Replacing a Disk
When a data drive associated with a HyperScale node fails due to disk issues, it can be replaced with a new empty disk.
If multiple disks need to be replaced, replace them one-by-one, using the steps described in this page.
- The capacity of the new disk must be equal to, or larger than the disk that you want to replace. However, the replacement process for a larger disk can take an excessive amount of time.
- Ensure that the disk being replaced has its entry unchanged in
- No file system should exist on the new disk.
- For Service Pack 15 (and below) the disk will have a different UUID for the old disk and the new disk. The new disk will be given a different disk number from that of the original disk.
From Service Pack 16 onwards, the UUID of the old disk and the new disk will be the same. The new disk will be given the same disk number as the disk being replaced.
Before You Begin
- Attach the new disk in an unmounted state to the MediaAgent that has the failed disk.
Note: To physically identify the failed disk, look for the drive with a lit drive light. The drive light will be automatically lit when there is a disk failure.
- Use the following command to verify whether the disk is in the unmounted state:
The output will list all the block devices. The new device would most likely be the last device in the list. Verify that the MOUNTPOINT is left blank for the new device in the output.
- In the case of HyperScale 1.5 Appliance, where the Hosted Engine (oVirt or the virtualization management application that manages multiple virtual machines on multiple hosts) is installed in the HyperScale setup, restart the glusterd daemon on the node in which the disk is being replaced, using the following command:
# systemctl restart glusterd
Caution: Disk replacement must be performed using the Command Center as described in the following steps. Manual disk replacement attempts may make the disk ineligible for replacement.
- Open the Command Center.
- From the navigation, click Storage > HyperScale.
The HyperScale page appears.
- Click the storage pool name that has the disk you want to replace.
The storage pool page appears.
- Under Nodes, click the node that has the disk you want to replace.
The node page appears.
- Under Disks, locate the disk you want to replace.
- Click the Actions button , select Replace, and then click Yes on the confirmation dialog box.
- The new disk appears with the Disk replaced and data is being recovered disk status initially when the data recovery from the failed disk is in progress.
- After successful data recovery, the new disk appears with the Ready disk status.
Caution: If multiple disks are being replaced, wait until the new disk is successfully discovered and the disk status is displayed as Ready, before replacing another disk.
Last modified: 3/12/2021 3:21:35 PM