VM-AWARE FTL DESIGN FOR SR-IOV NVME SSD

• US 20170242722A1
• Filed: 04/19/2016
• Published: 08/24/2017
• Est. Priority Date: 02/24/2016
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Qiu, Sheng, Ki, Yang Seok
• Granted Patent: US 10,235,198 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 12/00 Accessing, addressing or allocating within memory systems or architectures digital input from, or digital output to record carriers, e.g. to disk storage units, G06F3/06
• CPC Class Codes: G06F 13/385 for adaptation of a particular data processing system to different peripheral devices
• CPC Class Codes: G06F 2009/45579 I/O management, e.g. providing access to device drivers or storage
• CPC Class Codes: G06F 2009/45583 Memory management, e.g. access or allocation
• CPC Class Codes: G06F 3/061 Improving I/O performance
• CPC Class Codes: G06F 3/0655 Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
• CPC Class Codes: G06F 3/0679 Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

A mass storage device for providing persistent storage. The system includes a plurality of instances of virtual flash translation layers, each associated with a namespace and configured to provide, to one or more virtual machines executing in a host connected to the mass storage device, access to read and write operations in the persistent storage.

Claims

1. A mass storage device, comprising:

a flash memory comprising a plurality of physical blocks, each of the physical blocks comprising a plurality of physical pages, the plurality of physical blocks comprising;

a first set of physical blocks; and

a second set of physical blocks,

the first set of physical blocks and the second set of physical blocks being disjoint; and

a controller configured to implement a first virtual flash translation layer instance and a second virtual flash translation layer instance,

the first virtual flash translation layer instance being configured to receive storage access requests directed to logical page numbers and to generate storage access requests directed to physical page numbers within the first set of physical blocks, and

the second virtual flash translation layer instance being configured to receive storage access requests directed to logical page numbers and to generate storage access requests directed to physical page numbers within the second set of physical blocks.

2. The device of claim 1, wherein each of the first virtual flash translation layer instance and the second virtual flash translation layer instance is further configured to perform garbage collection operations on its respective set of physical blocks.

3. The device of claim 2, wherein:

the first virtual flash translation layer instance has a first weight and the second virtual flash translation layer instance has a second weight,

the first set of physical blocks comprises a first plurality of overprovision blocks reserved for storage of data for garbage collection operations,

the second set of physical blocks comprises a second plurality of overprovision blocks reserved for storage of data for garbage collection operations, and

the product of;

the ratio of the number of physical blocks in the first plurality of overprovision blocks, to the number of physical blocks in the first set of physical blocks, and

the ratio of the number of physical blocks in the second set of physical blocks to the number of physical blocks in the second plurality of overprovision blocks,

is equal to the ratio of the first weight to the second weight.

4. The device of claim 1, comprising:

one or more first packages, each comprising a page allocated to the first virtual flash translation layer instance, and

one or more second packages, each not comprising any pages allocated to the first virtual flash translation layer instance,

the device further comprising a page allocator, the page allocator being configured to allocate pages to the first virtual flash translation layer instance in response to storage access requests, the allocation of pages to the first virtual flash translation layer instance comprising;

allocating a page in a package of the one or more first package, when a page is available in any of the one or more first packages; and

allocating a page in a package of the one or more second packages, when no page is available in any of the one or more first packages.

5. A system comprising:

a host comprising;

a first virtual machine associated with a first namespace; and

a second virtual machine associated with a second namespace, the first virtual machine being configured to send storage access requests to a first virtual flash translation layer instance instantiated in a mass storage device, and

the second virtual machine being configured to send storage access requests to a second virtual flash translation layer instance instantiated in the mass storage device.

6. The system of claim 5, wherein the host further comprises:

a third virtual machine associated with the second namespace,

the third virtual machine being configured to send storage access requests to the second virtual flash translation layer instance.

7. The system of claim 5, wherein the host comprises a hypervisor, configured to assign to each of the first virtual machine and the second virtual machine a weight in proportion to an anticipated write activity level of the virtual machine.

8. The system of claim 5, wherein the host comprises a hypervisor, configured to assign each of the first virtual machine and the second virtual machine a weight in proportion to a persistent storage quality of service of the virtual machine.

9. The system of claim 5, further comprising the mass storage device, wherein:

the mass storage device comprises;

a flash memory comprising a plurality of physical blocks, each of the physical blocks comprising a plurality of physical pages, the plurality of physical blocks comprising;

a first set of physical blocks; and

a second set of physical blocks,

the first set of physical blocks and the second set of physical blocks being disjoint;

a controller configured to implement a first virtual flash translation layer instance and a second virtual flash translation layer instance, and

each of the first virtual flash translation layer instance and the second virtual flash translation layer instance is further configured to perform garbage collection operations on its respective set of physical blocks.

10. The system of claim 9, wherein:

the first virtual flash translation layer instance has a first weight and the second virtual flash translation layer instance has a second weight,

the first set of physical blocks comprises a first plurality of overprovision blocks reserved for storage of data for garbage collection operations,

the second set of physical blocks comprises a second plurality of overprovision blocks reserved for storage of data for garbage collection operations, and

the ratio of;

the ratio of;

the number of physical blocks in the first plurality of overprovision blocks, to

the number of physical blocks in the first set of physical blocks, to

the ratio of;

the number of physical blocks in the second plurality of overprovision blocks, to

the number of physical blocks in the second set of physical blocks, is equal to the ratio of the first weight to the second weight.

11. The system of claim 9, comprising:

one or more first packages each comprising a page allocated to the first virtual flash translation layer instance and

one or more second packages, each not comprising any pages allocated to the first virtual flash translation layer instance,

the first packages and the second packages together comprising all packages of the device and comprising the plurality of physical blocks,

the device further comprising a page allocator, the page allocator being configured to allocate pages to the first virtual flash translation layer instance in response to storage access requests, the allocation of pages to the first virtual flash translation layer instance comprising;

allocating a page in a package of the one or more first package, when a page is available in any of the one or more first packages; and

allocating a page in a package of the one or more second packages, when no page is available in any of the one or more first packages.

12. A method, comprising:

receiving, by a first virtual flash translation layer instance in a mass storage device comprising a plurality of physical blocks, a first storage access request directed to a first logical page number;

generating, by the first virtual flash translation layer instance, a storage access request directed to a physical page number within a first subset of the plurality of physical blocks;

receiving, by a second virtual flash translation layer instance in the mass storage device, a second storage access request directed to a second logical page number; and

generating, by the second virtual flash translation layer instance, a storage access request directed to a physical page number within a second subset of the plurality of physical blocks,

the first subset and the second subset being disjoint.

13. The method of claim 12, further comprising:

creating, by a host, a first namespace and a second namespace;

attaching, by the host, the first namespace to a first virtual function configured to interact with the first virtual flash translation layer instance; and

attaching, by the host, the second namespace to a second virtual function configured to interact with the second virtual flash translation layer instance.

14. The method of claim 13, further comprising:

creating, by the host, a first virtual machine and a second virtual machine,

assigning, by the host, the first virtual function to the first virtual machine, and

assigning, by the host, the second virtual function to the second virtual machine.

15. The method of claim 14, further comprising,

composing, by a guest device driver on the host, a read request comprising the first namespace;

receiving, by the mass storage device, through the first virtual function, the read request;

processing, by the first virtual flash translation layer instance, the read request; and

returning, through the first virtual function, a result of the read request.

16. The method of claim 14, further comprising,

composing, by a guest device driver on the host, a write request comprising:

data to be written and

an identifier for a first namespace;

receiving, by the mass storage device, through the first virtual function, the write request;

processing, by the first virtual flash translation layer instance, the write request;

requesting, by the first virtual flash translation layer instance, from a page allocator of the mass storage device, one or more physical page numbers corresponding to physical pages available for writing;

receiving one or more physical page numbers from the page allocator; and

writing the data to one or more physical pages corresponding to the received one or more physical page numbers.

17. The method of claim 12, further comprising performing garbage collection operations, by the first virtual flash translation layer instance, on the first subset.

18. The method of claim 17, wherein:

the first virtual flash translation layer instance has a first weight and the second virtual flash translation layer instance has a second weight,

the first subset comprises a first plurality of overprovision blocks reserved for storage of data for garbage collection operations,

the second subset comprises a second plurality of overprovision blocks reserved for storage of data for garbage collection operations, and

the product of;

the ratio of the number of physical blocks in the first plurality of overprovision blocks, to the number of physical blocks in the first set of physical blocks, and

the ratio of the number of physical blocks in the second set of physical blocks to the number of physical blocks in the second plurality of overprovision blocks,

is equal to the ratio of the first weight to the second weight.