Storage device and control method of storage controller

By designing address conversion and queue switching mechanisms of multiple storage controllers and front-end interfaces in the storage device, the I/O communication interruption problem when the controller processor is stopped is solved, and the automatic switching and continuous response capabilities of the storage device are realized.

CN119025448AInactive Publication Date: 2024-11-26HITACHI VANTARA LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410212275.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2023-05-17
Filing Date
2024-02-27
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

When the controller processor is stopped, the existing storage device cannot automatically switch I/O communication, resulting in the system being temporarily shut down and unable to continue to respond to I/O requests from the host device.

Method used

Design multiple storage controllers, one of the front-end interfaces of one controller can automatically switch I/O requests to another controller processor when the processor is stopped, ensuring the continuous I/O communication. sex.

Benefits of technology

Even if one controller's processor stops, the other controller's processor can take over I/O requests, ensuring that the storage device continues to respond to the host device's request and avoid system downtime.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025448A_ABST
    Figure CN119025448A_ABST
Patent Text Reader

Abstract

The invention provides a storage device and a control method of a storage controller, which can continue to respond to an I / O request from a host device without a special control function part such as a local router when a processor of the controller is stopped. The front-end interface of the controller is provided with a plurality of queue interfaces corresponding to each processor of the controller, so that the enqueue destination of the I / O instruction of the host can be switched according to the instruction from the processor. When the controller OS is restarted, the controller waits for completion of ongoing host I / O and performs controller shutdown and restart, so in order to determine whether or not it is possible, the controller waits for a switching source queue to become null after the processor instructs queue switching.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a storage device comprising a front-end interface and a processor. Background Art

[0002] In order to improve reliability, a storage device is provided with multiple storage controllers (hereinafter referred to as controllers) responsible for controlling the storage device. Even if a controller fails and stops operating, the remaining controllers can continue the operation. The typical number of such controllers is two.

[0003] For example, Patent Document 1 discloses an example in which the controller unit of a storage device is composed of two controllers. In this case, each controller is equipped with a processor for controlling the system, and a switch connected to each processor is provided with a non-transparent bridge (NTB: NON-TRANSPARENT BRIDGE).

[0004] One of the roles of the NTB is to convert the addresses used by the processors connected to each switch to other addresses. In particular, when two controllers are connected, the addresses used by each processor are controlled and managed by each processor, so it is necessary to convert the addresses used by the processor of one controller to the addresses used by the processor of the other controller. For example, each processor assigns independent addresses to the connected memory or protocol chip, etc., so they cannot be directly connected and use the same address. The two processors are connected by the NTB, and the destination address of the data packet passing through the NTB is converted, so that data can be transmitted to each other.

[0005] Furthermore, when each processor is stopped, the addresses used by each processor are not available because the controlling processor is absent. Therefore, for example, it is unclear which address is assigned to which memory, and the addresses become unusable.

[0006] In addition, in a storage device, in addition to a failure, it is sometimes necessary to stop the processor of the controller due to updating the OS (operating system) of the controller. In the storage device, I / O (Input / Output) communication is performed to input and output data with a host device. The protocol used in the I / O communication with the host device is, for example, Fiber Channel.

[0007] The controller is equipped with a front-end interface together with the processor, and the front-end interface includes a protocol chip for controlling the protocol used for such I / O communication. In such a situation, if the processor of the controller is stopped, the processor connected to the front-end interface does not exist and does not respond to the I / O request from the host device. When viewed from the host device, the storage device temporarily appears to be shut down. In order to recover, the host device needs to be reconnected to the remaining controller that is still running.

[0008] In order to solve this problem, it is conceivable that the controller automatically reconnects the I / O request from the host device to the controller having a processor that has not been stopped due to OS update or the like.

[0009] For example, in the configuration described in Patent Document 1, two processors for controlling the system are mounted. Therefore, when the processor of one controller stops, it is considered that an I / O request from the host device is transferred to the processor of the other controller.

[0010] Patent Document 2 discloses an example of a storage device including a local router having a function of automatically allocating access destinations from a front-end interface to each controller, thereby automatically switching I / O communication even when a processor of a controller stops.

[0011] Prior art literature

[0012] Patent Document 1: U.S. Patent No. 8,700,856

[0013] Patent Document 2: U.S. Patent No. 8,423,677 Summary of the invention

[0014] Problems to be solved by the invention

[0015] In the technology described in Patent Document 1, each controller is equipped with at most one NTB, and each front-end interface uses the address controlled by the processor of each controller, and does not use the address controlled by another processor. Therefore, when the processor of the controller stops, its address becomes unusable, and there is no address that can be used, so each front-end interface cannot pass the request from the host device to another processor.

[0016] In addition, in the technology described in Patent Document 2, the local router requires complex control, so there may be a case where some processor is installed to perform program control. In this case, when the OS of the local router itself is updated, the local router needs to be stopped and restarted. In this case, the function of automatically switching the I / O communication with the host device based on the local router cannot be used, and the I / O communication with the host device is cut off.

[0017] Therefore, a technology is desired that can continue to respond to I / O requests from a host device without providing a special control function unit such as a local router when a processor of a controller stops.

[0018] Means for solving problems

[0019] One embodiment of the present invention is a storage device for processing requests from a host device, wherein the storage device includes multiple storage controllers, wherein a first storage controller among the multiple storage controllers includes: a first front-end interface, which controls a communication protocol with the host device; and a first processor, which controls the storage device, and a second storage controller among the multiple storage controllers includes a second processor that controls the storage device, and the first storage controller also includes: a first address conversion unit, which converts a first address used by the first processor and a second address used by the first front-end interface; a second address conversion unit, which converts a third address used by the second processor and the second address used by the first front-end interface; a first outbound queue, which controls the transmission of data from the first front-end interface to the host device through the first address conversion unit. The first front-end interface includes a first processor, a first in-queue, and a second in-queue. The first front-end interface includes a first processor, a first processor, and a first processor. The first processor includes a first processor, a first processor, and a first processor. The first processor includes a first processor, a first processor, and a first processor. The first processor includes a first processor, a first processor, and a first processor. The first processor includes a first processor, a first processor, and a first processor. The first processor includes a first processor, a first processor, and a first processor. The first processor includes a first processor, a first processor, and a first processor. The first processor includes a first processor, a first processor, and a first processor.

[0020] Effects of the Invention

[0021] According to one aspect of the present invention, even if the processor of one controller constituting the dual controller stops processing, the processor of the other controller can take over the I / O request from the host device and continue to respond. Other issues, structures, and effects will become clear from the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a diagram showing a storage device according to the first embodiment.

[0023] Figure 2 This is a diagram showing the structure of the controller of Example 1.

[0024] Figure 3 It is a diagram for explaining the structure of OQ and IQ in the first embodiment.

[0025] Figure 4 This is a diagram for explaining a normal data transmission path between a front-end interface and a memory in the storage device of the first embodiment.

[0026] Figure 5 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device of the first embodiment is switched.

[0027] Figure 6 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device of the first embodiment is switched.

[0028] Figure 7 This is a diagram for explaining the data transfer sequence in the normal state in the storage device of the first embodiment.

[0029] Figure 8 This is a diagram for explaining the data transfer sequence after the enqueue destination queue of the front-end interface in the storage device of the first embodiment is switched.

[0030] Fig. 9 This is a diagram for explaining the flow when one controller is restarted in the storage device of the first embodiment.

[0031] Fig.10 This is a diagram for explaining a normal data transmission path between a front-end interface and a memory in the storage device of the second embodiment.

[0032] Fig.11 This is a diagram for explaining the flow when one controller is restarted in the storage device of the second embodiment.

