metric: CPU Usage
  • Occurrences: 1
  • Monitoring Interval: 5 minutes
  • Threshold: > 10%
  • Alarm Notification: Yes
  • Notification Object: select the Topic SMN created in step 8
  • Trigger Condition: General alarm, Cleared alarm,

 

Click the button OK.