WO2022001138A1 - Ssd disk controller reset method, system and apparatus, and readable storage medium - Google Patents

Abstract

An SSD disk controller reset method, system and apparatus, and a computer-readable storage medium. The method is applied to a front-end protocol interface module, and comprises: receiving an IO sent by a rear-end service module (S11); determining whether an IO queue corresponding to a feature code carried by the IO is present (S12); and if not, discarding the IO (S13). Whether an IO queue corresponding to an IO is present is determined by using a feature code in the IO, and according to whether the IO queue is present, whether the IO is a timeout IO and whether the IO can be sent to a Host end are determined, such that the IO queue can be successfully deleted for a reset operation even before a rear-end service module releases all the IOs, and the situation in which a timeout IO is sent to the Host end is thus avoided, thereby improving the reset speed and reliability.

Description

A kind of SSD disk controller reset method, system, device and readable storage medium

This application claims the priority of the Chinese patent application filed on June 30, 2020, with the application number of 202010616527.2 and the invention titled "An SSD disk controller reset method, system, device and readable storage medium" , the entire contents of which are incorporated herein by reference.

technical field

The present invention relates to the field of computer technology, and in particular, to a method, system, device and computer-readable storage medium for resetting an SSD disk controller.

Background technique

Compared with traditional mechanical hard drives, SSDs (Solid-state drives) are more and more widely used in enterprise high-performance storage systems and personal consumer fields due to their significantly lower read and write latency and high bandwidth.

SSD has a variety of interface forms, among which SSD based on PCIe bus (PCIe, peripheral component interconnect express, high-speed serial computer expansion bus standard) NVMe protocol interface (NVMe, Non-Volatile Memory express, NVMe storage system) SSD has more Low latency, higher bandwidth, and higher random read and write IOPS (Input/Output Operations Per Second, the number of read and write operations per second) occupy an important position in the market.

The SSD disk using the NVMe protocol interface provides an interface for controller reset according to the protocol: The Host can trigger the disk to do the controller reset process by configuring the disk control register. After the disk receives the reset request sent by the Host, it will Stop the current service and perform the action of deleting the queue. After completing these processes, the Host is notified of the reset completion information through the status register.

In the actual application process, the controller reset is more common in abnormal scenarios, such as the scenario in which the IO (Input/Output, input/output) in the SSD is overtime: the Host will abort the overtime IO when it detects an IO timeout. After failure, the process of queue deletion and controller reset will be performed. Generally speaking, the SSD disk is divided into front-end protocol interface module and back-end business module. IO timeout usually means that some kind of abnormality has occurred in the back-end business module. After receiving the queue deletion request, the front-end protocol module usually needs to wait and confirm. The subsequent process of queue deletion can be completed only after the IO resources of the terminal have been released. Otherwise, if only the front-end resources are deleted and released without waiting for and confirming the release of the back-end resources, there may be abnormal back-end IO in a certain The time is returned to the front-end interface. More specifically, if the host happens to recreate the queue at this time, it is possible to return to the host through the newly created queue, which will lead to serious consequences of data inconsistency. Therefore, under this mechanism, waiting for and confirming the release of back-end IO resources when the controller resets and deletes the queue is an important link. However, in the actual scenario, since IO timeout occurs, it means that some kind of failure has occurred in the back-end module of the SSD disk. In this failure state, the action of resource release may not be completed, which will cause the front-end waiting conditions to fail to be obtained. Satisfied, causing the process to get stuck and unable to continue processing. Therefore, in practical applications, it is often encountered that the queue deletion fails due to the inability to confirm the release of the back-end IO resources, which in turn leads to the failure of the controller reset.

Therefore, a more reliable and effective reset method is required.

SUMMARY OF THE INVENTION

In view of this, the purpose of the present invention is to provide a method, system, device and computer-readable storage medium for resetting an SSD disk controller, which are more reliable and effective. Its specific plan is as follows:

A method for resetting an SSD disk controller, applied to a front-end protocol interface module, includes:

Receive the IO sent by the back-end business module;

Determine whether there is an IO queue corresponding to the feature code carried by the IO;

If it does not exist, the IO is discarded.

Optionally, before the receiving IO, the method further includes:

Receive the host's queue creation command;

According to the queue creation instruction, create the IO queue and the feature code of the IO queue;

Receive the IO request sent by the Host;

write the feature code into the IO request;

Send the IO request with the feature code to the back-end service module through the IO queue.

Optionally, the feature code is formulated according to the queue ID of the IO queue, the CPU tick count information and the current power-on times of the SSD device.

Optionally, the process of writing the feature code into the IO request includes:

Write the feature code into the context information in the IO request.

