CN116450349A

CN116450349A - Method, device, system, equipment and medium for improving availability of system

Info

Publication number: CN116450349A
Application number: CN202310378825.6A
Authority: CN
Inventors: 王志浩
Original assignee: Suzhou Inspur Intelligent Technology Co Ltd
Current assignee: Suzhou Inspur Intelligent Technology Co Ltd
Priority date: 2023-04-11
Filing date: 2023-04-11
Publication date: 2023-07-18

Abstract

The invention provides a method, a device, a system, equipment and a medium for improving the availability of a system, which are applied to a storage system, wherein the storage system comprises a plurality of main control modules, relates to the technical field of storage, and aims to solve the problem of unbalanced load among the plurality of main control modules in an asymmetric service scene; the method comprises the following steps: obtaining current performance parameters of a current main control module for the current main control modules in the plurality of main control modules; the current main control module is any main control module in a plurality of main control modules, and the current performance parameter is used for representing the current service processing performance of the current main control module; determining the current load state of the current main control module based on the current performance parameter; executing a load balancing strategy corresponding to the load state so as to adjust the load state of the current main control module to the load balancing state; the load balancing strategy at least comprises the following steps: and forwarding at least one service which is not processed currently in the current main control module to a plurality of target main control modules.

Description

Method, device, system, equipment and medium for improving availability of system

Technical Field

The present invention relates to the field of storage systems, and in particular, to a method, apparatus, system, device, and medium for improving usability of a system.

Background

When the traffic volume of the user is huge and the requirements on the reliability of the traffic are high, the data are accessed through the array storage. In the related art, the load balancing condition among a plurality of main control modules is strictly required by the array storage, particularly by the high-end multi-control array storage, and in general, the load balancing effect is achieved by adopting a host plug-in and other modes to balance IO (input output) distribution, but under the extremely asymmetric service scene, if only part of main control modules can accept front-end service, the service loads among the plurality of main control modules can generate larger difference, so that the load imbalance among the plurality of main control modules is caused, and the usability of the whole system is influenced.

Disclosure of Invention

In view of this, the present invention aims to propose a method, apparatus, system, device and medium for improving the availability of a system, so as to solve the problem that the load balancing method in the related art has limited effect and is difficult to implement load balancing between multiple master control modules in an asymmetric service scenario.

In a first aspect of an embodiment of the present invention, a method for improving usability of a system is provided, where the method is applied to a storage system, and the storage system includes a plurality of main control modules, and the method includes:

Obtaining current performance parameters of a current main control module in a plurality of main control modules; the current main control module is any main control module of a plurality of main control modules, and the current performance parameter is used for representing the current service processing performance of the current main control module;

determining the current load state of the current main control module based on the current performance parameter;

executing a load balancing strategy corresponding to the load state so as to adjust the load state of the current main control module to the load balancing state;

wherein, the load balancing strategy at least comprises: forwarding at least one service which is not processed currently in the current main control module to a plurality of target main control modules; the target main control module is any main control module except the current main control module in the plurality of main control modules.

Further, the determining, based on the current performance parameter, a load state in which the current main control module is currently located includes:

determining that the load state is a first overload state under the condition that the current performance parameter exceeds a first preset threshold value and does not exceed a second preset threshold value;

And determining that the load state is a second overload state under the condition that the current performance parameter exceeds the second preset threshold value.

Further, the executing the load balancing policy corresponding to the load state includes:

executing a first load balancing strategy corresponding to the first overload state under the first overload state; the first load balancing strategy comprises the following steps: switching the management mode of the current main control module to enable the current main control module to enter a service to-be-forwarded state;

executing a second load balancing strategy corresponding to the second overload state under the second overload state; the second load balancing strategy comprises the following steps: and forwarding at least one service which is not processed currently in the current main control module to a plurality of target main control modules.

Further, the forwarding the at least one currently unprocessed service in the master control module to the plurality of target master control modules includes:

determining the target number of the currently unprocessed business to be forwarded based on the current performance parameter and the preset performance parameter of the main control module; the preset performance parameters are performance parameters of the main control module in the load balancing state;

Forwarding the target number of the services which are not processed currently in the main control module to a plurality of target main control modules.

acquiring current performance parameters of each main control module except the current main control module;

determining at least one target master control module from a plurality of master control modules except the current master control module based on the current performance parameter of each master control module;

and forwarding at least one service which is not processed currently in the current main control module to the target main control module.

and sending a service forwarding instruction to the current main control module so that the current main control module forwards at least one currently unprocessed service to a plurality of target main control modules based on the service forwarding instruction.

