Troubleshooting > Troubleshooting enclosures and enclosure groups

  

 next

Troubleshooting enclosures and enclosure groups

Add or remove enclosure is unsuccessful

Symptom Possible cause and recommendation
Unable to add a c7000 enclosure

If adding a c7000 enclosure is not successful, a notification panel provides the reason why and provides a solution to the problem. Often, the resolution is to click the add link embedded in the message; the add action rediscovers all components and updates its knowledge of the enclosure.

Enclosure is already being managed by some other management software and is claimed by that software

  1. If a first-time enclosure add does not succeed, verify that the enclosure prerequisites listed in add an enclosure to manage its contents are met. Verify that the data you entered on the screen is correct, and try the action again.

  2. Follow the guidance in the notification panel for the corrective action you need to take to successfully add the enclosure.

    Failures can occur during the add action if all information about an enclosure, its servers, or interconnect modules cannot be acquired. When this happens, an explanation of the problem and the component that caused the problem (the enclosure, a server, an interconnect) is provided in a notification panel.

  3. To re-add an enclosure, click the add link in the notification message panel (if there is one), or start the add action again from the Add Enclosure screen, supplying the address and credentials for the enclosure's Onboard Administrator.

To forcibly add the enclosure to the appliance, see Claim a c7000 enclosure currently being managed by another management system.
Unable to forcibly add a c7000 enclosure

You forcibly added a c7000 enclosure but received an error message. This happens in cases where there is a VCMode set and Virtual Connect (VC) is managing the enclosure.

  1. Manual clean-up of the configuration is needed, investigate the following items:

    • The management URL might still point to the appliance. If so, it needs to be reset to point at the first interconnect in the enclosure. To fix this, use the following ssh commands to go into the Onboard Administrator (using administrator credentials) and change the management URL to point to the first active VC interconnect's IP address:

      clear vcmode

      Disassociates the enclosures from the appliance.
      restart interconnect N

      (where N is the bay number of a VC interconnect) Performing this step for every VC interconnect in the enclosure causes the interconnect to revert to a default configuration.
      restart oa N

      (where N is the bay number of the active Onboard Administrator) This causes the OA to obtain the management URL from the first VC interconnect.

  2. After manual configuration, Add a c7000 enclosure or refresh the enclosure.
An existing enclosure is detected as being new after a midplane is replaced

You replaced the enclosure midplane but did not follow the recommended procedure in the hardware documentation.

Recommendation: Add a c7000 enclosure.
Unable to remove a c7000 enclosure

You might be unable to remove a c7000 enclosure for the following reasons:

  • Lack of communication with the hardware during the remove action can prevent the appliance from being able to properly manage the interconnect, server hardware, and enclosure settings.

    To forcibly remove an enclosure from the appliance due to lack of communication, see Forcibly remove a c7000 enclosure if a remove action fails.

  • The enclosure is not removed from the appliance. This is typically a problem on the appliance itself, and the best resolution is to follow instructions in the notification panels.

  • The enclosure is removed but due to a communication failure, the configuration requires manual intervention to correct.

If manual clean-up of the configuration is needed, investigate the following items:

  • The management URL might still point to the appliance. If so, it needs to be reset to point at the first interconnect in the enclosure. To fix this, use the following ssh commands to go into the Onboard Administrator (using administrator credentials) and change the management URL to point to the first active VC interconnect's IP address:

    clear vcmode

    Disassociates the enclosures from the appliance.
    restart interconnect N

    (where N is the bay number of a VC interconnect) Performing this step for every VC interconnect in the enclosure causes the interconnect to revert to a default configuration.
    restart oa N

    (where N is the bay number of the active Onboard Administrator) This causes the OA to obtain the management URL from the first VC interconnect.

  • The interconnects might still be claimed by the appliance. If this is the case, you have to remove the interconnects manually.
Unable to unconfigure single sign-on (SSO) on the Onboard Administrator when adding or removing an enclosure

Resolution: Remove all certificates and restart the OA (Onboard Administrator).

  1. From the OA user interface, select Users/Authentication.

  2. Select HPE SSO integration, in the right pane.

  3. Verify that Settings, Trust mode is set to Trust by Certificate.

  4. Select the Certification Information tab and remove all HPE SSO Certificates.

  5. Reboot the OA.

  6. Re-add the enclosure.

Unassigned server profile connections cannot be migrated

Symptom

The HPE OneView migration compatibility report shows a warning or blocking issue due to an unassigned server profile connection. To determine how to resolve an unassigned server profile connection depends on the reason for the unassigned connection and the ramifications to the OS configuration if the connection is not migrated. Review the following solutions to determine the cause of the issue and how to resolve it.

Resolve server profile connections associated with a specific adapter port

Cause

An unassigned server profile connection cannot be migrated. The connection was created in VCM to associate with a specific adapter port.

VCM associates Ethernet connections to server adapter ports with a round-robin algorithm. If a connection is needed to map to a mezzanine port instead of a LOM port, an unassigned connection is created in VCM to force the mapping. The first two connections are unassigned and the next two connections map to the mezzanine ports.

HPE OneView allows a server profile connection to be directly associated with a specific adapter port. During migration, HPE OneView maps the specific connections associated with the server in VCM to the specific adapter ports in HPE OneView. Because the unassigned connection is not migrated, the removal of the corresponding interface within the operating system might cause issues.

