Replacement procedure for SFP port

This topic describes the process of replacing the SFP port on an HPE server node. The SFPs are hot-swappable, but you need to disconnect the network cable from the node.

Identifying an SFP failure (performed by the customer)

There might be a problem with the cable connection or the SFP port. To isolate the issue, first disconnect and reconnect the cables. If you face the same issue, replace the cables and try again.

SFP private port (eth4)

After you unplug the cable from the SFP private port (eth4), the node status shows unhealthy. In the NetBackup Flex Scale infrastructure management UI, navigate to Monitor > Infrastructure > Nodes.

Navigate to Infrastructure > Hardware > Network Card. On the Hardware tab, the eth4 status is shown Unplugged.

Click Dashboard. The private network status is shown unhealthy:

In the iLO remote console, navigate to System information > Network. The status for the affected port is shown Link down.

An event is generated stating that eth4 is down. In the NetBackup Flex Scale infrastructure management UI, navigate to Settings > Events:

Plug in the cable in the affected port. Review the cabling guidelines before you plug in the cable.

After plugging in the eth4 cable, the node status is shown healthy. Navigate to Monitor > Infrastructure > Nodes:

The eth4 status changes to Plugged on the Hardware tab. Navigate to Monitor > Infrastructure > Hardware > Network Card:

In the iLO remote console, navigate to System information > Network. The status for the affected port is shown OK.

An event is generated, which shows that eth4 is online. navigate to Settings > Events:

On the Dashboard, the private network status is shown healthy. It might take approximately five minutes to refresh the status in the UI.

SFP public port (eth5)

After you unplug the cable from the SFP public port (eth5), the node status shows unhealthy. In the NetBackup Flex Scale infrastructure management UI, navigate to Monitor > Infrastructure > Nodes.

Navigate to Infrastructure > Hardware > Network Card. On the Hardware tab, the eth5 status is shown Unplugged.

Click Dashboard. The public network status is shown unhealthy:

In the iLO remote console, navigate to System information > Network. The status for the affected port is shown Link down.

Plug in the cable in the affected port. Review the cabling guidelines before you plug in the cable. After plugging in the eth5 cable, the node status is shown healthy. Navigate toMonitor > Infrastructure > Nodes:

The eth5 status changes to Plugged on the Hardware tab. Navigate to Monitor > Infrastructure > Hardware > Network Card:

In the iLO remote console, navigate to System information > Network. The status for the affected port is shown OK.

On the Dashboard, the public network status is shown healthy. It might take approximately five minutes to refresh the status in the UI.

If the issue persists it implies that the SFP port is faulty. The CHS team will need to involve HPE for replacement of the port. Contact Veritas TSE to replace the node with the faulty iLO port.

Collecting HPE Active Health System (AHS) logs (performed by Veritas TSE)

Before you contact the hardware vendor for replacing the failed component, collect AHS logs. To collect the AHS logs, in the NetBackup Flex Scale infrastructure management UI, navigate to Settings > Diagnostics > Basic > Appliance.

Noting the MAC IDs (performed by Veritas TSE)

From the node-level CLI use the system hardware-health command to note the eth4 and eth5 MAC addresses.

Shutting down the node (performed by Veritas TSE)

Before an HPE representative can replace the SFP port, you must shut down the node.

To shut down the node:

Sign in to the NetBackup Flex Scale infrastructure management UI and navigate to Monitor > Infrastructure > Nodes.
On the node where the failure occurred, click the Actions menu (vertical ellipsis) from the right side of the row in the UI and click Shutdown node.
Confirm that the node is shut down successfully. In the UI, you can view the notification at the top of the page.

Replacing the SFP port (performed by the HPE vendor)

The HPE representative replaces the SFP port.

Completing the post-replacement tasks (performed by Veritas TSE)

After the hardware vendor notifies you that the hardware component is replaced, verify that the issue is resolved.

To verify that the issue is resolved, complete the following steps:

Restart the node from the iLO remote console using the Power > Momentary Press option.
The green color power symbol indicates that the node has started.
As the port is changed physically, the MAC address is also changed, which results in the following cases:
The node status is shown unhealthy in the UI:
An event is shown on the Settings > Events page of the UI:
From the node-level CLI, the system hardware-health command shows that eth4 was plugged and eth5 was unplugged:
Click Dashboard. The public network status is shown unhealthy:
In the iLO remote console go to System information > Network:
If you are on version 3.0 or earlier, complete the following steps to bring up eth4 and eth5. You need to elevate to root access to perform these steps.
- SSH to another cluster node:
  ssh primary@node_name
- su
- Enter the maintenance password.
- Open the /etc/sysconfig/network-scripts/ifcfg-eth1 file and you will see old MAC ID in the HWADDR field. Change the MAC ID in the HWADDR field to the new MAC ID.
- Run following commands:
  ifdown eth4
  ifup eth4
Repeat the same steps for eth5.
After completing the steps for eth4 and eth5, navigate to Settings > Services management and click Run full discovery.
Verify that the node status is shown healthy in the UI:
The public network status is shown healthy:
Verify that eth4 and eth5 both are shown PLUGGED on the Hardware tab.
Verify that the changed MAC ID can be seen in the system hardware-healthand eth4 and eth5 are shown plugged:
An event is generated notifying that eth4 and eth5 are online:

Replacement procedure for SFP port

Feedback

Feedback