Completing the Procedure - Appliance for Hadoop

Appliance for Hadoop 6 Platform Hardware Replacement Guide for Customers

Product
Appliance for Hadoop
Published
February 2018
Language
English (United States)
Last Update
2018-04-16
dita:mapPath
gkl1471885608819.ditamap
dita:ditavalPath
hadoop6.ditaval.ditaval
dita:id
ynp1471650878435
Product Category
Hardware
  1. Return to Server Management Client and, in the Replace FRU window, click Ok.
  2. When the Replace FRU confirmation dialog box appears, click Ok to close the FRU replacement session in Server Management Client.
  3. To stop Hadoop services on a node in a cluster, follow one of these procedures:
    Platform Procedure
    Hortonworks HDP:
    1. Log on to Ambari.
    2. Select the Hosts tab.
    3. Select the host you want to stop services on.
    4. Select Host Actions > Stop All Components.
    Cloudera CDH:
    1. Log on to Cloudera Manager.
    2. On the Cloudera Manager Home page, select Hosts.
    3. From the Name list, select the node you want to stop Hadoop services on.
    4. Select the checkbox next to the targeted host.
    5. Click the Actions for Selected tab.
    6. Select one of the following:
      • In CDH 5.4 and 5.5, Stop All Roles.
      • In CDH 5.7 and later, Stop Roles on Hosts.
  4. Configure replaced disks to make them available to Hadoop services:
    1. Return to Server Management Client.
    2. In the Server Management Client Tree View, select the node.
    3. Select Functions > Connect.
    4. In the Connect dialog box, select Secure Shell (SSH) then click Ok.
    5. In the New Connection dialog box, ensure Display Connections Manager is selected and click Ok.
    6. In the Client Connections Window, in the Console, type the root username and password and press Enter.
    7. Type hcli support configureDisk
    8. Answer yes to each prompt.
    If any problems are encountered, notify Teradata Customer Support. Use a contact method that is in accordance with your service contract. If your system is not set up with Automatic Incident Creation (AIC), log on to https://access.teradata.com and initiate or update an incident. If your system is set up with AIC, close the maintenance window and add a comment.
  5. To start or restart Hadoop services on a node:
    Platform Procedure
    Hortonworks HDP:
    1. Log on to Ambari.
    2. Select the Hosts tab.
    3. Select the host you want to start or restart services on.
    4. Select Host Actions > Start.
    Cloudera CDH:
    1. Log on to Cloudera Manager.
    2. On the Cloudera Manager Home page, select Hosts.
    3. Select the checkbox next to the targeted host.
    4. Click the Actions for Selected tab.
    5. Select one of the following:
      • In CDH 5.4 and 5.5, Start all Roles
      • In CDH 5.7 and later, Start Roles on Hosts
  6. Verify that Ambari or Cloudera Manager recognizes all the storage available on the node:
    Platform Procedure
    Hortonworks HDP:
    1. Log on to Ambari.
    2. Select the Hosts tab.
    3. Verify that the host is listed and its status is .
    4. Verify that the correct storage capacity is listed for the host.
    Cloudera CDH:
    1. Log on to Cloudera Manager.
    2. Select the Hosts tab.
    3. Verify that the host is listed and its status is .
    4. Verify that the correct storage capacity is listed for the host.
  7. Add a comment to the summary alert:
    1. In the Server Management Client window, select the Summary Alerts tab. A list of summary alerts appears.
    2. Locate the appropriate summary alert.
    3. On the right side, click the three-bar icon.
    4. Select Add Comment. The Comments window appears.
    5. You must add entries in both the Author and the Comment box.
    6. Click Submit. The comment is attached to the summary alert.
  8. Clear the summary alert:
    1. In the Summary Alerts tab, select the alert containing the corresponding failed drive. The synopsis indicates a failed drive (similar to the following): A Node Drive Failure has been detected in SMP.x.x.x.x
    2. In the Summary Alerts tab, select Functions > Clear Summary Alerts.
    3. At the confirmation prompt, click Yes. The status of the alert group changes from Active to Cleared (has checkmark). Clearing an alert group removes it from the list of problem scenarios being monitored.
  9. Close the Maintenance Window:
    1. In the Server Management Client window, click the Overview tab.
    2. In the Maintenance Window on the right side of the column, click the trash can icon.
    3. In the End Maintenance Window, click End.
    4. In the Information window, click OK.
  10. Log off Server Management Client.