Viewing Cluster Details

Node(s) Info Tab

The Node(s) Info tab of the Cluster Details screen appears as shown in the following image.

This tab contains the following button:

• Actions: Clicking this button opens a pull-down menu with the following options:
  • Download Public Endpoints: Downloads a list of public service endpoints, including the virtual node, IP address, service name, and port used to access each service. See Downloading Endpoints.
  • Add cluster service: Adds information about a cluster service. See Adding a Cluster Service.
  • Edit cluster service: Edits information about a cluster service. See Editing a Cluster Service.
  • Remove cluster service: Removes information about a cluster service. See Removing a Cluster Service.

The Node List table on this tab displays the following information for each of the nodes in the selected cluster:

• Public/Private Endpoints: This pull-down menu only appears when the floating IP address range of the cluster is non-routable, meaning that it cannot be accessed from the corporate network.
  • Selecting Private Endpoints lists the private virtual node IP addresses and service ports. For example, an SSH service might list 172.18.0.8:22 as the private IP address and port. You can also click the Private Endpoints button to download a list of the virtual nodes in the cluster along with IP address and port information in comma-delimited (.csv) format.
  • Selecting Public Endpoints lists the Gateway hostname and port that users who are not in the private network will use to access services. For example, the SSH service at 172.18.0.8:22 may be publicly accessible using yav-008.lab.HPE.com:10031. You can also click the Public Endpoints button to download a list of the virtual nodes in the cluster along with Gateway hostname and port information in comma-delimited (.csv) format.
• Name: Name of the virtual node. This name will appear in orange with the notation Node in debug mode if you created the cluster in debug mode (see Creating a New Cluster) and virtual node creation failed when creating or editing the cluster.
• Distribution: Distribution type used for the virtual node.
• Role: Role of the virtual node, such as Controller (controller), Worker (worker), or Edge (edge). If custom roles are defined for this cluster, those will appear here as well. If YARN High Availability is enabled, this column may also say arbiter or standby, as appropriate.
• Instance IP: IP address used to access the node.
• Services: List of currently running services. Clicking an item in this list opens the Web interface for the selected service (including third-party tools such as Cloudera Manager or Ambari, if included in the cluster), if your network includes a route from the computer you are using to view Container Platform to the Controller node.

  • If Public Endpoints is selected in the pull-down menu, this column lists a hostname:port combination that allows users to access those services when the virtual node IP addresses are not accessible via the corporate network. For example, in the above image, you can access the SSH service on the HPE-31.bdlocal virtual node using yav-008.lab.HPE.com:10042. The hostname refers to a Gateway set. Large deployments may require multiple Gateway sets. In this case, you may see different hostnames for different virtual nodes. See Network Requirements for additional information about routable and non-routable networks.

    

  • If Private Endpoints is selected in the pull-down menu, this column lists the non-routable (private) IP address and port number of the services on each virtual node.

    

ActionScript(s) Tab

The ActionScript(s) tab of the Cluster Details screen allows you to run and view ActionScripts on the current cluster. See About ActionScripts. This tab appears as shown in the following image.

This tab contains the following buttons:

• Run New ActionScript: Clicking the Run New ActionScript button opens the Run ActionScript(s) popup. See Creating a New ActionScript.
• Delete: Selecting one or more ActionScripts and then clicking the Delete ActionScripts button deletes the selected ActionScripts from the cluster. See Deleting an ActionScript.

The table on this screen contains the following information and functions:

• Action Name: Name of the ActionScript. Clicking a script name opens the ActionScript Source popup for the selected ActionScript, which displays the ActionScript contents. See Viewing ActionScript Contents.
• User: Name of the user who ran the ActionScript. The notation (As Root) appears if the user ran the ActionScript as the root user.
• Start Time: Date and time the ActionScript began.
• End Time: Date and time the ActionScript ended.
• Run Time: How long the action took to run, in MM:SS format.
• Action Status: Status of the ActionScript. See ActionScript Statuses.
• Actions: The following actions are available for each ActionScript:
  • Kill: While an ActionScript script is running, clicking the Kill button (X) immediately stops ActionScript execution.
  • Delete: Clicking the Delete icon (trash can) deletes the ActionScript. See Deleting an ActionScript.
  • Logs: Clicking the Logs icon (envelope) opens the Logs popup, which displays the ActionScript-related log entries from each virtual node in the cluster. See Viewing ActionScript Logs.

Services Status Tab

The Services Statustab of the Cluster Detailsscreen appears as shown in the following image.

The Services Status table has the following two groups of information:

• Application Services: Contains any services that are specific to that distribution.
• System Services: Contains the common HPE management services, such as the HPE Agent.
• Additional groups may also appear, depending on the cluster configuration. These are beyond the scope of this manual.

The status of a service can be either OK (green dot), CRITICAL (red dot), or DISABLED (intentionally not running; gray dot). Hovering the mouse over the status button opens a popup with additional information. In general:

• The Controller node for the cluster must not display any red dots. If the Controller node has one or more errors, then the cluster may not function properly.
• If all of the dots for a Worker node are red, then that node will not be able to participate in jobs that run on the cluster; however, the cluster as a whole should still perform correctly with a reduced capacity. This situation usually occurs because the host is powered off or has lost network connectivity.
• A Worker node with some red and some green dots may cause some jobs to fail, unless the errors are transient conditions caused by the host powering on or regaining network connectivity.

Please contact HPE Technical Support if a cluster that is reporting service errors meets all of the following criteria:

• The HPE Ezmeral Container Platform software is completely installed on the affected host(s).
• The host(s) is/are powered on.
• The host(s) has/have network connectivity.

See Common Cluster Errors for more information about resolving errors.

Cluster History Tab

The Cluster History tab of the Cluster Details screen appears as shown in the following image.

The Cluster History table on this tab displays the following information for available historical events:

• Operation: The operation that was performed on the cluster, such as Cluster Created or Reconfigure.
• User: Username of the person who performed the action.
• Details: Details of the action performed. For example, if the cluster was reconfigured, then this column will list the configuration of the cluster.
• Started at: Date and time the operation began.
• Completed at: Date and time the operation completed.