[0033] Fig.12 This is a diagram for explaining the structure of a controller in the storage device of the third embodiment.

[0034] Fig.13 This is a diagram for explaining a normal data transmission path between a front-end interface and a memory in a storage device according to a third embodiment.

[0035] Fig.14 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device of the third embodiment is switched.

[0036] Fig.15 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device of the third embodiment is switched. DETAILED DESCRIPTION

[0037] Hereinafter, several embodiments of the present invention will be described with reference to the accompanying drawings. In addition, in all embodiments, the components marked with the same symbols are substantially the same. In addition, the processing performed by the processor is appropriately performed using storage resources (e.g., memory) and communication interface devices (e.g., communication ports), so the subject of the processing may also be the processor. The processor may also have dedicated hardware in addition to the CPU (Central Processing Unit).

[0038] [Example 1]

[0039] Reference Figures 1 to 9 , the storage device of Example 1 is described.

[0040] Figure 1 This is a diagram showing a storage device according to the first embodiment.

[0041] The storage device 100 includes controllers 110 and 120 and a drive box 130 having a plurality of hard disk drives or a plurality of solid state disk drives.

[0042] The controller 110 includes a processor 111, a memory 112, a front-end interface 113, a back-end interface 114, a PCIe (PCI Express, PCI bus) switch 115, and a management processor 116. Similarly, the controller 120 has a processor 121, a memory 122, a front-end interface 123, a back-end interface 124, a PCIe switch 125, and a management processor 126. The processors 111 and 121 each have a plurality of processor cores (not shown) inside.

[0043] A host device (not shown) accessing the storage device 100 is connected to the storage device 100 via the front-end interfaces 113 and 123. The host device and the front-end interfaces 113 and 123 are connected via a transmission line such as a fiber channel cable or an Ethernet cable.

[0044] Alternatively, the host device and the front-end interface 113, 123 may be connected via a storage area network composed of a plurality of transmission lines and a plurality of switches. The front-end interface 113, 123 converts the data transmission protocol between the host device and the storage device 100 and the data transmission protocol within the controller 110, 120. The front-end interface 113, 123 includes a protocol chip.

[0045] The driver box 130 is connected to the controllers 110 and 120 via the back-end interfaces 114 and 124. The back-end interfaces 114 and 124 convert the data transmission protocol within the controllers 110 and 120 and the data transmission protocol between the controllers 110 and 120 and the driver box 130. In addition, in the case where the drive in the driver box is a PCIe-connected NVMe SSD, the back-end interfaces 114 and 124 are PCIe switches that do not perform protocol conversion.

[0046] The processors 111 and 121 control data transmission between the host device connected via the front-end interfaces 113 and 123 and the drive cartridge 130 connected via the back-end interfaces 114 and 124. Furthermore, the processors 111 and 121 control data transmission between controllers.

[0047] Memories 112 and 122 are main memories of processors 111 and 121, respectively, and store programs (such as storage control programs) executed by processors 111 and 121 and management tables referenced by processors 111 and 121. Memories 112 and 122 are also used as cache memories of controllers 110 and 120, respectively.

[0048] Figure 2 It is a diagram showing the structure of the controllers 110 and 120 according to the first embodiment.

[0049] exist Figure 2 In order to simplify the description, Figure 1 The back-end interfaces 114, 124 in.

[0050] The processors 111 and 121 have NTBs 211 and 231, respectively. The NTB 211 and the NTB 231 are connected via a link 140. The processors 111 and 121 can communicate with each other via the link 140. In this way, in the storage device 100, a dual controller is formed by the two controllers 110 and 120. The processors 111 and 121 transmit user data and control data of the controllers, which are duplicated between the two controllers, on the link 140.

[0051] The front-end interface 113 has interrupt setting registers 217 and 218. The front-end interface 113 can send an interrupt such as MSI (Message Signaled Interrupt) to the addresses set in the interrupt setting registers 217 and 218. For example, the address of the interrupt controller (not shown) built into the processor 111 is set in the interrupt setting register 217, and the address of the interrupt controller (not shown) built into the processor 121 is set in the interrupt setting register 218. Therefore, the front-end interface 113 can send an interrupt to the processor 111 or 121. These interrupts are used in association with error notification from the front-end interface to the processor and the operation of the control queue for data transfer described later.

[0052] Likewise, the front-end interface 123 includes interrupt setting registers 237 and 238. The front-end interface 123 can send an interrupt such as an MSI to the addresses set in the interrupt setting registers 237 and 238. For example, the address of the interrupt controller built into the processor 121 is set in the interrupt setting register 237, and the address of the interrupt controller built into the processor 111 is set in the interrupt setting register 238. Therefore, the front-end interface 123 can send an interrupt to the processor 111 or 121.

[0053] Furthermore, the interrupts sent by the front-end interfaces 113 and 123 to the processors 111 and 121 use only MSI or MSI-X that can pass through the NTB as PCIe write requests, and do not use traditional interrupts (INTx).

[0054] The PCIe switch 115 has NTBs 214 and 215. The NTB 214 is connected to the port 212 of the processor 111 via the link 117. The NTB 215 is connected to the port 233 of the processor 121 via the link 128.

[0055] Likewise, the PCIe switch 125 has NTBs 234 and 235. The NTB 234 is connected to the port 232 of the processor 121 via the link 127. The NTB 235 is connected to the port 213 of the processor 111 via the link 118.

[0056] The port 213 of the processor 111 and the port 233 of the processor 121 set the DPC (Downstream Port Containment) function of PCIe to be valid. Therefore, for example, when the controller 110 is removed from the storage device 100, even if the links 118 and 128 are disconnected, the controller 120 remaining in the storage device 100 does not need to treat the link disconnection (linkdown) as an error, and can continue to operate.

[0057] Similarly, for example, when the controller 120 is removed from the storage device 100, even if the links 118 and 128 are disconnected, the controller 110 remaining in the storage device 100 does not need to process the link disconnection as an error, and can therefore continue to operate.

[0058] The management processor 116 connects its root port 216 to the PCIe switch 115 , and functions as a root complex for the front-end interface 113 and the PCIe switch 115 .

[0059] Similarly, the management processor 126 connects its root port 236 to the PCIe switch 125 , and functions as a root complex for the front-end interface 123 and the PCIe switch 125 .

[0060] The management processor 116 performs initial settings of the front-end interface 113 and the PCIe switch 115 including the NTBs 214 and 215 independently of data transfer control between the host device and the storage device 100. The replacement function of the management processor 116 may be built into the PCIe switch 115.

[0061] Likewise, the management processor 126 performs initial settings of the front-end interface 123 and the PCIe switch 125 including the NTBs 234 and 235 independently of data transfer control between the host device and the storage device 100. The replacement function of the management processor 126 may be built into the PCIe switch 125.

[0062] The memories 112 and 122 have queues, namely OQ (Outbound Queue) and IQ (Inbound Queue), for controlling data transmission between the processors 111 and 121 and the front-end interfaces 113 and 123. The OQ is a queue for controlling data transmission from the front-end interface to the processor, and the IQ is a queue for controlling data transmission from the processor to the front-end interface.

[0063] The memory 112 has OQ 201 and IQ 202 for controlling data transfer between the processor 111 and the front-end interface 113. In addition, the memory 112 has OQ 203 and IQ 204 for controlling data transfer between the processor 111 and the front-end interface 123.

[0064] Similarly, the memory 122 has OQ 221 and IQ 222 for controlling data transfer between the processor 121 and the front end interface 123. In addition, the memory 122 has OQ 223 and IQ 224 for controlling data transfer between the processor 121 and the front end interface 113. Thus, in the storage device of the first embodiment, the control queue of the front end interface is arranged across two controllers.