Optionally, the process of judging whether there is an IO queue corresponding to the feature code carried by the IO includes:

Obtain the feature code carried in the IO context information;

Determine whether there is an IO queue corresponding to the feature code.

Optionally, also include:

Receive the queue deletion command sent by the Host;

The corresponding IO queue is deleted directly according to the queue deletion instruction.

The invention also discloses an SSD disk controller reset system, which is applied to the front-end protocol interface module, including:

The IO receiving module is used to receive the IO sent by the back-end business module;

A feature code judgment module, for judging whether there is an IO queue corresponding to the feature code carried by the IO;

The IO discarding module is configured to discard the IO if the feature code judging module determines that the IO queue does not exist.

Optionally, also include:

The instruction receiving module is used to receive the queue creation instruction of the Host;

a queue creation module for creating the IO queue and the feature code of the IO queue according to the queue creation instruction;

The IO request receiving module is used to receive the IO request sent by the Host;

a feature code writing module for writing the feature code into the IO request;

An IO request sending module, configured to send the IO request with the feature code to the back-end business module through the IO queue.

The invention also discloses an SSD disk controller reset device, comprising:

memory for storing computer programs;

The processor is configured to execute the computer program to implement the foregoing method for resetting the SSD disk controller.

The present invention also discloses a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the foregoing method for resetting an SSD disk controller is implemented.

In the present invention, the SSD disk controller reset method is applied to the front-end protocol interface module, including: receiving the IO sent by the back-end service module; judging whether there is an IO queue corresponding to the feature code carried by the IO; if not, discarding IO.

The present invention uses the feature code in the IO to judge whether there is an IO queue corresponding to the IO, and according to whether the IO queue exists, it is judged whether the IO is a timeout IO and whether it can be sent to the Host side. Before the module releases all IOs, it can also delete the IO queue for reset operation, instead of sending IOs to the host in the primary selection timeout, which improves the reset speed and reliability.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only It is an embodiment of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to the provided drawings without creative work.

1 is a schematic flowchart of a method for resetting an SSD disk controller disclosed in an embodiment of the present invention;

2 is a schematic flowchart of another method for resetting an SSD disk controller disclosed in an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of a reset system for an SSD disk controller disclosed in an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

An embodiment of the present invention discloses a method for resetting an SSD disk controller. Referring to FIG. 1, the method is applied to a front-end protocol interface module. The method includes:

S11: Receive the IO sent by the back-end business module.

Specifically, after the front-end protocol interface module in the SSD device sends the IO request from the Host side to the back-end business module in the SSD device, the back-end business module will return the corresponding IO, for example, the data required by the IO request or according to The response information after the IO request processing is completed. Therefore, the front-end protocol interface module will receive the IO sent by the back-end business module.

The Host side may be an operating system that directly interacts with the user.

S12: Determine whether there is an IO queue corresponding to the feature code carried by the IO.

Specifically, when the SSD disk controller needs to be reset, the back-end service module may have a timeout phenomenon, so there will be IO accumulation. At this time, the back-end service module may send the time-out IO to the front-end protocol interface module. , in order to directly delete the IO queue for reset when the SSD disk controller is reset, to avoid waiting too long for the back-end business module to release IO, and the IO that takes too long to reset fails and times out passes the front-end protocol interface module from the back-end business When the module is sent to the Host, it is necessary to determine whether there is an IO queue corresponding to the feature code carried by the IO.

It is understandable that an IO request corresponds to an IO queue. Under normal conditions, the IO returned by the back-end business module according to the IO request will also use the same IO queue used when the IO request is sent to the back-end business module; In order to directly delete the IO queue during reset, without waiting for the back-end business module to release the IO, in the embodiment of the present invention, when the IO times out, the IO queue corresponding to the IO may have been deleted due to the reset operation. At this time, the back-end business The time-out IO returned by the module will not have a corresponding IO queue, and can only be sent to the front-end protocol interface module through the IO queue created to process other IO requests. Therefore, by judging the characteristics of the IO corresponding to the IO queue code, you can know whether the IO is a timeout IO and whether it should be discarded.

Among them, the feature code is the identification information corresponding to the IO queue when the IO queue is created according to the IO request. Each IO queue has a unique feature code. Therefore, the corresponding IO queue can be found through the feature code.

S13: If it does not exist, discard the IO.

Specifically, if the corresponding IO queue cannot be found by using the feature code in the current IO, it means that the IO queue corresponding to the current IO has been deleted, indicating that the current IO is a timeout IO and should not be sent to the Host side, but directly in the front-end protocol interface module. Abandoning IO avoids the situation that IO will be sent to the host over time after the IO queue is deleted, which provides a prerequisite for the rapid reset of the SSD disk controller.