Further, the forwarding step includes:

determining the number of the businesses to be received of each target main control module based on the target number;

Forwarding a corresponding number of unprocessed services to each target master control module;

or forwarding the current unprocessed business to each target main control module one by one until the target number is reached.

Further, after forwarding the at least one service that is not currently processed in the current master control module to a plurality of the target master control modules, the method further includes:

re-acquiring the current performance parameters of the current main control module at intervals of preset time length; the preset duration is the duration of stable performance of the current main control module after the forwarding service operation is executed;

based on the current performance parameters obtained again, determining the load state of the current main control module again;

and repeatedly executing the second load balancing strategy under the condition that the load state is still the second overload state until the current main control module is in the load balancing state.

Further, the business processed by the main control module has a plurality of business models; under the condition of processing the services of different service models, the preset performance parameters of the main control module are different; the method for acquiring the preset performance parameters comprises the following steps:

Determining the type of a service model of the service received by the current main control module;

and acquiring the equalization parameters corresponding to the service model types from a plurality of equalization performance parameters as the preset performance parameters.

In a second aspect of an embodiment of the present invention, there is provided an apparatus for improving usability of a system, the apparatus including:

the acquisition module is used for acquiring current performance parameters of a current main control module from a plurality of main control modules; the current main control module is any main control module of a plurality of main control modules, and the current performance parameter is used for representing the current service processing performance of the current main control module;

the determining module is used for determining the current load state of the current main control module based on the current performance parameter;

the execution module is used for executing a load balancing strategy corresponding to the load state so as to adjust the load state of the current main control module to the load balancing state;

In a third aspect of embodiments of the present invention, there is provided a storage system, the system comprising:

the system comprises a plurality of main control modules, a balanced load management module and a serial port module;

the main control modules are in communication connection with each other, and are used for processing or forwarding the service;

the balanced load management module is correspondingly connected with the plurality of main control modules, and is used for executing the method for improving the availability of the system in the first aspect;

the serial port module is connected with the balanced load management module and is used for communicating with external equipment.

In a fourth aspect of the embodiments of the present invention, there is provided an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method for improving the usability of the system according to the first aspect.

A fifth aspect of an embodiment of the present invention provides a computer-readable storage medium storing a computer program for executing the method of improving usability of a system as described in the first aspect above.

Compared with the prior art, the method for improving the usability of the system has the following advantages:

The method for improving the availability of the system is applied to a storage system, wherein the storage system comprises a plurality of main control modules, and the method comprises the following steps: obtaining current performance parameters of a current main control module in a plurality of main control modules; the current main control module is any main control module of a plurality of main control modules, and the current performance parameter is used for representing the current service processing performance of the current main control module; determining the current load state of the current main control module based on the current performance parameter; executing a load balancing strategy corresponding to the load state so as to enable the current main control module to be kept in the load balancing state; wherein, the load balancing strategy at least comprises: forwarding at least one currently unprocessed service in the master control module to a plurality of target master control modules; the target main control module is any main control module except the current main control module in the plurality of main control modules.

Therefore, the invention monitors the performance of the main control module in real time, and executes corresponding balanced load strategy to the main control module according to the monitored load state of the main control module, thereby forwarding the service of the main control module to other main control modules for processing when the main control module has overload risk, so that the load among a plurality of main control modules is balanced, and meanwhile, as the service can be mutually forwarded among the plurality of main control modules, the problem of unbalanced load among a plurality of main control modules caused by unbalanced distribution of read-write service among the plurality of main control modules in an extremely asymmetric scene is avoided; meanwhile, as the business can be forwarded among the plurality of main control modules, under the condition that part of the main control modules are idle, part of the business can be forwarded to the idle main control modules for processing, and the utilization rate of the whole machine is improved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention. In the drawings:

FIG. 1 is a schematic diagram of a storage system architecture of an asymmetric business scenario in the related art;

FIG. 2 is a flowchart illustrating steps of a method for improving usability of a system according to a first embodiment of the present invention;

FIG. 3 is a flowchart illustrating a method for improving usability of a system according to a second embodiment of the present invention;

FIG. 4 is a flowchart illustrating steps of a method for improving usability of a system according to a third embodiment of the present invention;

fig. 5 is a schematic diagram of a service forwarding process of a master control module in a third embodiment of the present invention;

FIG. 6 is a schematic structural diagram of a device for improving usability of a system according to a fourth embodiment of the present invention;

fig. 7 is a schematic structural diagram of a memory system according to a fifth embodiment of the present invention.