[0065] In addition, Figure 2 , OQ201 is displayed as "OQ00", IQ202 is displayed as "IQ00", OQ203 is displayed as "OQ01", IQ204 is displayed as "IQ01", OQ221 is displayed as "OQ11", IQ222 is displayed as "IQ11", OQ223 is displayed as "OQ10", and IQ224 is displayed as "IQ10".

[0066] In addition, the front end interface 113 can switch between using OQ201 and IQ202 or using OQ223 and IQ224 according to an instruction from the processor 111. Similarly, the front end interface 123 can switch between using OQ221 and IQ222 or using OQ203 and IQ204 according to an instruction from the processor 121.

[0067] The instruction of the processor is executed, for example, by the processor writing a predetermined value to a PCIe register of the front-end interface using a PCIe write request.

[0068] The above-mentioned switching setting status is stored in registers of the front-end interfaces 113 and 123 that can be accessed via PCIe, and can be read by the processors 111 and 121 .

[0069] The front-end interfaces 113 and 123 are connected to the processors 111 and 121 respectively via NTB. Therefore, even if the processor 111 or 121 stops and the links 117 and 118 are disconnected or the links 127 and 128 are disconnected, the links between the front-end interfaces 113 and 123 and the PCIe switches 115 and 125 will not be disconnected. That is, the state of the link connecting the front-end interfaces 113 and 123 and the PCIe switches 115 and 125 is not affected by the stop of the processors 111 and 121.

[0070] Figure 3 It is a diagram for explaining the structure of OQ and IQ in the first embodiment.

[0071] OQ201 and IQ202 both store entries in a total of N elements from the 0th to the N-1th. In OQ201, the content of the entry is, for example, a host I / O instruction received from the host device. In IQ202, the content of the entry is, for example, a response corresponding to a completed host I / O instruction, and a data transfer list indicated by the processor to the front-end interface. In addition, each entry contains identification information (i.e., exchange ID) indicating which exchange of the host I / O the entry is associated with. An exchange represents a series of operations associated with a read action, a write action, etc. between the host and the memory.

[0072] exist Figure 3In OQ201, as an example, entries are stored in the i-1th element to the i+5th element. The other entries of OQ201 are empty. OQ_PI (Producer Index) 301 indicates the position of the element where the front-end interface will store the entry next. OQ_CI (Consumer Index) 302 indicates the position of the element where the entry that the processor will read next is stored. In the case where OQ_PI301 and OQ_CI302 represent the same element, it means that an unprocessed entry in OQ201 is not stored and is in an empty state. In addition, the latest entry that has been processed is stored in the i-1th element. The processor can determine which exchange the processing of the entry associated with has been completed by investigating the exchange ID of the entry.

[0073] exist Figure 3 In IQ202, as an example, entries are stored in the j-1th to i+4th elements. The other entries of IQ202 are empty. IQ_PI311 indicates the position of the element where the processor is going to store the entry next. IQ_CI312 indicates the position of the element where the entry to be read out by the front-end interface next is stored. In the case where IQ_PI311 and IQ_CI312 represent the same element, it means that an unprocessed entry is not stored in IQ202 and is in an empty state. In addition, the latest entry that has been processed is stored in the j-1th element. The processor can determine which exchange the entry associated with has been processed by investigating the exchange ID of the entry.

[0074] Figure 4 This is a diagram for explaining a normal data transmission path between the front-end interface and the memory in the storage device 100 of the first embodiment.

[0075] Here, it is assumed that the front-end interface 113 and the processor 111 use Figure 2 The OQ201 and IQ202, the front-end interface 123 and the processor 121 use Figure 2 The situation of OQ221 and IQ222.

[0076] The front-end interface 113 accesses the memory 112 via a data transmission path 400 passing through the PCIe switch 115, the NTB 214, the link 117, and the processor 111. The front-end interface 123 accesses the memory 122 via a data transmission path 401 passing through the PCIe switch 125, the NTB 234, the link 127, and the processor 121. In addition, between the memory 112 and the memory 122, the processors 111 and 121 transmit data via a data transmission path 402 passing through the link 140.

[0077] The front-end interface 113 sends an interrupt to the processor 111 via the PCIe switch 115 , the NTB 214 , and the link 117 . The front-end interface 123 sends an interrupt to the processor 121 via the PCIe switch 125 , the NTB 234 , and the link 127 .

[0078] Figure 5 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device 100 of the first embodiment is switched.

[0079] Here, it is assumed that the front-end interface 113 and the processor 121 use Figure 2 The OQ223 and IQ224, the front-end interface 123 and the processor 121 use Figure 2 The situation of OQ221 and IQ222.

[0080] The front-end interface 113 accesses the memory 122 via a data transmission path 500 passing through the PCIe switch 115, the NTB 215, the link 128, and the processor 121. The front-end interface 123 accesses the memory 122 via a data transmission path 401 passing through the PCIe switch 125, the NTB 234, the link 127, and the processor 121. In addition, between the memory 112 and the memory 122, the processors 111 and 121 transmit data via a data transmission path 402 passing through the link 140.

[0081] The front-end interface 113 sends an interrupt to the processor 121 via the PCIe switch 115, the NTB 215, and the link 128. The front-end interface 123 sends an interrupt to the processor 121 via the PCIe switch 125, the NTB 234, and the link 127.

[0082] Figure 6 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device 100 of the first embodiment is switched.

[0083] Here, it is assumed that the front-end interface 113 and the processor 111 use Figure 2 The OQ201 and IQ202, the front-end interface 123 and the processor 111 use Figure 2 The situation of OQ203 and IQ204.

[0084] The front-end interface 113 accesses the memory 112 via a data transmission path 400 passing through the PCIe switch 115, the NTB 214, the link 117, and the processor 111. The front-end interface 123 accesses the memory 112 via a data transmission path 600 passing through the PCIe switch 125, the NTB 235, the link 118, and the processor 111. In addition, between the memory 112 and the memory 122, the processors 111 and 121 transmit data via a data transmission path 402 passing through the link 140.

[0085] The front-end interface 113 sends an interrupt to the processor 111 via the PCIe switch 115, the NTB 214, and the link 117. The front-end interface 123 sends an interrupt to the processor 111 via the PCIe switch 125, the NTB 235, and the link 118.

[0086] Figure 7 1 is a diagram showing a data transfer sequence in a normal state in the storage device 100 according to the first embodiment.

[0087] Here, as an example, a case where the front end interface 113 and the processor 111 use OQ 201 and IQ 202 will be described.

[0088] First, the host device 700 sends a host I / O command 701 to the front end interface 113 .

[0089] The front-end interface 113 that has received the host I / O command 701 enqueues an entry 702 containing the content of the command into the OQ 201 (step 703 ).

[0090] Next, the front end interface 113 sends an interrupt to the address set in the interrupt setting register 217 to notify the processor 111 that an entry has been enqueued in the OQ (step 704). The front end interface 113 also updates the OQ_PI of the OQ 201 in the memory 112 (step 705).

[0091] The processor 111 that has received the interruption reads out the entry storing the content of the host I / O command from the OQ 201 (step 706 ) The processor 111 also updates the OQ_CI of the OQ 201 in the front end interface 113 (step 707 ).

[0092] Next, the processor 111 queues an entry including a data transfer list corresponding to the host I / O command 701 to the IQ 202 (step 708 ). Furthermore, the processor 111 updates the IQ_PI of the IQ 202 in the front-end interface 113 (step 709 ).

[0093] The front-end interface 113 having updated the IQ_PI reads out the entry containing the data transfer list from the IQ 202 (step 710).