Action

In VCM, perform one of the following actions based on the type of connection.

  • Fibre Channel — If an OS impact is not expected, proceed with the migration without the connection. If an OS impact is expected, resolve the issue after migration.

  • FCoE — Assign a network with an uplink set containing an uplink port.

  • FCoE FC SAN and native Fibre Channel — Assign a SAN fabric with an uplink port or proceed with the migration without the connection.

  • iSCSI — Assign a network to the server profile connection or proceed with the migration without the connection.

  • Ethernet — Do one or more of the following:

    • If the downlink port status is irrelevant, assign an unused, private network to eliminate server-to-server traffic.

    • If the downlink port must be disabled, associate a private network with an uplink set with an unused port. Enable Smart link on the network.

    • If the connection is not necessary, delete the connection.

Assign a network for pre-allocated virtual MAC/WWNs

Cause

An unassigned server profile connection cannot be migrated. The connection was created in VCM as a placeholder for pre-allocation. Pre-allocation of virtual MAC/WWNs requires a network or fabric association with a profile connection.

Action

In VCM, perform one of the following actions based on the type of connection.

  • FCoE — Assign a network with an uplink set containing an uplink port.

  • FCoE FC SAN and native Fibre Channel — Assign a SAN fabric with an uplink port or proceed with the migration without the connection.

  • iSCSI — Assign a network to the server profile connection or proceed with the migration without the connection.

  • Ethernet — Do one or more of the following:

    • If the downlink port status is irrelevant, assign an unused, private network to eliminate server-to-server traffic.

    • If the downlink port must be disabled, associate a private network with an uplink set with an unused port. Enable Smart link on the network.

    • If the connection is not necessary, delete the connection.

Plan for mitigation for place holders created for an absent interconnect

Cause

An unassigned server profile connection cannot be migrated. The connection was created in VCM as a redundant FC or FCoE connection between vertical interconnects.

If redundant FC or FCoE connections are created between vertical interconnects (for example, when using two adapters for redundancy) instead of horizontally-adjacent interconnects, VCM requires a connection placeholder for the adapter even though the adapter port is not associated with an interconnect. HPE OneView does not require a placeholder since connections can be directly associated with a specific adapter port.

For FC connections, the detection of a placeholder connection to an absent interconnect is a warning issue on the HPE OneView compatibility report, regardless of the server power status because the interface associated with the placeholder connection is present for the adapter port, even after server reboot. For FCoE connections, if the server power is on, the detection is a blocking issue on the compatibility report because the removal of the connection might result in the removal of the OS interface when the server is rebooted.

Action

  • FC SAN — Proceed with the migration.

  • FCoE and offline migration — Determine if there will be OS impact and plan for mitigation after the migration.

  • FCoE and in-service migration — Power off the server and determine if there will be OS impact and plan for mitigation after the migration.

Remove FCoE, iSCSI, and FC connections created automatically by default

Cause

An unassigned server profile connection cannot be migrated. The connection was created in VCM because the VCM GUI automatically associates connections with FCoE, iSCSI, and FC interconnects.

Action

If they are not needed, delete the FCoE, iSCSI, and FC connections from the server profile.

More information
About unassigned VCM server profile connections during migration
About blocking issues during migration
Migrate a c7000 enclosure currently managed by VCM

Migration is unsuccessful

Symptom Possible cause and recommendation
You see a message indicating HPE OneView is unable to migrate the VCM enclosure

Failures can occur during the add action if all information about an enclosure, its servers, or interconnect modules cannot be acquired.

When this happens, an explanation of the problem and the component that caused the problem (the enclosure, a server, an interconnect) is provided in the compatibility report.

  1. Review each issue listed in the compatibility report and perform the corrective action.

  2. Retry the migration.

  3. If the migration cannot be performed, revert back to Virtual Connect Manager by performing the following steps:

    1. Remove all server profiles that were created during the migration.

    2. Remove the enclosure from the appliance

    3. From the OA UI, reset the lowest bay Virtual Connect interconnect.

    4. From the VC module, log in with the factory default credentials and then recover the VC configuration from the backup file.
You see a message indicating the migration of the enclosure did not complete

Migration has occurred but the migration task in HPE OneView does not show completed. This error can appear if an appliance restart occurred during the migration.

  1. View the migration tasks in the Activity view and follow any proposed resolutions.

  2. If the problem persists, refresh the enclosure.

  3. If the problem continues, reset the migrationsubstate by performing the following steps:

    1. Obtain the “auth:{token}”.

      GET /rest/login-sessions

    2. Enter the following command using the {AUTH} token obtained from step 1:

      curl -ik -X PATCH -H "Content-Type:application/json" -H "X-API-Version:300" -H "auth:${AUTH}" -d '{"op":"replace","path":"/migrationState","value":"NotApplicable"}

Invalid OA certificate

Symptom

Invalid certificate message is displayed

Cause

The OA single sign-on certificate can be corrupted when the OA firmware is downgraded to a lower version and then is upgraded to a higher version.

Action

Reset the OA:

  1. From the OA UI, select security→HPESIM SSO.

  2. Delete the corrupted certificate, which is shown in yellow.

  3. To re-install the original certificate, refresh the enclosure.

prev

 

up

 next

Troubleshooting email notifications 

home

 Troubleshooting firmware bundles

Copyright © 2013, 2016 Hewlett Packard Enterprise Development LP