Detailed Description

It should be noted that, without conflict, the embodiments of the present invention and features of the embodiments may be combined with each other.

In the related art, in an extremely asymmetric service scenario, as shown in fig. 1, the present control main control module can carry front-end service, and the opposite control main control module is disconnected from the front-end module, so that the front-end service can be only processed by the present control main control module, and the opposite control main control module is idle, so that the performances of each main control module can not be fully utilized, and under the condition of large traffic, load imbalance among main control modules is easily caused, and the usability of the whole system is affected.

In view of this, the present invention provides a method, apparatus, system, device and storage medium for improving the availability of a system, by implementing service forwarding of a corresponding number between main control modules, even if the front end of the main control module cannot bear the service, load balancing between the main control modules can be implemented, and the availability of the whole machine is improved.

A method, apparatus, system, device and medium for improving system availability of the present invention will be described in detail with reference to the accompanying drawings in combination with embodiments.

Example 1

Referring to fig. 2, fig. 2 is a flowchart illustrating steps of a method for improving usability of a system according to an embodiment of the present invention, where the method according to the embodiment of the present invention is applied to a storage system, and the storage system includes a plurality of master control modules, as shown in fig. 2, and the method includes:

s101, for a current main control module in a plurality of main control modules, acquiring current performance parameters of the current main control module.

The current main control module is any one of a plurality of main control modules, and the current performance parameter is used for representing the current service processing performance of the current main control module. The current performance parameter may be a current bandwidth, a current time delay, a current read/write efficiency, i.e. IOPS (Input/Output Operations Per Second, read/write times per second), a current CPU utilization, etc. of the current main control module, which may represent performance of the main control module, and the present invention is not limited in detail. Because the performance parameters of the main control module change along with the service carried by the main control module, the performance of the main control module can be monitored in real time to obtain the performance parameters of the main control module.

S102, determining the current load state of the current main control module based on the current performance parameters.

Specifically, the current performance parameter reflects the current performance of the main control module, and whether the current main control module has overload risk can be determined based on the current performance parameter; in general, when the main control module deviates from the load balancing performance more or approaches the theoretical maximum load performance, it indicates that the main control module may be overloaded or has an overload risk, so the current performance parameter may be compared with the load balancing performance or the theoretical maximum performance to determine the current load state of the main control module, so as to determine whether to execute the load balancing policy on the current main control module.

For example, in the case of judging whether there is an overload risk based on the theoretical maximum performance of the main control module, since the main control module reaches the theoretical maximum performance already in an overload state, a value smaller than the theoretical maximum performance is taken as a judgment basis; if the performance parameter of the main control module is greater than 80% of the theoretical maximum performance parameter, determining that the overload risk exists in the current main control module, and determining the load state as a first load state which indicates that the overload risk exists in the main control module; when the performance parameter of the main control module is greater than 95% of the theoretical maximum performance, the main control module is in an overload state at this time, so that the state of the main control module is determined to be a second load state, namely the overload state, and the high load risk is indicated.

Therefore, different load balancing strategies are formulated for different load states, so that loads among a plurality of main control modules are balanced; specifically, the load balancing policy may be forwarding the unprocessed service in the master control module with overload risk to other master control modules; or determining the master control module with lower performance, and forwarding the master control module with overload risk to the master control module with lower performance.

In some embodiments, for each master module, it may also be possible to determine its performance when theoretically balancing the load, and determine whether the master module is at risk of overload based on how far the master module deviates from the theoretical load performance.

And S103, executing a load balancing strategy corresponding to the load state so as to enable the load state of the current main control module to be the load balancing state.

Wherein, the load balancing strategy at least comprises: forwarding at least one currently unprocessed service in the master control module to a plurality of target master control modules; the target main control module is any main control module except the current main control module in the plurality of main control modules.