[0094] Next, the front end interface 113 performs data transfer between the host device 700 and the memory 112 according to the data transfer list included in the entry read from the IQ 202 (step 711 ).

[0095] When the data transmission is completed, the front end interface 113 updates the IQ_CI of the IQ 202 in the memory 112 (step 712).

[0096] In this way, the processor 111 is able to process the host I / O instruction 701 received by the front-end interface 113 .

[0097] Figure 8 This is a diagram for explaining the data transfer sequence after the enqueue destination queue of the front-end interface in the storage device 100 of the first embodiment is switched.

[0098] Here, as an example, a case where the front end interface 113 and the processor 121 use the OQ 223 and the IQ 224 will be described.

[0099] First, the host device 700 sends a host I / O command 801 to the front end interface 113 .

[0100] The front-end interface 113 that has received the host I / O command 801 enqueues an entry 802 containing the content of the command into the OQ 223 (step 803 ).

[0101] Next, the front end interface 113 sends an interrupt to the address set in the interrupt setting register 218 to notify the processor 121 that an entry has been enqueued in the OQ (step 804). The front end interface 113 also updates the OQ_PI of the OQ 223 in the memory 122 (step 805).

[0102] The processor 121 that has received the interruption reads out the entry storing the content of the host I / O command from the OQ 223 (step 806 ) The processor 121 also updates the OQ_CI of the OQ 223 in the front end interface 113 (step 807 ).

[0103] Next, the processor 121 queues an entry including a data transfer list corresponding to the host I / O command 801 to the IQ 224 (step 808 ). Furthermore, the processor 121 updates the IQ_PI of the IQ 224 in the front-end interface 113 (step 809 ).

[0104] The front-end interface 113 having updated the IQ_PI reads out the entry containing the data transfer list from the IQ 224 (step 810).

[0105] Next, the front end interface 113 performs data transfer between the host device 700 and the memory 122 according to the data transfer list included in the entry read from the IQ 224 (step 811 ).

[0106] When the data transmission is completed, the front end interface 113 updates the IQ_CI of the IQ 224 in the memory 122 (step 812).

[0107] In this way, the processor 121 is able to process the host I / O instruction 801 received by the front-end interface 113 .

[0108] Fig. 9 This is a diagram for explaining the flow when one controller is restarted in the storage device 100 of the first embodiment.

[0109] Here, as an example, a process flow will be described in which the processor 121 of the controller 120 takes over the processing of the host I / O received by the front-end interface 113 when the controller 110 is restarted.

[0110] First, the processor 111 confirms whether the controller 120 is being restarted (step 901 ).

[0111] Next, in step 902, if the controller 120 is in the process of restarting, the process returns to step 901. If the controller 120 is not in the process of restarting, the process proceeds to step 903.

[0112] Next, the processor 111 notifies the processor 121 of the start of the restart process of the controller 110 (step 903 ).

[0113] Next, the processor 111 sends a queue switching instruction to the front end interface 113 to switch the operation target queue from OQ201 and IQ202 to OQ223 and IQ224 (step 904 ).

[0114] The front-end interface 113 that has received the queue switching instruction enqueues the host I / O instruction received thereafter to the OQ 223 (step 905). In other words, the front-end interface 113 switches the OQ of the enqueue destination according to the newly received exchange. Whether the exchange is newly started can be determined by checking the exchange ID included in the host I / O instruction.

[0115] That is, the front-end interface 113 enqueues into OQ223 the host I / O instructions that initially switched the exchange ID (i.e., the initial host I / O instruction assigned a new exchange ID) and the host I / O instructions that come after it, among the host I / O instructions received from the host device 700 after receiving the queue switching instruction from the processor 111.

[0116] Next, the processor 111 waits for OQ201 and IQ202 to become empty (step 906). When OQ_PI and OQ_CI of OQ201 are equal, the processor 111 can determine that OQ201 is empty. Similarly, when IQ_PI and IQ_CI of IQ202 are equal, the processor 111 can determine that IQ202 is empty. If OQ201 and IQ202 are empty, it can be known that there is no unfinished host I / O that is received by the front-end interface 113 and is associated with OQ201 and should be processed by the processor 111.

[0117] Furthermore, in order to make it reliable that there is no unfinished host I / O associated with OQ201, the processor 111 may also check whether the exchange IDs included in the entry corresponding to the host I / O instruction most recently processed in OQ201 and the entry corresponding to the response most recently processed in IQ202 are consistent. When OQ201 and IQ202 are empty and the two exchange IDs are consistent, it can be known that there is no unfinished host I / O that is received as the front-end interface 113 and that the processor 111 should process.

[0118] Next, the storage device 100 shuts down the controller 110 and performs necessary processing such as OS update. Then, the storage device 100 continues to operate only in one controller 120. After that, the storage device 100 restarts the controller 110. Accordingly, the processor 111 stops and restarts (step 907).

[0119] After the controller 110 is restarted, the processor 111 confirms the queue switching setting status of the front-end interface 113 (step 908 ).

[0120] Next, in step 909, if the host I / O instruction enqueue destination is OQ 223, the process proceeds to step 910. Otherwise, the process ends.

[0121] Next, the processor 111 notifies the processor 121 that the controller 110 in the state where the operation target queue of the front-end interface 113 is switched is restarted (step 910 ).

[0122] Next, the processor 111 sends a queue switching instruction to the front end interface 113 to switch the operation target queues from OQ223 and IQ224 to OQ201 and IQ202 (step 911 ).

[0123] The front-end interface 113 that has received the queue switching instruction enqueues the host I / O instruction subsequently received with the new switch ID into the OQ 201 (step 912 ).

[0124] Next, the processor 111 waits for OQ223 and IQ224 to become empty (step 913). When OQ_PI and OQ_CI of OQ223 are equal, the processor 111 can determine that OQ223 is empty. Similarly, when IQ_PI and IQ_CI of IQ224 are equal, the processor 111 can determine that IQ224 is empty. The fact that OQ223 and IQ224 are empty means that there is no uncompleted host I / O received by the front-end interface 113 that the processor 121 should process.

[0125] Furthermore, in order to make the situation where there is no unfinished host I / O associated with OQ223 reliable, processor 111 may also investigate whether the exchange IDs contained in the entry corresponding to the most recently processed host I / O instruction in OQ223 and the entry corresponding to the most recently processed response in IQ224 are consistent.

[0126] Finally, the processor 111 notifies the processor 121 that the restart process of the controller 110 is completed (step 914).

[0127] As described above, the storage device 100 of the first embodiment restarts the controller after the switching process of the enqueue destination OQ based on the front-end interface is completed and the host I / O being processed before the switching is completed. Therefore, even if one controller is restarted, the host I / O is not interrupted.

[0128] In the storage device 100 of the first embodiment, an example is described in which each front-end interface is configured with a pair of OQ and IQ for each controller, but the storage device 100 may be configured with more than two pairs of OQ and IQ. For example, when each front-end interface is configured with two pairs of OQ and IQ for each controller, the two pairs are used as the unit of switching processing.

[0129] In addition, for example, a case where a failure occurs in the processor 111, the memory 112, etc. of the controller 110 can be assumed. Even in such a case, the influence thereof will not be transmitted to the front-end interface 113 and the controller 120 through the NTB 211, 214, and 215. However, when a failure occurs in the processor 111, the processor 111 cannot send a queue switching instruction to the front-end interface 113. In this case, the processor 121 that detects the failure of the processor 111 through the heartbeat of the link 140 sends an instruction to the front-end interface 113 via the link 128, the NTB 215, and the PCIe switch 115 to switch the enqueue destination of the host I / O instruction to the normal controller 120 side. As a result, the storage device 100 of the first embodiment can continue to operate. However, in this case, it is necessary to send again the unfinished I / O that is being processed in the controller 110 where the host device has failed.