Specifically, if there is an IO queue corresponding to the feature code, it means that the IO currently has an IO queue corresponding to the threshold, and no reset operation is currently performed. The IO is valid, and the IO can be sent to the Host to complete the entire IO request.

It can be seen that the embodiment of the present invention uses the feature code in the IO to determine whether there is an IO queue corresponding to the IO, and according to whether the IO queue exists, it is determined whether the IO is a timeout IO and whether it can be sent to the Host side. Before waiting for the back-end business module to release all IOs, the IO queue can also be deleted for the reset operation, instead of sending the IOs to the host in the primary selection timeout, which improves the reset speed and reliability.

The embodiment of the present invention discloses a specific SSD disk controller reset method, which is applied to the front-end protocol interface module. Compared with the previous embodiment, this embodiment further describes and optimizes the technical solution. See Figure 2, specifically:

S201: Receive a queue creation instruction from the Host.

Specifically, before sending an IO request, the Host needs to allocate a corresponding IO queue for the IO request. Therefore, it will send the corresponding queue creation command to the front-end protocol interface module to create a new IO queue for subsequent IO requests.

S202: Create an IO queue and a feature code of the IO queue according to the queue creation instruction.

Specifically, when the IO queue is created according to the queue creation instruction, a corresponding feature code is created for the IO queue at the same time. , CPU clock, CPU, central processing unit, central processing unit) and the current power-on times of the SSD device to obtain a unique feature code. Of course, you can also set a unique feature code according to different parameters or setting methods according to actual application scenarios. .

S203: Receive the IO request sent by the Host;

S204: Write the feature code into the IO request.

Specifically, after receiving the IO request sent by the Host, the feature code can be written into the IO request, and the relationship between the IO request and the IO can be confirmed through the context information. To this end, the feature code can be written into the IO request. In this way, the feature code can be obtained while using the context to determine the relationship between IO and IO request.

S205: Send the IO request with the feature code to the back-end business module through the IO queue.

It can be understood that the IO request is sent to the back-end business module through the IO queue corresponding to the feature code.

S206: Receive the IO sent by the back-end business module;

S207: Obtain the feature code carried in the IO context information.

Specifically, the context information in the IO generated by the back-end business module also has the same feature code according to the context information in the IO request. Therefore, after receiving the IO sent by the back-end business module, the IO can be obtained from the IO context information. feature code for subsequent judgment.

S208: Determine whether there is an IO queue corresponding to the feature code;

S209: If it does not exist, discard IO;

S210: Receive a queue deletion instruction sent by the Host.

Specifically, during reset, the IO queue needs to be deleted. Therefore, the queue deletion instruction may be a reset instruction or an instruction used to delete a queue alone.

S211: Delete the corresponding IO queue directly according to the queue deletion instruction.

Specifically, by judging the feature code, it can effectively prevent the timeout IO from being sent to the Host. Therefore, the front-end protocol interface module does not need to wait for the back-end business module to release the IO, and can directly delete the corresponding IO queue after receiving the queue deletion instruction. The reset of the SSD disk controller is completed, the reset efficiency is improved, and the reliability is improved without being affected by the back-end service modules.

Among them, when the back-end business module is reset, it can release the IO according to the preset reset method by itself. Even if the release fails and the IO is blocked, it will not affect the deletion of the IO queue.

It should be noted that S210 can be executed at any time. Even if there is no IO queue currently, the queue deletion instruction can be received and executed. Of course, there is no change before and after execution. Therefore, S210 is not executed in a clear order relative to S201 to S209. The order is not limited here.

Correspondingly, an embodiment of the present invention also discloses a system for resetting an SSD disk controller, as shown in FIG. 3 , which is applied to a front-end protocol interface module, including:

The IO receiving module 11 is used to receive the IO sent by the back-end business module;

The feature code judgment module 12 is used to judge whether there is an IO queue corresponding to the feature code carried by the IO;

The IO discarding module 13 is configured to discard the IO if the feature code determination module 12 determines that the IO queue does not exist.

It can be seen that the embodiment of the present invention uses the feature code in the IO to determine whether there is an IO queue corresponding to the IO, and according to whether the IO queue exists, it is determined whether the IO is a timeout IO and whether it can be sent to the Host side. Before waiting for the back-end business module to release all IOs, the IO queue can also be deleted for the reset operation, instead of sending the IOs to the host in the primary selection timeout, which improves the reset speed and reliability.

