SolidFire and HCI

adding available drives in HCI

Mohisina01
1,804 Views

Hi Everyone,

 

We have the below errors, in HCI, should i power drain the node and add the available drives, also can i add the drives in bulk or do i need to add one by one starting from node ID 6.

 

can the add the drives without power drain.

Mohisina01_4-1670857749981.png

 

Mohisina01_3-1670857737409.png

 

1 ACCEPTED SOLUTION

elementx
1,790 Views

It depends what OS version you have; on older Element OS *and* on busy clusters, it's probably better to add one or two at once.

With Element 12.7 you can add all 5 data disks at once (I'd still add the metadata disk separately; either first or last).

 

The fact that you have 6 available means MD (Disk #0) has been kicked out as well (maybe due to extra downtime), which may be a sign of some h/w or s/w (driver, firmware, etc.) issues which is concerning.

 

You don't need to power-drain the node to add disks, but you could in order to clear those errors (since the disks aren't in use anyway).

 

- Power off the node

- Maybe consider rebooting BMC (not wiping its configuration) as per https://kb.netapp.com/Advice_and_Troubleshooting/Hybrid_Cloud_Infrastructure/NetApp_HCI/How_to_reset_BMC_on_NetApp_HCI_compute_nodes

- Then boot the node, give it 5-10 mins of standby time to ensure there are no other errors, then add disks

- After data rebalancing has completed, consider updating drivers/firmware/Element OS to latest

 

It's quite a few things (maybe even vCenter Plug-in needs to be removed, and then installed after Element OS upgrade), so for a production system maybe you could contact Support to confirm these steps or do a Zoom session for a guided upgrade.

View solution in original post

1 REPLY 1

elementx
1,791 Views

It depends what OS version you have; on older Element OS *and* on busy clusters, it's probably better to add one or two at once.

With Element 12.7 you can add all 5 data disks at once (I'd still add the metadata disk separately; either first or last).

 

The fact that you have 6 available means MD (Disk #0) has been kicked out as well (maybe due to extra downtime), which may be a sign of some h/w or s/w (driver, firmware, etc.) issues which is concerning.

 

You don't need to power-drain the node to add disks, but you could in order to clear those errors (since the disks aren't in use anyway).

 

- Power off the node

- Maybe consider rebooting BMC (not wiping its configuration) as per https://kb.netapp.com/Advice_and_Troubleshooting/Hybrid_Cloud_Infrastructure/NetApp_HCI/How_to_reset_BMC_on_NetApp_HCI_compute_nodes

- Then boot the node, give it 5-10 mins of standby time to ensure there are no other errors, then add disks

- After data rebalancing has completed, consider updating drivers/firmware/Element OS to latest

 

It's quite a few things (maybe even vCenter Plug-in needs to be removed, and then installed after Element OS upgrade), so for a production system maybe you could contact Support to confirm these steps or do a Zoom session for a guided upgrade.

Public