Pillar Axiom: How to use Guided Maintenance in R4 and How to identify a FRU

Asset ID:	1-71-1447672.1
Update Date:	2012-05-17
Keywords:

Solution Type Technical Instruction Sure

Solution 1447672.1 : Pillar Axiom: How to use Guided Maintenance in R4 and How to identify a FRU

Applies to:

Pillar Axiom 300 Storage System - Version Not Applicable and later
Pillar Axiom 500 Storage System - Version Not Applicable and later
Pillar Axiom 600 Storage System - Version Not Applicable and later
Information in this document applies to any platform.

Goal

This guide addresses service technicians, field engineers, and others who maintain, troubleshoot, and replace system components and field replaceable units (FRUs) in a Pillar Axiom Storage System operating on R4 Software.

Fix

About Guided Maintenance Initiation

You manage the servicing of a Pillar Axiom system by means of the Guided Maintenance module of the Pillar Axiom Storage Services Manager (GUI).

CAUTION!: Access by any other means is not supported and voids the warranty for your Pillar Axiom 600 system. Remote access (ssh, telnet, ftp, and others) may also compromise data integrity.

To maintain or restore reliability to a Pillar Axiom system, you sometimes need to replace a hardware component. To replace a hardware component, you must use Guided Maintenance (unless the Pillar Axiom Oracle Support Center gives you explicit instructions not to do so).

You access Guided Maintenance through the Pillar Axiom Storage Services Manager in one of three ways:

In the Health section, click the type of hardware of interest and locate a particular FRU. To enter Guided Maintenance for that FRU, select the link.
Click the system status icon on the far left side of the status bar, which takes you to the above Health section.
In the Support section, click the Slammers or Bricks link in the navigation pane and locate a particular FRU. To enter Guided Maintenance for that FRU, select the link.

Although you can enter Guided Maintenance in any of those ways, procedures in this guide assume that you are using the Health page. System Health screens in the GUI display the status of hardware and firmware components of the Pillar Axiom system. The overall system status icon on the bottom of the screen is a summary of the hardware status and does not reflect the status of LUNs or Filesystems.

A hardware problem will typically cause filesystems and LUNs to go offline or to a degraded state. Because this is not always the case, you should check the state of the filesystems and LUNs or any associated Administrator Actions that may be listed.

Guided Maintenance on an R4 Pillar Axiom Storage System

Log in to the graphical user interface (GUI).
Enter Guided Maintenance for the target FRU.

After you initiate Guided Maintenance, you can:

Identify the target FRU to find the FRU that needs to be replaced.
Prepare the system for FRU replacement.
Remove the FRU.
Fail back the target control unit (CU).

Tip: Automatic failback of NAS Slammers is the default configuration. If this is not desired, automatic failback of NAS Slammer CUs can be disabled in the GUI (System > Global Settings > Networking > Notification). Automatic failback of SAN Slammer CUs is always enabled.

Verify the status of the FRU.

Log In to the Graphical User Interface (GUI)

1. Start the browser software on your workstation.
2. Specify the public IP address of the Pilot as the address to open.

Tip: If the IP has not been changed to a customer-specified address, use 10.0.0.2, which was set at the factory. If that IP address is not successful, try 10.0.0.3 and 10.0.0.4. If you still cannot log in, ping those addresses and contact the Pillar Axiom Oracle Support Center.

3. When prompted, enter the Primary Administrator's login name and password.
4. If this is the first time the Primary system administrator account is accessed, use administrator for the login name and pillar for the password.

When logged in, you will be prompted to change the password. Choose an appropriate password to protect the security of the server. Keep that password in a safe location.

Tip: The following tips can improve your experience using the GUI:

Because the GUI uses popup windows, configure your browser to allow popups. Current versions of Internet Explorer block pop-ups by default.
For Internet Explorer and Mozilla-based browsers, select the smallest text size you can view comfortably. Doing so helps you to see all menu items.
For Internet Explorer, disable Script Debugging in Tools > Internet Options > Advanced > Browsing.

About Guided Maintenance for the Target FRU

The Summary page in the Health section displays the status of the Pilot and all Slammers and Bricks. For any component of interest (for example, one that has a critical status), you can inspect all of its FRUs by selecting the component. You can then enter Guided Maintenance for any of its FRUs.

When you enter Guided Maintenance, if the Enable Automatic Failback of NAS Control Units option has been selected, Guided Maintenance warns you of that fact on the introductory page. This option is found in the System > Global Settings > Networking tab in the Pillar Axiom Storage Services Manager and is described in Pillar Axiom Administratorâs Guide.