Specifically, it may also include an instruction receiving module, a queue creation module, an IO request receiving module, a feature code writing module, and an IO request sending module; wherein,

The instruction receiving module is used to receive the queue creation instruction of the Host;

The queue creation module is used to create the IO queue and the feature code of the IO queue according to the queue creation instruction;

The IO request receiving module is used to receive the IO request sent by the Host;

The feature code writing module is used to write the feature code into the IO request;

The IO request sending module is used to send the IO request with the characteristic code to the back-end business module through the IO queue.

Among them, the feature code can be formulated according to the queue ID of the IO queue, CPU tick count information, and the current number of power-on times of the SSD device.

Specifically, the above feature code writing module is specifically configured to write the feature code into the context information in the IO request.

Specifically, the above-mentioned feature code judgment module 12 may include a feature code acquisition unit and a feature code judgment unit; wherein,

A feature code obtaining unit, used to obtain the feature code carried in the IO context information;

The feature code judgment unit is used to judge whether there is an IO queue corresponding to the feature code.

Specifically, it may also include a deletion instruction receiving module and a queue deletion module; wherein,

The deletion instruction receiving module is used to receive the queue deletion instruction sent by the Host;

The queue deletion module is used to delete the corresponding IO queue directly according to the queue deletion instruction.

The embodiment of the present invention also discloses a device for resetting an SSD disk controller, comprising:

memory for storing computer programs;

The processor is used for executing the computer program to implement the foregoing method for resetting the SSD disk controller.

An embodiment of the present invention further discloses a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the foregoing method for resetting an SSD disk controller is implemented.

Finally, it should also be noted that in this document, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply these entities or that there is any such actual relationship or sequence between operations. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device that includes a list of elements includes not only those elements, but also includes not explicitly listed or other elements inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

Professionals may further realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two, in order to clearly illustrate the possibilities of hardware and software. Interchangeability, the above description has generally described the components and steps of each example in terms of functionality. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each particular application, but such implementations should not be considered beyond the scope of the present invention.

The technical content provided by the present invention is described in detail above, and specific examples are used in this paper to illustrate the principles and implementations of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific embodiments and application scope. To sum up, the content of this specification should not be construed as a limitation of the present invention.

Claims (10)

1. A method for resetting an SSD disk controller, characterized in that it is applied to a front-end protocol interface module, comprising:

   Receive the IO sent by the back-end business module;

   Determine whether there is an IO queue corresponding to the feature code carried by the IO;

   If it does not exist, the IO is discarded.
2. The method for resetting an SSD disk controller according to claim 1, wherein before the receiving IO, the method further comprises:

   Receive the host's queue creation command;

   According to the queue creation instruction, create the IO queue and the feature code of the IO queue;

   Receive the IO request sent by the Host;

   write the feature code into the IO request;

   Send the IO request with the feature code to the back-end service module through the IO queue.
3. SSD disk controller reset method according to claim 1, is characterized in that, described feature code is formulated according to the queue ID of described IO queue, CPU tick count information and current power-on times of SSD equipment.
4. The method for resetting an SSD disk controller according to claim 2, wherein the process of writing the feature code into the IO request comprises:

   Write the feature code into the context information in the IO request.
5. The method for resetting an SSD disk controller according to claim 4, wherein the process of judging whether there is an IO queue corresponding to the feature code carried by the IO comprises:

   Obtain the feature code carried in the IO context information;

   Determine whether there is an IO queue corresponding to the feature code.
6. The method for resetting an SSD disk controller according to any one of claims 1 to 5, further comprising:

   Receive the queue deletion command sent by the Host;

   The corresponding IO queue is deleted directly according to the queue deletion instruction.
7. An SSD disk controller reset system, characterized in that, applied to a front-end protocol interface module, comprising:

   The IO receiving module is used to receive the IO sent by the back-end business module;

   A feature code judgment module, for judging whether there is an IO queue corresponding to the feature code carried by the IO;

   The IO discarding module is configured to discard the IO if the feature code judging module determines that the IO queue does not exist.
8. The SSD disk controller reset system according to claim 7, further comprising:

   The instruction receiving module is used to receive the queue creation instruction of the Host;

   a queue creation module for creating the IO queue and the feature code of the IO queue according to the queue creation instruction;

   The IO request receiving module is used to receive the IO request sent by the Host;

   a feature code writing module for writing the feature code into the IO request;

   An IO request sending module, configured to send the IO request with the feature code to the back-end business module through the IO queue.
9. A device for resetting an SSD disk controller, characterized in that it includes:

   memory for storing computer programs;

   The processor is configured to execute the computer program to implement the method for resetting an SSD disk controller according to any one of claims 1 to 6.
10. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the SSD disk according to any one of claims 1 to 6 is implemented Controller reset method.

Images