Data-protection-aware capacity provisioning of shared external volume
AssigneeHitachi, Ltd.
InventorsNaruki KURATA, Tomohiro KAWAGUCHI
Life of this patent
Estimating the term…
Abstract
Example implementations described herein facilitate a drive unit to share physical mediums (e.g. solid state drives) among multiple storage controllers or storage nodes in a distributed storage system, while keeping them physically independent in a redundancy group to avoid single point of failure, by providing an interface to notify redundancy group to the drive unit.
Claims10
A drive unit, comprising:
- a plurality of physical mediums;
- a plurality of virtual mediums distributed across the plurality of physical mediums;
- and
- a processor, configured to:
- determine, from a virtual medium allocation request received from a storage controller of the one or more storage controllers, a controller identifier of the storage controller;
- determine a first set of the plurality of virtual mediums allocated to the storage controller based on the controller identifier;
- determine a second set of the plurality of physical mediums corresponding to the first set of the plurality of virtual mediums allocated to the storage controller;
- and
- respond to the virtual medium allocation request with a third set of the plurality of virtual mediums corresponding to remaining ones of the plurality of physical mediums outside of the second set of the plurality of physical mediums.
A storage controller managing a redundant array of inexpensive disk (RAID) group of virtual mediums and connected to a drive unit configured to provide physical mediums for the RAID group of virtual mediums, the storage controller comprising:
- a processor, configured to:
- provide a virtual medium allocation request comprising a controller identifier of the storage controller to the drive unit;
- receive a set of virtual mediums eligible for inclusion into the RAID group, the set of virtual mediums corresponding to physical mediums in the drive unit that are unused in the RAID group;
- and
- allocate one or more virtual mediums from the set of virtual mediums to the RAID group.
A distributed system comprising a plurality of distributed storage nodes managing a chunk group of virtual mediums for a corresponding pool volume, and connected to a drive unit configured to provide physical mediums for the chunk group of virtual mediums, each of the plurality of distributed nodes comprising:
- a processor, configured to:
- provide a virtual medium allocation request comprising an identifier of the each of the plurality of distributed nodes to the drive unit;
- receive a set of virtual mediums eligible for inclusion into the chunk group, the set of virtual mediums corresponding to physical mediums in the drive unit that are unused in the chunk group;
- and
- allocate one or more virtual mediums from the set of virtual mediums to the chunk group.
Dependencies are inferred from the claim text.