Specifically, since the load balancing policy is preset, after the corresponding load balancing policy is determined according to the load state, the corresponding load balancing policy is directly executed. The main load balancing strategy is to forward part of unprocessed business in the main control module to other main control modules so as to forward part of business load of the current main control module to other main control modules.

In some embodiments, because the current performance of each master control module is different, in the process of forwarding the service to the other master control modules, the number of services forwarded to each master control module can be determined according to the performance of each master control module, and then the service is forwarded, so that the master control module with higher service load per se is prevented from generating overload risk.

In some embodiments, the master control module capable of receiving the service may be determined based on the current performance of each master control module, and only the unprocessed service is forwarded to the master control module with lower performance parameters, so that the idle master control modules may be utilized, and the utilization rate of the system may be improved.

According to the method for improving the availability of the system, the real-time performance of each main control module is monitored, when the overload risk of the current main control module is monitored, the corresponding balancing load strategy is executed based on the determined load state of the main control module, so that the main control module with the overload risk is adjusted to the load balancing state, and therefore when the overload risk of the main control module exists, the unprocessed part of business in the current main control module is forwarded to a plurality of main control modules, other main control modules except the current main control module balance the performance of the current main control module and the performance of other main control modules, and the overload of the current main control module is avoided, so that the load imbalance among the main control modules is avoided, and the system efficiency is affected; meanwhile, because the service among the plurality of main control modules can be mutually forwarded, the idle main control modules can be utilized even in an asymmetric service scene, and the utilization rate of the whole machine is improved.

Example two

Referring to fig. 3, fig. 3 is a flowchart illustrating steps of a method for improving usability of a system according to an embodiment of the present invention, where the method according to the embodiment of the present invention is applied to a storage system, and the storage system includes a plurality of master control modules, as shown in fig. 3, and the method includes:

s201, for a current master control module in a plurality of master control modules, acquiring current performance parameters of the current master control module.

Specifically, a CIF (Controller Intelligent Forwarding, intelligent forwarding) management module is set for each main control module, and the CIF management module is used for monitoring the performance of the main control module.

S202, determining that the load state is a first overload state under the condition that the current performance parameter exceeds a first preset threshold value and does not exceed a second preset threshold value.

Specifically, a plurality of thresholds can be set in the CIF management module in advance, and the current load state of the current main control module is determined according to different thresholds. The multiple thresholds may be determined according to a theoretical maximum load performance of the main control module, or may be determined according to an equilibrium load performance of the main control module.

In the embodiment of the invention, the first preset threshold value and the second preset threshold value are set according to the balanced load performance, so that overload risk exists when the main control module has a certain value of the ultra-high balanced load performance, and the main control module is regulated according to the overload degree. According to the balanced load performance and the parameters preset by the system pre-stored in the CIF management module, a first preset threshold and a second preset threshold are determined, wherein the first preset threshold is D1 times of the balanced load performance, the second preset threshold is D2 times of the balanced load performance, the values of D1 and D2 are determined according to the actual performance of the main control module, and the method is not particularly limited.

After the first preset threshold value and the second preset threshold value are determined, comparing the current performance parameter of the main control module with the first preset threshold value and the second preset threshold value, and determining the load state of the main control module according to the comparison result. Under the condition that the current performance parameter is larger than a first threshold value and not larger than a second preset threshold value, the main control module is in a first load state; and under the condition that the current performance parameter is larger than a second threshold value, the main control module is in a second load state. And further, different load balancing strategies are executed according to different loads.

In some embodiments, in order to avoid the situation that the CIF management module has a detection error, determining whether the main control module reaches the first overload state through two preset parameters, specifically, determining the magnitude of a preset multiple of the current performance parameter and the maximum theoretical performance again when judging that the current performance parameter of the main control module is greater than a first preset threshold, for example, determining whether the current performance parameter is greater than E1 times of the theoretical maximum performance of the main control module; if the current performance parameter is not more than E1 times of the theoretical maximum performance of the main control module, the main control module does not have overload risk, and if the current performance parameter is not more than E1 times of the theoretical maximum performance of the main control module, the main control module does not actually have overload risk; under the condition, the problems of detection failure or main control module failure and the like may exist, and when the failure condition occurs, the CIF management module reports the failure information so as to facilitate the user to conduct failure troubleshooting.