When automatic recovery is enabled, Guided Maintenance will automatically fail back the target control unit (CU) of a NAS Slammer after you replace the FRU.

Tip: If you want to fail back the CU manually, de-select the Enable Automatic Failback of NAS Control Units option before you continue with Guided Maintenance. However, be sure to re-select this option when you have completed maintenance.

Note: SAN Slammers always have automatic recovery enabled.

Starting Guided Maintenance

1. Click the Health icon in the top context pane and Summary on the left hand pane

IMPORTANT! Check the background tasks at the bottom of the screen (Running Man Icon). If any tasks are running, those tasks may interfere with Guided Maintenance. Wait for them to complete. If any of these tasks do not complete, before you enter Guided Maintenance, contact the Pillar Axiom Oracle Support Center.

2. Review the hardware summary information that is displayed in the content pane.

3. Under Hardware in the navigation pane, click the component type of interest.

4. In the content pane, click the name of the component of interest to open its status page. This page shows detailed information about the FRUs.

5. Click the link for the FRU of interest.

6. Choose one of:

To enter Guided Maintenance for that FRU, click Replace Component.
To close the window and return to the component status page, click Close.

IMPORTANT! Do not exit Guided Maintenance by closing the window if you selected the step to prepare the system for FRU Replacement in Guided Maintenance, because this may leave the LUNs or filesystems in Conservative (write-through cache) mode.

About Target FRU Identification

Using Guided Maintenance, you can identify which FRU in the system needs to be replaced. If you already know which FRU to replace, you may still want to perform this step to verify the identity of the target FRU. Though this step is not required, it is highly recommended. Both Identify and Reverse Identify are available for Bricks and Slammers.

When identifying a target FRU, Guided Maintenance will:

Blink the LEDs on that FRU and on the bezel of the chassis containing that FRU.
Shut off all other LEDs on the front and back of all other Bricks and Slammers.

NOTE: Guided Maintenance does not control the LEDs for power supplies, fans, and Slammer motherboards.

To reverse identify the target FRU, Guided Maintenance, blinks all Brick and Slammer LEDs, except for the LEDs on the target FRU. Use this method if you are having trouble spotting the light patterns of Identify.

When identifying Pilots, Guided Maintenance lights the drive-activity LED on the target Pilot CU to a steady red.

IMPORTANT! The drive-activity LED may be flickering quite rapidly at times, appearing nearly steady red. This is the case during event log synchronization or Call-Home activity. Look closely to distinguish whether the LED is steady or flickering. When in doubt, you can have Guided Maintenance reverse identify the Pilot CU, which lights the drive-activity LED to a steady red on the partner CU. If that is not sufficient, compare the Pilot serial numbers to the values on the Pilot Details page in the GUI.

For Bricks, Guided Maintenance blinks all LEDs to identify the FRU except for the following fault-related LEDs (which light solid amber instead):

Fault LED on the bezel
Power LED that is adjacent to the triangular icon on the power supply.
FLT LED on the RAID controller.
Left LED on the Enclosure Services (ES) module.
FLT LED on the spare drive.

IMPORTANT! When using Guided Maintenance to identify a FRU, be sure to stand in front of the Axiom system to locate the target control unit (CU). The LEDs on the bezel will be blinking.

IMPORTANT! Axiom systems ensure continuous data access during single points of failure. A second failure, including operator errors, may cause system failure. Use care and take full advantage of Guided Maintenance features.

About System Preparation for FRU Replacements

Before Guided Maintenance prompts you to replace a FRU, the Axiom system prepares the system by first performing some preparatory steps that reallocate or move resources so that you can safely remove a FRU. For example, these steps:

Allow the administrator to power down a Slammer control unit, if necessary, before replacing a network interface module.
Fail over LUNs to the peer controller when a RAID controller is about to be replaced.
Spin down a drive when it is about to be replaced.

During some replacement procedures, the Axiom system turns off access to components, halts components, or powers off components within the system. During Guided Maintenance, watch out for specific Administrator Actions that tell you when to remove or replace a particular FRU.

CAUTION! When Guided Maintenance creates an Administrator Action that instructs you to remove the FRU, do so even if you are only testing fault injection and recovery. Otherwise, automatic recovery may be inhibited and data loss may result.

If the Guided Maintenance creates an Administrator Action that instructs you to replace a component, you may be able to recover from Guided Maintenance by canceling that Administrator Action.

NOTE: You should not exit Guided Maintenance prematurely. Otherwise, you may leave that component out of service, which may impact redundancy and recovery in the event of another fault. If you exit Guided Maintenance prematurely, you may also leave LUNs and Filesystems in Conservative mode.

Attachments

This solution has no attachment