As an example, we will set up an alert for cluster nodes that are running out of disk space in it’s root file system.

  • Log into the CycleCloud web application, and click on the Message icon at the top right corner

  • Click on the “Configure Alerting” link at the top right of the page

  • Click on the “+” icon at the bottom left corner of the panel to add another alert

  • Create a new alert:

    • Provide a name for this rule: “Root FS Full”
    • Select a Priority level for this rule — in this example, “Error”
    • Query Ganglia host data by using the drop down in the “Query from…” line.
    • Construct a SQL-like query


    a. Choosing the attributes to query

    SELECT Cluster, Name, disk_free_percent_rootfs

    b. Construct the query clause, in this example, hosts where the free space in the root FS is less than 5%

    WHERE disk_free_percent_rootfs < 5

    c. Fill in a useful subject line for the alert message

    ERROR: Host root file system is full

    d. Construct the message body

    The following {%= size(Results) %} instances(s) have used more than 95% of their root partition:<br>
    
    {% for host in Results %}
        Host: {%= host.Name %}<br>
        Cluster: {%= host.Cluster %}<br>
        Root FS % Free: {%= host.disk_free_percent_rootfs  %}<br>
    {% endfor %}
    1. Click on the “Apply” button at the bottom right hand corner to save the alert

Email Notifications

  • In the alert creation page, select “Email Settings”
  • Fill in the “Email Settings” pop-up. Use the Test button to test if the SMTP server settings are correct.
  • Click “Save” to save the email settings.