Wherein, because the performances of the plurality of main control modules are the same, the theoretical maximum performance of a single main control module can be directly determined according to the maximum theoretical performance of the cluster; if the theoretical maximum performance of the cluster of the N main control modules is B, the theoretical maximum performance of the single main control module is B/N.

In some embodiments, the services carried by the main control module have multiple service models, such as sequential reading/writing of large/small data blocks, random reading/writing of large/small data blocks, and the like, where service processing performance of each service model is different, so when determining the first preset threshold value and the second preset threshold value, a step of acquiring preset performance parameters needs to be executed first, and the step includes:

s2021, determining the service model type of the service received by the main control module.

Specifically, since the CIF management module is further connected to the front end module, the CIF management module may determine a specific type of the service model sent to the current main control module according to the service data received by the front end module.

S2022, obtaining preset performance parameters corresponding to the service model types from a plurality of balance performance parameters.

The CIF management module is internally provided with equalization performance parameters of the current main control module corresponding to different types of service models, so that after the specific types of the service models are obtained, the corresponding equalization performance parameters are determined according to the types. The CIF management module obtains a service model of a service received by the front end module, where different service models correspond to different identifiers, and after the service module is obtained, obtains preset performance parameters corresponding to the service model types from a plurality of equalization performance parameters according to preset performance parameters corresponding to the axillary service model identifiers.

S203, executing a first load balancing strategy corresponding to the first overload state under the first overload state; the first load balancing strategy comprises the following steps: and switching the management mode of the main control module so as to enable the main control module to enter a service waiting forwarding state.

In the embodiment of the invention, the management mode of the main control module comprises the following steps: a common mode and a strong equalization mode, wherein in the common mode, service forwarding between main control modules is generally not performed; in the strong equalization mode, executing service forwarding between the main control modules; therefore, when the main control module is in the first overload state, in order to prevent overload risk, the management mode of the main control module is switched to the strong equalization mode, so that the service of the current main control module can be forwarded conveniently in the strong equalization mode.

S204, determining that the load state is a second overload state under the condition that the current performance parameter exceeds the second preset threshold value.

Specifically, when the current performance parameter exceeds D2 times of the balanced load performance, it is determined that the current main control module is in the second overload state, and a higher overload risk exists, and a load balancing strategy corresponding to the second overload state is executed.

S205, executing a second load balancing strategy corresponding to the second overload state under the second overload state; the second load balancing strategy comprises the following steps: and forwarding at least one service which is not processed currently in the main control module to a plurality of target main control modules.

The target main control module is any main control module except the current main control module in the plurality of main control modules.

Specifically, in the second overload state, in order to adjust the main control module to a load balancing state, part of the unprocessed service in the main control module is forwarded to the rest main control modules, so that the overload risk of the current main control module is reduced.

In the embodiment of the invention, a CIF management module sends a service forwarding command to a main control module, and the main control module sends the service to other main control modules through corresponding interfaces according to the forwarding command; the physical path adopted by the forwarding is from a high-speed data link such as NTB (Non-Transparent Bridge ), roCE (RDMA over Converged Ethernet, remote memory direct access technology of the converged Ethernet) among multiple controllers.

In some embodiments, the number of services forwarded to the other master control modules may be determined according to the performance parameters of load balancing of the current master control module, so that the current master control module reaches a load balancing state, where the forwarding step includes:

S2051, determining the target number of the current unprocessed business to be forwarded based on the current performance parameter and the preset performance parameter of the main control module.

The preset performance parameters are performance parameters of the main control module in the load balancing state.

S2052, forwarding the currently unprocessed target number of services in the main control module to a plurality of target main control modules.

Illustratively, the number of unprocessed traffic M to be forwarded is calculated using a PID (Proportional, integral, differential, proportional, integral, derivative) algorithm: m=k1×|a-c|; wherein M is the number of unprocessed services, A is the current performance parameter of the current main control module, C is the preset performance parameter, and K1 is the PID forwarding factor. After the value of M is obtained through calculation, the CIF management module sends the value of M to the main control module, so that the main control module forwards M unprocessed services to other main control modules except the current main control module.

It should be noted that, in the storage system, the service received by the main control module is a data block, and after the service data is calculated through the performance, the number of the services to be forwarded is determined according to the size of the service data block.

