Dell ObjectScale 1.3 Administration Guide

Disk replacement service procedure

ObjectScale generates messages in ObjectScale Portal when failed persistent volumes (PVs) are found. Once ObjectScale finds failed PVs, the disk replacement service procedure automatically begins to recover data on the failed drive. You can monitor the status of the recovery and confirm a successful recovery using the Events within the ObjectScale Portal user interface.

There are two types of disk replacement service procedures ObjectScale uses when necessary. One method is used when the disk contains a StorageServer (SS) or NVMeEngine pod, and the other is used when the disk contains only a non-SS pod.

When ObjectScale uses the disk replacement service procedure for SS pods, the status of the object store changes. This affects the availability of maintenance mode service procedures until the disk replacement service procedure completes. When ObjectScale uses the disk replacement service procedure for non-SS pods, ObjectScale creates a DR SP. This does not affect your ability to place nodes into a maintenance mode.

Shown below is the high-level overview of the disk replacement service procedure.

• When a disk's health becomes FAILED or SUSPECT ObjectScale Portal displays an issue in the Alerts section with the details of the degraded disk status.
• During the disk replacement service procedure for SS pods, the object store status changes from Available to ReplacingPV. Otherwise, the object store status is unchanged.
• Then the data recovery portion of the service procedure is initiated automatically, if applicable.
  • If there is available capacity (free disks) in the node, a volume is re-created once the recovery is completed using a new available disk in the system. The old disk is now waiting to be replaced (pod and object store are in a good state, and no longer impacted).
  • If there is no available capacity on the node (no free disks), the system waits for the user to perform the necessary actions required before initiating the volume recreation.
• The alert is updated with a message that the Disk is ready for replacement
• Once you receive this message, you can replace the hard drive in the node.
• After you replace the disk and ObjectScale completes the disk replacement SP, the alert message is updated to Disk has been successfully replaced.