[0130] [Example 2]

[0131] Next, refer to Figure 10 to Figure 11 The storage device of the second embodiment is described. The storage device of the second embodiment has the same structure as the storage device of the second embodiment except for the differences described below. Figures 1 to 9 The storage device of the illustrated embodiment 1 is the same, and therefore the description thereof is omitted.

[0132] The storage device 100 of the second embodiment uses both the group of OQ201 and IQ202 and the group of OQ223 and IQ224 to control data transfer between the front-end interface 113 and the processors 111 and 121 in a normal state. In this case, either the processors 111 and 121 receive the host I / O command received by the front-end interface 113 via OQ201 or OQ223. Then, according to the instruction from the processor 111, it is possible to switch between using the group of OQ201 and IQ202 or using the group of OQ223 and IQ224 during data transfer.

[0133] Similarly, the storage device 100 of the second embodiment uses both the group of OQ221 and IQ222 and the group of OQ203 and IQ204 in normal state to control data transfer between the front-end interface 123 and the processors 111 and 121. In this case, either the processors 111 and 121 receive the host I / O command received by the front-end interface 123 via OQ221 or OQ203. In addition, according to the instruction from the processor 121, it is possible to switch between using the group of OQ221 and IQ222 or using the group of OQ203 and IQ204 during data transfer.

[0134] In addition, the switching process in the storage device of the second embodiment is called a "queue offset" process.

[0135] Fig.10 This is a diagram for explaining a normal data transmission path between a front-end interface and a memory in the storage device of the second embodiment.

[0136] Here, it is assumed that the front-end interface 113 and the processor 111 use Figure 2 The OQ201 and IQ202, the front-end interface 113 and the processor 121 use Figure 2 The situation of OQ223 and IQ224.

[0137] In addition, it is assumed that the front-end interface 123 and the processor 121 use Figure 2 The OQ221 and IQ222, the front-end interface 123 and the processor 111 use Figure 2 The situation of OQ203 and IQ204.

[0138] Front end interface 113 accesses memory 112 via data transmission path 400 through PCIe switch 115, NTB 214, link 117, and processor 111. In addition, front end interface 113 accesses memory 122 via data transmission path 1000 through PCIe switch 115, NTB 215, link 128, and processor 121.

[0139] Front end interface 123 accesses memory 122 via data transmission path 401 through PCIe switch 125, NTB 234, link 127, and processor 121. In addition, front end interface 123 accesses memory 112 via data transmission path 1001 through PCIe switch 125, NTB 235, link 118, and processor 111.

[0140] Furthermore, between the memory 112 and the memory 122 , the processors 111 and 121 transmit data via a data transmission path 402 through the link 140 .

[0141] The front-end interface 113 sends an interrupt to the processor 111 via the PCIe switch 115, the NTB 214, and the link 117. The front-end interface 113 sends an interrupt to the processor 121 via the PCIe switch 115, the NTB 215, and the link 128. The front-end interface 123 sends an interrupt to the processor 111 via the PCIe switch 125, the NTB 235, and the link 118. The front-end interface 123 sends an interrupt to the processor 121 via the PCIe switch 125, the NTB 234, and the link 127.

[0142] Fig.11 This is a diagram for explaining the flow when one controller is restarted in the storage device of the second embodiment.

[0143] Here, as an example, a process flow will be described in which the processor 121 of the controller 120 takes over the processing of the host I / O received by the front-end interface 113 when the controller 110 is restarted.

[0144] First, the processor 111 confirms whether the controller 120 is being restarted (step 1101 ).

[0145] Next, in step 1102, if the controller 120 is in the process of restarting, the process returns to step 1101. If the controller 120 is not in the process of restarting, the process proceeds to step 1103.

[0146] Next, the processor 111 notifies the processor 121 of the start of the restart process of the controller 110 (step 1103 ).

[0147] Next, the processor 111 sends a queue offset instruction for switching the operation target queues to only OQ223 and IQ224 to the front end interface 113. The processor 121 sends a queue offset instruction for switching the operation target queues to only OQ221 and IQ223 to the front end interface 123 (step 1104).

[0148] The front-end interface 113 that has received the queue offset instruction will only queue the host I / O instructions received with the new switch ID to OQ 223. The front-end interface 123 that has received the queue offset instruction will only queue the host I / O instructions received with the new switch ID to OQ 221 (step 1105).

[0149] Next, the processor 111 waits for OQ201, IQ202, OQ203, and IQ204 to become empty (step 1106). When OQ_PI and OQ_CI of OQ201 are equal, the processor 111 can determine that OQ201 is empty. Similarly, when IQ_PI and IQ_CI of IQ202 are equal, the processor 111 can determine that IQ202 is empty. If OQ201 and IQ202 are empty, it can be known that there is no unfinished host I / O that is received by the front-end interface 113 and is associated with OQ201 and should be processed by the processor 111.

[0150] Furthermore, in order to make it reliable that there is no uncompleted host I / O associated with OQ201, the processor 111 may also check whether the exchange IDs included in the entry corresponding to the host I / O instruction most recently processed in OQ201 and the entry corresponding to the response most recently processed in IQ202 are consistent. When OQ201 and IQ202 are empty and the two exchange IDs are consistent, it can be known that there is no uncompleted host I / O that is received by the front-end interface 113 and is associated with OQ201 and IQ202 and should be processed by the processor 111.

[0151] In addition, when OQ_PI and OQ_CI of OQ203 are equal, the processor 111 can determine that OQ203 is empty. Similarly, when IQ_PI and IQ_CI of IQ204 are equal, the processor 111 can determine that IQ204 is empty. If OQ203 and IQ204 are empty, it can be known that there is no unfinished host I / O that is received by the front-end interface 123 and is associated with OQ203 and should be processed by the processor 111.

[0152] Furthermore, in order to make it reliable that there is no unfinished host I / O associated with OQ203, the processor 111 may also check whether the exchange IDs included in the entry corresponding to the host I / O instruction most recently processed in OQ203 and the entry corresponding to the response most recently processed in IQ204 are consistent. When OQ203 and IQ204 are empty and the two exchange IDs are consistent, it can be known that there is no unfinished host I / O that is received by the front-end interface 123 and is associated with OQ203 and IQ204 and should be processed by the processor 111.

[0153] Next, the storage device 100 of the second embodiment shuts down the controller 110 and performs necessary processing such as OS update. Then, the storage device 100 of the second embodiment continues to operate only in one controller 120. With this shutdown process, the processor 111 stops. Thereafter, the storage device 100 restarts the controller 110. Accordingly, the processor 111 stops and restarts (step 1107).

[0154] After the controller 110 is restarted, the processor 111 confirms the bias setting status of the front-end interface 113 (step 1108 ).

[0155] Next, in step 1109, if the host I / O instruction enqueue destination is only OQ 223, the process proceeds to step 1110. Otherwise, the process ends.

[0156] Next, the processor 111 notifies the processor 121 that the controller 110 is restarted in a state where the operation target queue of the front-end interface 113 is offset (step 1110 ).

[0157] Next, the processor 111 sends an instruction to release the queue bias state to the front-end interface 113. The processor 121 sends an instruction to release the queue bias state to the front-end interface 123 (step 1111).

[0158] The front-end interface 113 that has received the queue offset release command will enqueue the host I / O commands subsequently received with the new exchange ID to OQ201 or OQ223. The front-end interface 123 that has received the queue offset release command will enqueue the host I / O commands subsequently received with the new exchange ID to OQ221 or OQ203 (step 1112).