The forwarding of the target number of services can be batch forwarding or one-by-one forwarding; under the condition of batch forwarding, determining the number of the businesses to be received by each target main control module based on the target number; forwarding a corresponding number of unprocessed services to each target master control module. Under the condition of forwarding one by one, forwarding the current unprocessed business to each target main control module one by one until the target quantity is reached; and forwarding a service to a target main control module each time according to a preset sequence until the target number is reached.

In some embodiments, in order to avoid forwarding the service of the current main control module to other main control modules with overload risk or to other main control modules with higher performance parameters, before forwarding the service, the CIF management module may further obtain the current performance parameters of each main control module except the current main control module; and determining at least one target main control module from a plurality of main control modules except the current main control module based on the current performance parameter of each main control module. Therefore, the performance parameters of the current main control module can be forwarded to other main control modules with lower performance parameters, and the situation that service is forwarded for many times to reduce the service performance of the main control module is avoided.

The method for improving the availability of the system provided by the embodiment of the invention judges the risk degree of the load state of the current main control module by setting a plurality of thresholds, adopts a first load balancing strategy in the first load state, adjusts the management mode of the main control module in advance, and facilitates the subsequent service forwarding; forwarding the service to other main control modules except the current main control module in a second load state; therefore, the current main control module is in a load balance state, and the occurrence of load unbalance among a plurality of main control modules is avoided; meanwhile, under an asymmetric service scene, service forwarding can be performed among the main control modules to balance loads among the plurality of main control modules, so that the idle main control modules are utilized, and the utilization rate of the whole system is improved. In addition, the PID algorithm is adopted to precisely control the service forwarding quantity, so that the performance parameters of the current main control module can be precisely controlled below the overload performance, and the overload risk of the current main control module is effectively avoided.

Example III

Referring to fig. 4, fig. 4 shows a flowchart of steps of a method for improving usability of a system according to an embodiment of the present invention, where the method according to the embodiment of the present invention is applied to a storage system, and the storage system includes a plurality of master control modules, as shown in fig. 4, and the method includes:

s301, for a current master control module in a plurality of master control modules, acquiring current performance parameters of the current master control module.

S302, based on the current performance parameters, determining the current load state of the current main control module.

S303, executing a load balancing strategy corresponding to the load state so as to adjust the load state of the current main control module to the load balancing state; wherein, the load balancing strategy at least comprises: and forwarding at least one service which is not processed currently in the current main control module to a plurality of target main control modules.

In the embodiment of the invention, the common mode and the strong balancing mode are set to manage the current main control module, so that different thresholds can be set for different management modes to execute different balancing adjustment strategies corresponding to different modes.

Specifically, the threshold value can be set according to the function of the management mode, for example, the ordinary mode is only used for preventing the current main control module from being in an overload state, so that in the ordinary mode, the performance parameter of the current main control module, which is overloaded, is taken as the threshold value, or the performance parameter with overload risk is taken as the threshold value; the strong balancing mode is to keep the current main control module in the balancing mode, so that the performance parameter of the balancing load performance close to the current main control module is used as a threshold value in the strong balancing mode.

For example, when the management mode of the current main control module is determined to be the normal mode, a first preset threshold is adopted to determine whether the current main control module is in an overload risk state, i.e. a first overload state.

Executing a first load balancing strategy under the condition that the current main control module is in a first overload state: and switching the management mode of the current main control module, and restricting the load state of the current main control module in the balanced load state by adopting a strong balancing mode to avoid overload risks.

Specifically, a request for switching the management mode is sent to the user, and when confirmation information of the user is received, the management mode of the current main control module is switched from a common mode to a strong equalization mode.

And when the management mode of the current main control module is determined to be the strong balance mode, judging whether the current main control module is in a non-balance load state or not, namely, a second overload state by adopting a second preset threshold value.

And executing a second load balancing strategy under the condition that the current main control module is in a second overload state: and forwarding at least one service which is not processed currently in the current main control module to a plurality of target main control modules.

The purpose of the strong balancing mode is to restrict the load state of the current main control module to a balanced load state, so that the second preset threshold is close to or equal to the balanced load performance parameter; in the strong balancing mode, determining whether the performance of the main control module needs to be restrained according to whether the main control module is larger than D2 times of the balanced load performance of the main control module or not; the parameter setting of D2 is usually between 1 and 1.1, so as to avoid the performance of the master control module exceeding 10% of the balanced load performance, so that the load among the plurality of master control modules is balanced.

