Managing and monitoring clusters

You can manage monitor your clusters from the Cloudbreak UI. To do that, click on the tile representing the cluster that you want to access. The actions available for your cluster are listed in the top right corner:

Tips

  • To add or remove nodes from your cluster click ACTIONS>Resize.
  • To synchronize your cluster with the cloud provider account click ACTIONS>Sync.
  • To temporarily stop your cluster click STOP.
  • To terminate your cluster click TERMINATE.

Retry a cluster

When stack provisioning or cluster creation failure occurs, the "retry" option allows you to resume the process from the last failed step.

In some cases the cause of a failed stack provisioning or cluster creation can be eliminated by simply retrying the process. For example, in case of a temporary network outage, a retry should be successful. In other cases, a manual modification is required before a retry can succeed. For example, if you are using a custom image but some configuration is missing, causing the process to fail, you must log in to the machine and fix the issue; Only after that you can retry the the process.

Only failed stack or cluster creations can be retried. A retry can be initiated any number of times on a failed creation process.

To retry provisioning a failed stack or cluster, follow these steps.

Steps

  1. Browse to the cluster details.

  2. Click Actions and select Retry.

    Only failed stack or cluster creations can be retried, so the option is only available in these two cases.

  3. Click Yes to confirm.

    The operation continues from the last failed step.

Resize a cluster

To resize a cluster, follow these steps.

Cluster resizing is not supported for HDF clusters.
To configure automated cluster scaling, refer to Configure autoscaling.

Steps

  1. Browse to the cluster details.

  2. Click Actions and select Resize. The cluster resize dialog is displayed.

  3. Using the +/- controls, adjust the number of nodes for a chosen host group.

    You can only modify one host group at a time.
    It is not possible to resize the Ambari server host group.

  4. Click Yes to confirm the scale-up/scale-down.

    While nodes are being added or removed, cluster status changes to "Update In Progress". Once the operation has completed, cluster status changes back to "Running".

Synchronize a cluster

Use the sync option if you:

To synchronize your cluster with the cloud provider, follow these steps.

Steps

  1. Browse to the cluster details.

  2. Click Actions and select Sync.

  3. Click Yes to confirm.

    Your cluster infrastructure is synchronized based on changes on the cloud provider. The updates are written to the "Event History".

Stop a cluster

Cloudbreak supports stopping and restarting clusters. To stop and restart a cluster managed by Cloudbreak, use the options available from the Cloudbreak UI.

Steps

  1. Browse to the cluster details.

  2. Click Stop to stop a currently running cluster.

  3. Click Yes to confirm.

  4. Your cluster status changes to "Stopping in progress" and then to "Stopped". Once stopping the infrastructure has completed, you will see a Start option to restart your cluster.

When a cluster is in the "stopped" state, you are not charged for the VMs, but you are charged for external storage.

Restart a cluster

If your cluster is in the "Stopped" state, you can restart the cluster by follow these steps.

Steps

  1. click Start. This option is only available when the cluster has been stopped.

  2. Click Yes to confirm.

    Your cluster status changes to "Start in progress" and then to "Running".

Terminate a cluster

To terminate a cluster managed by Cloudbreak, use the option available from the Cloudbreak UI.

Steps

  1. Browse to the cluster details.

  2. Click Terminate.

  3. Click Yes to confirm.

    All cluster-related resources will be deleted, unless the resources (such as networks and subnets) existed prior to cluster creation or are used by other VMs in which case they will be preserved.

Force terminate a cluster

Cluster deletion may fail if Cloudbreak is unable to delete one or more of the cloud resources that were part of your cluster infrastructure. In such as case, you can use the Terminate > Force terminate option to remove the cluster entry from the Cloudbreak web UI, but you must also check your cloud provider account to see if there are any resources that must be deleted manually.

Steps

  1. Browse to the cluster details.

  2. Click Terminate.

  3. Check Force terminate.

  4. Click Yes to confirm.

    When terminating a cluster with Kerberos enabled, you have an option to disable Kerberos prior to cluster termination. This option removes any cluster-related principals from the KDC.

  5. This deletes the cluster tile from the UI.

  6. Log in to your cloud provider account and manually delete any resources that failed to be deleted.

View cluster history

From the navigation menu in the Cloudbreak UI, you can access the History page that allows you to generate a report showing basic information related to the clusters that were running within the specified range of dates.

To generate a report, follow these steps.

Steps

  1. From the Cloudbreak UI navigation menu, select History.

  2. On the History page, select the range of dates and click Show History to generate a tabular report for the selected period.

History report content

Each entry in the report represents one cluster instance group. For each entry, the report includes the following information:

The AGGREGATE RUNNING TIME is the sum of the Running Times, adjusted for the selected time range.

To learn about how your cloud provider bills you for the VMs, refer to their documentation: