Storage sever with hot plug and unplug capabilities

• US 9,936,024 B2
• Filed: 02/06/2017
• Issued: 04/03/2018
• Est. Priority Date: 11/04/2014
• Status: Active Grant

• Current Assignee: Pavilion Data Systems, Inc.
• Original Assignee: Pavilion Data Systems, Inc.
• Inventors: Malwankar, Kiron Balkrishna, Govindasamy, Raghuraman, Melnic, Dan M.
• Primary Examiner(s): Pollack, Melvin H
• Application Number:
  US15/425,917
• Publication Number:
  US 20170149897A1
• Time in Patent Office: 421 Days
• Field of Search: 709217
• CPC Class Codes: G06F 3/061 Improving I/O performance
• CPC Class Codes: G06F 3/0614 Improving the reliability of storage systems
• CPC Class Codes: G06F 3/0659 Command handling arrangements, e.g. command buffers, queues, command scheduling
• CPC Class Codes: G06F 3/0664 at device level, e.g. emulation of a storage device or system
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: G06F 3/0688 Non-volatile semiconductor memory arrays
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
• CPC Class Codes: H04L 67/568 Storing data temporarily at an intermediate stage, e.g. caching

Abstract

A storage server generates a device tree representing a maximal topology of the storage server, wherein the maximal topology is a maximum number of storage controllers and solid state drives (SSDs) that are supported by the storage server, and wherein a plurality of entries in the device tree contain placeholder information. The storage server discovers one or more SSDs of the storage server and determines drive information for the one or more SSDs that have been discovered. The storage server replaces the placeholder information in one or more of the plurality of entries in the device tree with the drive information for the one or more SSDs.

Claims

1. A method comprising:

generating, by a storage server, a device tree representing a maximal topology of the storage server, wherein the maximal topology is a maximum number of storage controllers and solid state drives (SSDs) that are supported by the storage server, and wherein a plurality of entries in the device tree contain placeholder information;

discovering one or more SSDs of the storage server;

determining drive information for the one or more SSDs that have been discovered; and

replacing the placeholder information in one or more of the plurality of entries in the device tree with the drive information for the one or more SSDs.

2. The method of claim 1, further comprising:

detecting a new SSD that has been plugged into the storage server while the storage server is powered on;

applying power to the new SSD;

determining drive information about the new SSD; and

replacing the placeholder information in an entry of the plurality of entries in the device tree with the drive information for the new SSD.

3. The method of claim 2, further comprising:

determining whether there is sufficient power to operate the new SSD and other components of the storage server; and

applying the power to the new SSD responsive to determining that there is sufficient power.

4. The method of claim 1, further comprising:

detecting a new SSD that has been plugged into the storage server while the storage server is powered on;

determining that there is insufficient power to operate the new SSD and other components of the storage server; and

failing to apply power to the new SSD.

5. The method of claim 1, further comprising:

detecting that an SSD of the one or more SSDs is being unplugged from the storage server while the storage server is powered on;

sending a message to a storage controller and a host that are using the SSD, wherein the message causes the storage controller and the host to reach a quiescent state and deallocate resources being used with reference to the SSD;

sending a second message to the storage controller to cause the storage controller to disconnect from the SSD;

replacing the drive information for the SSD with placeholder information in an entry in the device tree associated with the SSD; and

removing power from the SSD.

6. The method of claim 5, wherein detecting that the SSD is being unplugged comprises detecting that a physical presence link to the SSD has become inactive.

7. The method of claim 5, wherein detecting that the SSD is being unplugged comprises determining that the SSD has failed to respond to one or more messages.

8. The method of claim 7, wherein the one or more messages comprise one or more peripheral component interconnect express (PCIe) messages.

9. The method of claim 1, wherein the one or more SSDs comprise one or more nonvolatile memory express (NVMe) drives.

10. A storage server comprising:

a plurality of solid state drives (SSDs);

a plurality of storage controllers operatively coupled to the plurality of SSDs; and

an array manager, operatively coupled to the plurality of SSDs and to the plurality of storage controllers, the array manager to;

generate a device tree representing a maximal topology of the storage server, wherein the maximal topology is a maximum number of storage controllers and SSDs that are supported by the storage server, and wherein a plurality of entries in the device tree contain placeholder information;

discover the plurality of SSDs;

determine drive information for the plurality of SSDs that have been discovered; and

replace the placeholder information in at least a subset of the plurality of entries in the device tree with the drive information for the plurality of SSDs.

11. The storage server of claim 10, wherein the array manager is further to:

detect a new SSD that has been plugged into the storage server while the storage server is powered on;

apply power to the new SSD;

determine drive information about the new SSD; and

replace the placeholder information in an entry of the plurality of entries in the device tree with the drive information for the new SSD.

12. The storage server of claim 11, wherein the array manager is further to:

determine whether there is sufficient power to operate the new SSD and other components of the storage server; and

apply the power to the new SSD responsive to determining that there is sufficient power.

13. The storage server of claim 10, wherein the array manager is further to:

detect a new SSD that has been plugged into the storage server while the storage server is powered on;

determine that there is insufficient power to operate the new SSD and other components of the storage server; and

fail to apply power to the new SSD.

14. The storage server of claim 10, wherein the array manager is further to:

detect that an SSD of the plurality of SSDs is being unplugged from the storage server while the storage server is powered on;

send a message to a storage controller of the plurality of storage controllers using the SSD and to a host using the SSD, wherein the message causes the storage controller and the host to reach a quiescent state and deallocate resources being used with reference to the SSD;

send a second message to the storage controller to cause the storage controller to disconnect from the SSD;

replace the drive information for the SSD with placeholder information in an entry in the device tree associated with the SSD; and

remove power from the SSD.

15. The storage server of claim 14, wherein detecting that the SSD is being unplugged comprises detecting that a physical presence link to the SSD has become inactive.

16. The storage server of claim 14, wherein detecting that the SSD is being unplugged comprises determining that the SSD has failed to respond to one or more messages.

17. The storage server of claim 16, wherein the one or more messages comprise one or more peripheral component interconnect express (PCIe) messages.

18. The storage server of claim 10, wherein the plurality of SSDs comprise a plurality of nonvolatile memory express (NVMe) drives.

19. The storage server of claim 10, further comprising:

a switch, wherein the switch couples the plurality of SSDs to the plurality of storage controllers and to the array manager.

20. The storage server of claim 10, wherein:

each storage controller of the plurality of storage controllers comprises a system on a chip (SoC) comprising a memory, a processing device, a network interface controller (NIC) and a peripheral component interconnect express (PCIe) port; and

the array manager comprises an additional SoC comprising an additional memory, an additional processing device and an additional PCIe port.