In some embodiments, to facilitate a user to view the state of the storage system, a load imbalance warning message is generated based on the load state of the current master control module to notify the user that a load imbalance condition may exist in the system; wherein, different load states correspond to different load unbalance warning information.

For example, when the current main control module is determined to be in a first overload state, a primary load unbalance warning is reported; and when the current main control module is determined to be in the second overload state, reporting a secondary load early warning.

In some embodiments, if a high load condition occurs, i.e., approaching the theoretical maximum performance of the current master control module, a high load warning is reported; for example, when the current performance parameter of the current main control module is greater than 95% of the theoretical performance, determining that the current main control module is in a high load state, reporting a high load warning, and immediately forwarding unprocessed traffic in the current main control module.

S304, the current performance parameters of the current main control module are acquired again at intervals of preset time length.

The preset duration is a duration with stable performance of the main control module after the forwarding service operation is executed.

Specifically, after forwarding the service of the current main control module to the rest of main control modules except the current main control module, the performance of the current main control module is changed, so that the current performance parameters of the current main control module need to be acquired again after the performance of the current main control module is changed; wherein, because of the situation that other main control modules also transmit the service to the current main control module, after the service is transmitted, the current performance parameters are acquired again at intervals of preset time length; the preset duration can be a performance stabilization duration preset in advance, or can be a performance parameter of the current main control module obtained in real time, and when the performance parameter is stabilized within a certain range, the performance of the current main control module is considered to be stable, and then the performance parameter at the moment is taken as the current performance parameter.

S305, based on the fact that the current performance parameters are acquired again, the load state of the current main control module is determined again.

And S306, repeatedly executing the second load balancing strategy under the condition that the load state is still the second overload state until the current main control module is in the load balancing state.

After the current performance parameter is re-acquired, comparing the current performance parameter with a second preset threshold value to determine whether the current main control module is still in the second overload state. If the current main control module is still in the second overload state, the current main control module is not in the balanced load state, and the performance of the current main control module needs to be continuously adjusted, so that part of the unprocessed business in the current main control module is forwarded to other main control modules except the current main control module again until the current main control module is in the load balance state, namely, the current performance parameter of the current main control module is close to or equal to the performance parameter of the current main control module when the load balance is carried out.

The method for improving the availability of the system provided by the embodiment of the invention judges the load state by adopting different thresholds for the main control modules with different management modes, adopts a first load balancing strategy in the first load state, and monitors the load of the current main control module by adopting a strong balancing mode when the overload risk exists in the current main control module, so that the current main control module can be regulated to the load balancing state; forwarding the service to other main control modules except the current main control module in a second load state so as to balance the load of the current main control module; therefore, the current main control module is kept in a load balance state, and the occurrence of load unbalance among a plurality of main control modules is avoided; meanwhile, under an asymmetric service scene, service forwarding can be performed among the main control modules to balance loads among the plurality of main control modules, so that the idle main control modules are utilized, and the utilization rate of the whole system is improved.

The above process is specifically described below in connection with a specific example:

for the current master control module in the plurality of master control modules, a CIF management module is adopted to obtain performance parameters, wherein the performance parameters are parameters representing the current performance of the current master control module, such as the current bandwidth, the current time delay, the current IOPS, the current CPU utilization rate and the like of the current master control module.

After the current performance of the current main control module is obtained, determining the current load state of the current main control module according to the relation between the current performance and a preset threshold value. Specifically, a first preset threshold value c×d1 and a second preset threshold value c×d2 are set; wherein, C is the equalization performance parameter of the current main control module, D1 and D2 are the preset parameters of the system; under the condition that the management mode of the current main control module is a common mode, a first preset threshold value is adopted to determine the current load state of the current main control module; and under the condition that the management mode of the current main control module is a strong balance mode, determining the current load state of the current main control module by adopting a second preset threshold value.