[0159] Finally, the processor 111 notifies the processor 121 that the restart process of the controller 110 is completed (step 1113).

[0160] As described above, the storage device 100 of Example 2 restarts the controller after the offset processing of the enqueue destination OQ performed by the front-end interface is completed and the host I / O that was being processed before the switching of the enqueue destination OQ based on the offset processing is completed. Even if the controller is restarted, the host I / O will not be interrupted.

[0161] In the storage device 100 of the second embodiment, an example is described in which each front-end interface is configured with a pair of OQ and IQ for each controller, but the storage device 100 may be configured with two or more pairs of OQ and IQ. For example, when each front-end interface is configured with two pairs of OQ and IQ for each controller, the two pairs are used as the unit of the offset processing.

[0162] [Example 3]

[0163] Next, refer to Figures 12 to 15 The storage device of the third embodiment is described. The storage device of the third embodiment has the same structure as the storage device of the third embodiment except for the differences described below. Figures 1 to 9 The storage device of the illustrated embodiment 1 is the same, and therefore the description thereof is omitted.

[0164] Fig.12 This is a diagram for explaining the structure of a controller in the storage device of the third embodiment.

[0165] exist Fig.12 In order to simplify the description, Figure 2 The back-end interface is also omitted. The storage device of the third embodiment is characterized in that there is no equivalent Figure 2 The link 140 in the connection processor link ( Figure 2 Instead, processor 111 and processor 121 perform data transmission between processors via a path through links 118, 128 and PCIe switches 115, 125.

[0166] Fig.13 This is a diagram for explaining a normal data transmission path between a front-end interface and a memory in a storage device according to a third embodiment.

[0167] In normal operation, the front-end interface 113 and the processor 111 use Fig.12 The OQ201 and IQ202, the front-end interface 113 and the processor 121 use Fig.12 The OQ223 and IQ224 are used to control data transmission.

[0168] The front-end interface 113 accesses the memory 112 via a data transmission path 1300 through the PCIe switch 115, the NTB 214, the link 117, and the processor 111. The front-end interface 123 accesses the memory 122 via a data transmission path 1301 through the PCIe switch 125, the NTB 234, the link 127, and the processor 121. In addition, between the memory 112 and the memory 122, the processor 111 and the processor 121 transmit data via a data transmission path 1302 through the processor 111, the link 117, the NTB 214, the PCIe switch 115, the NTB 215, the link 128, and the processor 121. In addition, between the memory 112 and the memory 122, the processor 111 and the processor 121 transmit data via a data transmission path 1303 through the processor 121, the link 127, the NTB 234, the PCIe switch 125, the NTB 235, the link 118, and the processor 111. The two data transfer paths 1302 and 1303 between processors may be used to transfer user data with a large amount of data and control data with a small amount of data for storage devices. This can reduce the time required for control data transfer.

[0169] The front-end interface 113 sends an interrupt to the processor 111 via the PCIe switch 115 , the NTB 214 , and the link 117 . The front-end interface 123 sends an interrupt to the processor 121 via the PCIe switch 125 , the NTB 234 , and the link 127 .

[0170] Fig.14 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device of the third embodiment is switched.

[0171] Here, it is assumed that the front-end interface 113 and the processor 121 use Fig.12 The OQ223 and IQ224, the front-end interface 123 and the processor 121 use Fig.12 The situation of OQ221 and IQ222.

[0172] The front-end interface 113 accesses the memory 122 via a data transmission path 1400 through the PCIe switch 115, the NTB 215, the link 128, and the processor 121. The front-end interface 123 accesses the memory 122 via a data transmission path 1301 through the PCIe switch 125, the NTB 234, the link 127, and the processor 121. In addition, between the memory 112 and the memory 122, the processor 111 and the processor 121 transmit data via a data transmission path 1303 through the processor 121, the link 127, the NTB 234, the PCIe switch 125, the NTB 235, the link 118, and the processor 111. In addition, between the memory 112 and the memory 122, the processor 111 and the processor 121 may also transmit data via a data transmission path 1302 ( Fig.13 ) to transmit data.

[0173] The front-end interface 113 sends an interrupt to the processor 121 via the PCIe switch 115, the NTB 215, and the link 128. The front-end interface 123 sends an interrupt to the processor 121 via the PCIe switch 125, the NTB 234, and the link 127.

[0174] Fig.15 This is a diagram for explaining a data transmission path after the enqueue destination queue of the front-end interface in the storage device of the third embodiment is switched.

[0175] Here, it is assumed that the front-end interface 113 and the processor 111 use Fig.12 The OQ201 and IQ202, the front-end interface 123 and the processor 111 use Fig.12 The situation of OQ203 and IQ204.

[0176] The front-end interface 113 accesses the memory 112 via a data transmission path 1300 passing through the PCIe switch 115, the NTB 214, the link 117, and the processor 111. The front-end interface 123 accesses the memory 112 via a data transmission path 1500 passing through the PCIe switch 125, the NTB 235, the link 118, and the processor 111. In addition, between the memory 112 and the memory 122, the processor 111 and the processor 121 transmit data via a data transmission path 1302 passing through the processor 111, the link 117, the NTB 214, the PCIe switch 115, the NTB 215, the link 128, and the processor 121. In addition, between the memory 112 and the memory 122, the processor 111 and the processor 121 may also transmit data via a data transmission path 1303 passing through the processor 121, the link 127, the NTB 234, the PCIe switch 125, the NTB 235, the link 118, and the processor 111. Fig.13 ) to transmit data.

[0177] The front-end interface 113 sends an interrupt to the processor 111 via the PCIe switch 115, the NTB 214, and the link 117. The front-end interface 123 sends an interrupt to the processor 111 via the PCIe switch 125, the NTB 235, and the link 118.

[0178] The storage device 100 of the third embodiment, like the storage device 100 of the first embodiment, restarts the controller after the switching process of the enqueue destination OQ by the front-end interface is completed and the host I / O being processed before the switching is completed, so that even if one controller is restarted, the host I / O is not interrupted. In addition, the storage device 100 of the third embodiment has fewer links connecting the controllers than the storage device 100 of the first embodiment, so installation becomes easier.

[0179] Furthermore, the system according to the embodiment of the present invention may be configured as follows.

[0180] (1) A storage device (e.g., storage device 100) that processes requests from a host device, the storage device comprising a plurality of storage controllers, a first storage controller (e.g., controller 110) of the plurality of storage controllers comprising: a first front-end interface (e.g., front-end interface 113) that controls a communication protocol with the host device; and a first processor (e.g., processor 111) that controls the storage device, a second storage controller (e.g., controller 120) of the plurality of storage controllers comprising a second processor (e.g., processor 121) that controls the storage device, the first storage controller further comprising: a first address conversion unit (e.g., NTB214) that converts a first address used by the first processor and a second address used by the first front-end interface; a second address conversion unit (e.g., NTB215) that converts a third address used by the second processor and the second address used by the first front-end interface; a first outbound queue (e.g., OQ201) that controls the outbound requests from the first front-end interface to the host device; The first front-end interface controls data transmission from the first processor through the first address conversion unit; and a first inbound queue (e.g., IQ202) that controls data transmission from the first processor through the first address conversion unit to the first front-end interface, the second storage controller also includes: a second outbound queue (e.g., OQ223) that controls data transmission from the first front-end interface to the second processor through the second address conversion unit; and a second inbound queue (e.g., IQ224) that controls data transmission from the second processor to the first front-end interface through the second address conversion unit, after the first front-end interface receives a first enqueue destination switching indication that specifies the second outbound queue as the enqueue destination of a request from the host device, the first front-end interface switches the enqueue destination to the second outbound queue (e.g., steps 904, 905 or steps 1104, 1105) according to a request from the host device that is assigned an identifier of a series of jobs associated with a new host request to be received next.

