This section explains how to replace hardware components in HPE Alletra 4110 nodes.
- We strongly recommend engaging an on-site engineer to replace failed hardware components including but not limited to any procedure that:
- This guide doesn't cover
- You haven't received training on
- Requires precautions to avoid damage caused by electrostatic discharge (ESD) by using industry standard anti-static equipment (such as gloves or wrist straps)
- We don’t recommend updating firmware on Qumulo-certified hardware nodes unless the equipment manufacturer or a member of the Qumulo Care team advises you to do so. For questions about this process, contact the Qumulo Care team.
To Perform the Part Replacement Procedure by Using the FVT
When you replace a component such as the motherboard or a NIC card in your node, you must ensure that the firmware version and configuration are correct for your new components. To do this, you must perform the part replacement procedure by using the FVT.
-
Boot by using the latest version of the Qumulo Core USB Drive Installer.
-
Select [*] Perform maintenance.
-
Select [2] Perform automatic repair after part replacement (non-destructive).
The part replacement procedure runs and the FVT passed! message appears.
In some cases, after the part replacement procedure, the message
FIX: Run the FVT flash command.
appears. Enter 1
as you would for a fixable issue to reboot the node and then repeat the part replacement procedure.Replacing an NVMe Drive
Your HPE Alletra 4110 chassis contains either 20 or 10 NVMe drives.
For information about replacing an NVMe drive, see Storage Drives (NVMe) in the HPE documentation.
Replacing an M.2 Boot Drive
Your HPE Alletra 4110 chassis contains one NVMe boot drive in a server module.
For information about replacing an M.2 boot drive, see Installing an M.2 Solid State Drive in the HPE documentation.
Initializing the Replacement Boot Drive
After you replace the boot drive, you must initialize the replacement boot drive by using the Qumulo Core Installer and then rebuild the replacement boot drive by using a script on the node in your cluster.
Step 1: Initialize the Replacement Boot Drive
To get the correct version of the Qumulo Core Installer for the node in your cluster, contact the Qumulo Care team
-
Power on your node, enter the boot menu, and select your USB drive.
The Qumulo Core Installer begins to run automatically.
-
When prompted, take the following steps:
-
Select
[x] Perform maintenance
. -
Select
[1] Boot drive reset
and then follow the prompts.
The Qumulo Core Installer initializes the boot drive.
-
-
When the process is complete, the node is powered down automatically.
Step 2: Rebuild the Replacement Boot Drive
-
Power on your node and log in to the node by using the
qq
CLI. -
To get
root
privileges, run thesudo qsh
command. -
To stop the Qumulo Networking Services, run the
service qumulo-networking stop
command. -
To configure the IP address for the node, run the
ip addr add
command and specify the node’s IP address. For example:ip addr add 203.0.113.0/CDR dev bond0
-
Ensure that the node can ping other nodes in the cluster.
-
Run the
rebuild_boot_drive.py
script and specify the IP address of another node in the cluster, the ID of the node whose boot drive has been replaced, and the password of the administrative account of the cluster. For example:Note
If your password includes special characters such as the parenthesis ((
) or the asterisk (*
), use the backslash (\
) to escape these characters./opt/qumulo/rebuild_boot_drive.py \ --address 203.0.113.1 \ --node-id 2 \ --username admin \ --password my\(Special\*Password
Follow the prompts.
-
When the process is complete, reboot the node.
Replacing a Power Supply Unit (PSU)
Your HPE Alletra 4110 chassis contains two PSUs.
For information about replacing a PSU, see Power Supply in the HPE documentation.
Replacing a Fan
Your HPE Alletra 4110 chassis has seven fans at the back of the chassis.
For information about replacing a fan, see System Fans in the HPE documentation.
Replacing a DIMM
Your HPE Alletra 4110 chassis has 24 DIMM slots.
For information about replacing a DIMM, see DIMM Installation in the HPE documentation.
To identify which DIMM failed, you must use the baseboard management controller (BMC) on the node or another hardware monitoring solution.