[0181] Thus, even if the processor of one controller constituting the dual controller stops processing, the processor of the other controller can take over the I / O request from the host device and continue responding.

[0182] (2) In the above (1), the first processor performs the following processing: before stopping processing, sending the first enqueue destination switching indication to the first front-end interface (for example, step 904 or step 1104); and stopping processing after the first outbound queue and the first inbound queue become empty (for example, steps 906, 907 or steps 1106, 1107).

[0183] Thus, when the processor of one controller is stopped in a planned manner for the purpose of updating the OS, etc., the processor of the other controller can take over and continue to respond to I / O requests from the host device.

[0184] (3) In the above (2), when the identifier of a series of jobs associated with the request from the host device that was most recently processed in the first outbound queue and the identifier of a series of jobs associated with the host request that was most recently processed in the first inbound queue in the response to the request from the host device represent the same host request, the first processor stops processing (for example, steps 906, 907, or steps 1106, 1107).

[0185] This makes it possible to reliably determine that there is no unfinished host I / O and stop the processor at an appropriate timing.

[0186] (4) In the above (1), the first front-end interface respectively sets the sending destination address of the first interrupt associated with the first outgoing queue and the sending destination address of the second interrupt associated with the second outgoing queue, and sends the first interrupt to the first processor through the first address conversion unit (for example, via the data transmission path 400), and sends the second interrupt to the second processor through the second address conversion unit (for example, via the data transmission path 500).

[0187] Thus, when the processor of one controller stops processing, the processor of another controller can take over the I / O request from the host device.

[0188] (5) In the above (1), the first processor can read the enqueue destination setting of the first front end interface.

[0189] This allows you to confirm the current enqueue destination.

[0190] (6) In the above (1), after the first processor is restarted, if the result of reading the enqueue destination setting of the request from the host device from the first front-end interface is that the enqueue destination is the second outbound queue, the first processor switches the enqueue destination to the first outbound queue (for example, "Yes" in step 909 to step 912, or "Yes" in step 1109 to step 1112).

[0191] This makes it possible to restore the I / O of the restarted processor.

[0192] (7) In the above (1), the storage device also includes: a third address conversion unit (for example, NTB235), which converts the first address used by the first processor and the fourth address used by the second front-end interface; and a fourth address conversion unit (for example, NTB234), which converts the third address used by the second processor and the fourth address used by the second front-end interface, and the first processor and the second processor communicate with each other through the third address conversion unit and the fourth address conversion unit (for example, via the data transmission path 1303).

[0193] This can reduce the number of links between controllers, making installation easier.

[0194] (8) In the above (1), the storage device further includes a switch (e.g., switch 115) including the first address conversion unit and the second address conversion unit, and the first processor and the second processor communicate with each other through the switch (e.g., via the data transmission path 1302).

[0195] This can reduce the number of links between controllers, making installation easier.

[0196] (9) In the above (1), the first front-end interface enqueues the request from the host device into the first outbound queue or the second outbound queue (for example, transmission via the data transmission path 400 or 1000) before receiving the first enqueue destination switching indication, and the first processor sends the first enqueue destination switching indication to the first front-end interface before stopping processing (for example, step 1104). When the first front-end interface receives the first enqueue destination switching indication from the first processor, it enqueues only into the second outbound queue based on the request from the host device assigned an identifier of a series of jobs associated with a new host request to be received next (for example, step 1105). The first processor stops processing after the first outbound queue and the first inbound queue become empty (for example, steps 1106, 1107).

[0197] Thus, when a front-end interface is normally enqueued with respect to a queue associated with multiple processors, it is biased toward a queue associated with another processor before stopping one processor, so that even if the processor of one controller stops processing, the processor of another controller can take over the I / O request from the host device and continue to respond.

[0198] (10) In the above (9), the second storage controller also includes: a second front-end interface (e.g., front-end interface 123), which controls the communication protocol with the host device; a third address conversion unit (e.g., NTB235), which converts the first address used by the first processor and the fourth address used by the second front-end interface; a fourth address conversion unit (e.g., NTB234), which converts the third address used by the second processor and the fourth address used by the second front-end interface; a third outbound queue (e.g., OQ221), which controls data transmission from the second front-end interface to the second processor through the fourth address conversion unit; and a third inbound queue (e.g., IQ222), which controls data transmission from the second processor to the second front-end interface through the fourth address conversion unit, and the first storage controller also includes: a fourth outbound queue (e.g., OQ203), which controls data transmission from the second front-end interface to the first processor through the third address conversion unit ; and a fourth inbound queue (e.g., IQ204) that controls data transmission from the first processor to the second front-end interface through the third address conversion unit, the second front-end interface enqueues the request from the host device into the third outbound queue or the fourth outbound queue, the second processor sends a second enqueue destination switching indication that specifies the third outbound queue as the enqueue destination of the request from the host device to the second front-end interface before the first processor stops processing (e.g., step 1104), the second front-end interface, upon receiving the second enqueue destination switching indication, enqueues only the request from the host device to the third outbound queue according to the identifier of a series of jobs associated with a new host request to be received next (e.g., step 1105), the first processor stops processing after the first outbound queue, the first inbound queue, the fourth outbound queue, and the fourth inbound queue become empty (e.g., steps 1106, 1107).

[0199] Thus, when a front-end interface is normally enqueued with respect to a queue associated with multiple processors, it is biased toward a queue associated with another processor before stopping one processor, so that even if the processor of one controller stops processing, the processor of another controller can take over the I / O request from the host device and continue to respond.

[0200] (11) In the above (10), when the identifier of a series of jobs associated with the request from the host device that was most recently processed in the first outbound queue and the identifier of a series of jobs associated with the host request that was most recently processed in the first inbound queue in the response to the request from the host device represent the same host request, and when the identifier of a series of jobs associated with the request from the host device that was most recently processed in the fourth outbound queue and the identifier of a series of jobs associated with the host request that was most recently processed in the fourth inbound queue in the response to the request from the host device represent the same host request, the first processor stops processing (for example, steps 1106, 1107).

[0201] This makes it possible to reliably determine that there is no unfinished host I / O and stop the processor at an appropriate timing.

[0202] (12) In the above (1), when a failure occurs in the first storage controller, the second processor that detects the failure sends the first enqueue destination switching instruction to the first front-end interface.

[0203] Thus, even when a processor of one controller fails, the processor of the other controller can take over and continue to respond to I / O requests from the host device.

[0204] In addition, the present invention is not limited to the above-mentioned embodiments, and includes various modified examples. For example, the above-mentioned embodiments are described in detail for a better understanding of the present invention, and are not necessarily limited to all structures having the description. In addition, a part of the structure of a certain embodiment can be replaced with the structure of other embodiments, and the structure of other embodiments can be added to the structure of a certain embodiment. In addition, with respect to a part of the structure of each embodiment, other structures can be added, deleted, or replaced.

[0205] In addition, the above-mentioned structures, functions, processing parts, processing units, etc. can also be implemented in part or in whole by hardware, for example, by designing with integrated circuits. In addition, the above-mentioned structures, functions, etc. can also be implemented by software by interpreting and executing programs that implement each function through a processor. Information such as programs, tables, files, etc. that implement each function can be stored in a storage device such as a non-volatile semiconductor memory, a hard disk drive, an SSD (Solid State Drive), or a computer-readable non-temporary data storage medium such as an IC card, an SD card, and a DVD.

[0206] In addition, the control lines and information lines are those considered necessary for the explanation, and not all control lines and information lines are necessarily shown in the product. In fact, it can be considered that almost all the components are connected to each other.

[0207] Explanation of symbols

[0208] 100: Storage device

[0209] 110, 120: Controller

[0210] 111, 121: Processor

[0211] 112, 122: Memory

[0212] 113, 123, front-end interface

[0213] 114, 124: Backend interface

[0214] 115, 125: PCIe switch

[0215] 116, 126: Management processor

[0216] 130: Driver box

[0217] 217, 218, 237, 218: Interrupt setting register

[0218] 211, 214, 215, 231, 234, 235: Non-transparent bridge

[0219] 201, 203, 221, 223: Outgoing queues

[0220] 202, 204, 222, 224: Inbound queues.

Claims

1. A storage device that processes a request from a host device, characterized in that: The storage device includes a plurality of storage controllers. A first storage controller among the plurality of storage controllers comprises: a first front-end interface for controlling a communication protocol with the host device; and a first processor for controlling the storage device. A second storage controller among the plurality of storage controllers includes a second processor that performs control of the storage device, The first storage controller further includes: A first address conversion unit, which converts a first address used by the first processor and a second address used by the first front-end interface; a second address conversion unit, which converts a third address used by the second processor and the second address used by the first front-end interface; a first outbound queue, which controls data transmission from the first front-end interface to the first processor through the first address conversion unit; and a first inbound queue, which controls data transmission from the first processor to the first front-end interface through the first address conversion unit, The second storage controller further includes: a second outgoing queue, which controls data transmission from the first front-end interface to the second processor through the second address conversion unit; and a second inbound queue, which controls data transmission from the second processor to the first front-end interface through the second address conversion unit, After receiving a first enqueue destination switching indication specifying the second outbound queue as the enqueue destination of a request from the host device, the first front-end interface switches the enqueue destination to the second outbound queue in accordance with a request from the host device assigned an identifier of a series of jobs associated with a new host request to be received next.

2. The storage device according to claim 1, characterized in that The first processor performs the following processing: Before stopping processing, sending the first enqueue destination switching indication to the first front-end interface; and The processing is stopped after the first outgoing queue and the first incoming queue become empty.

3. The storage device according to claim 2, characterized in that: The first processor stops processing when the identifier of a series of jobs associated with the request from the host device that was most recently processed in the first outbound queue and the identifier of a series of jobs associated with the host request that was most recently processed in the first inbound queue in the response to the request from the host device represent the same host request.

4. The storage device according to claim 1, characterized in that: The first front-end interface sets a sending destination address of a first interrupt associated with the first outgoing queue and a sending destination address of a second interrupt associated with the second outgoing queue, respectively. sending the first interrupt to the first processor through the first address conversion unit, The second interrupt is sent to the second processor through the second address conversion unit.

5. The storage device according to claim 1, characterized in that: The first processor is capable of reading the enqueue destination setting of the first front-end interface.

6. The storage device according to claim 5, characterized in that: After the first processor is restarted, if the result of reading the enqueue destination setting of the request from the host device from the first front end interface shows that the enqueue destination is the second outbound queue, the first processor switches the enqueue destination to the first outbound queue.

7. The storage device according to claim 1, characterized in that: The storage device also includes: a third address conversion unit that converts the first address used by the first processor and a fourth address used by the second front-end interface; and a fourth address conversion unit, which converts the third address used by the second processor and the fourth address used by the second front-end interface, The first processor and the second processor communicate with each other through the third address conversion unit and the fourth address conversion unit.

8. The storage device according to claim 1, characterized in that: The storage device further includes a switch, the switch including the first address conversion unit and the second address conversion unit. The first processor and the second processor communicate with each other through the switch.

9. The storage device according to claim 1, characterized in that: The first front-end interface enqueues the request from the host device into the first outgoing queue or the second outgoing queue before receiving the first enqueue destination switching instruction, Before stopping processing, the first processor sends the first enqueue destination switching indication to the first front-end interface, When the first enqueue destination switching instruction is received from the first processor, the first front-end interface enqueues only in the second outgoing queue according to a request from the host device assigned with an identifier of a series of jobs associated with a new host request to be received next, The first processor stops processing after the first outbound queue and the first inbound queue become empty.

10. The storage device according to claim 9, characterized in that: The second storage controller further includes: a second front-end interface that controls a communication protocol with the host device; a third address conversion unit, configured to convert the first address used by the first processor and a fourth address used by the second front-end interface; a fourth address conversion unit, which converts the third address used by the second processor and a fourth address used by the second front-end interface, a third outgoing queue, which controls data transmission from the second front-end interface to the second processor through the fourth address conversion unit; and a third inbound queue, which controls data transmission from the second processor to the second front-end interface through the fourth address conversion unit, The first storage controller further includes: a fourth outbound queue, which controls data transmission from the second front-end interface to the first processor through the third address conversion unit; and a fourth inbound queue, which controls data transmission from the first processor to the second front-end interface through the third address conversion unit, The second front-end interface enqueues the request from the host device into the third outgoing queue or the fourth outgoing queue, the second processor sending a second enqueue destination switching instruction designating the third outgoing queue as an enqueue destination of a request from the host device to the second front-end interface before the first processor stops processing, When receiving the second enqueue destination switching instruction, the second front-end interface enqueues only the third outgoing queue according to a request from the host device to which an identifier of a series of jobs associated with a new host request to be received next is assigned, The first processor stops processing after the first outbound queue, the first inbound queue, the fourth outbound queue, and the fourth inbound queue become empty.

11. The storage device according to claim 10, characterized in that: In the following cases, the first processor stops processing: an identifier of a series of jobs associated with a request from the host device that was most recently processed in the first outbound queue and an identifier of a series of jobs associated with a host request that was most recently processed in the first inbound queue in response to a request from the host device represent the same host request, Furthermore, the identifier of a series of jobs associated with the request from the host device that was most recently processed in the fourth outbound queue and the identifier of a series of jobs associated with the host request in the response to the request from the host device that was most recently processed in the fourth inbound queue represent the same host request.

12. The storage device according to claim 1, characterized in that: When a failure occurs in the first storage controller, the second processor that detects the failure sends the first enqueue destination switching instruction to the first front-end interface.

13. A method for controlling a storage controller of a storage device, wherein the storage device processes a request from a host device, characterized in that: The storage device includes a plurality of storage controllers. A first storage controller among the plurality of storage controllers comprises: a first front-end interface for controlling a communication protocol with the host device; and a first processor for controlling the storage device. A second storage controller among the plurality of storage controllers includes a second processor that performs control of the storage device, The first storage controller further includes: A first address conversion unit, which converts a first address used by the first processor and a second address used by the first front-end interface; a second address conversion unit, which converts a third address used by the second processor and the second address used by the first front-end interface; a first outbound queue, which controls data transmission from the first front-end interface to the first processor through the first address conversion unit; and a first inbound queue, which controls data transmission from the first processor to the first front-end interface through the first address conversion unit, The second storage controller further includes: a second outgoing queue, which controls data transmission from the first front-end interface to the second processor through the second address conversion unit; and a second inbound queue, which controls data transmission from the second processor to the first front-end interface through the second address conversion unit, The control method of the storage controller comprises the following steps: The first front-end interface receives a first enqueue destination switch indication designating the second egress queue as an enqueue destination of a request from the host device; and After receiving the first enqueue destination switching instruction, the first front end interface switches the enqueue destination to the second outbound queue in response to a request from the host device to which an identifier of a series of jobs associated with a new host request to be received next is assigned.