And in the normal mode, judging whether the current performance parameter of the current main control module is larger than C multiplied by D1, when the current performance parameter of the current main control module is larger than C multiplied by D1, determining whether the current performance parameter is larger than (B multiplied by E1)/N, if so, indicating that overload risk exists, namely a first overload state, reporting a first-stage load warning, switching the management mode of the current main control module to a strong balance mode, and carrying out load balance through a plurality of main control modules in the strong balance mode. If the current performance parameter is not greater than (B×E1)/N, the current main control module does not have overload risk actually, and the performance detection of the current main control module may have faults and other conditions.

Judging whether the current performance parameter of the current main control module is larger than C multiplied by D2 in a strong equalization mode; when the current performance parameter of the current main control module is determined to be larger than C multiplied by D2, the current main control module deviates from the balanced load performance, namely, the current main control module is in a second overload state, a second-level load early warning is reported, the number M of the to-be-forwarded services is calculated according to K1 x|A-C| in a time period T1, wherein K1 is a PID forwarding factor, and the CIF management module sends M to the main control module so that the main control module forwards the unprocessed services according to the M value; therefore, the load state of the main control module is maintained in a load balance state by forwarding the service of the current main control module. When the number of the main control modules is four, the M services of the current main control module are equally distributed to the other three main control modules, or the weight of each main control module is determined according to the performance parameters of the other three main control modules, and the service number forwarded to each main control module is calculated; and then sending the corresponding number of services to the corresponding main control module. As illustrated in fig. 5, the current storage system includes four main control modules: the device comprises a main control module 1, a main control module 2, a main control module 3 and a main control module 4;

When the CIF management module detects that the main control module 1 is in the second overload state, a load strategy for forwarding the service to the rest main control modules is executed, so that the unprocessed service in the main control module 1 is forwarded to the main control module 2, the main control module 3 and the main control module 4 respectively through the high-speed link.

After M unprocessed services are forwarded, detecting current performance parameters of the current main control module again at intervals of a preset time length T2, wherein the preset time length T2 is determined according to the time length of stable performance of the current main control module after the services are forwarded;

comparing the re-detected current performance parameter with a second preset threshold C multiplied by D2 again to determine whether the current main control module is in a second overload state or not; if yes, reporting a secondary load warning, calculating the number M of the services to be forwarded according to K2-A-C in a time period T3, wherein K2 is a PID forwarding factor, and repeating the above process until the updated current performance parameter is smaller than a second preset threshold C multiplied by D2; and if not, the secondary load early warning is released, and the current main control module enters a load balancing state.

Example IV

Referring to fig. 6, fig. 6 is a schematic structural diagram of an apparatus for improving usability of a system according to an embodiment of the present invention, where the apparatus according to the embodiment of the present invention is applied to a storage system, and the storage system includes a plurality of main control modules, as shown in fig. 6, and the apparatus includes:

An obtaining module 401, configured to obtain, for a current master control module of the plurality of master control modules, a current performance parameter of the current master control module; the current main control module is any main control module of a plurality of main control modules, and the current performance parameter is used for representing the current service processing performance of the current main control module;

a determining module 402, configured to determine, based on the current performance parameter, a load state in which the current main control module is currently located;

an execution module 403, configured to execute a load balancing policy corresponding to the load state, so as to adjust the load state of the current main control module to a load balancing state;

In some possible embodiments, the determining module 402 further includes:

a first determining submodule, configured to determine that the load state is a first overload state if the current performance parameter exceeds a first preset threshold and does not exceed a second preset threshold;

And the second determining submodule is used for determining that the load state is a second overload state under the condition that the current performance parameter exceeds the second preset threshold value.

In some possible embodiments, the execution module 403 further includes:

a third determining sub-module, configured to determine a target number of currently unprocessed services to be forwarded based on the current performance parameter and a preset performance parameter of the main control module; the preset performance parameters are performance parameters of the main control module in the load balancing state;

and the first forwarding module is used for forwarding the target number of the services which are not processed currently in the main control module to a plurality of target main control modules.

In some possible embodiments, the execution module 403 further includes:

the acquisition sub-module is used for acquiring the current performance parameters of each main control module except the current main control module;

the third determining sub-module is used for determining at least one target main control module from a plurality of main control modules except the current main control module based on the current performance parameter of each main control module;

and the second forwarding module is used for forwarding at least one service which is not processed currently in the current main control module to the target main control module.

In some possible embodiments, the execution module 403 further includes:

and the sending module is used for sending a service forwarding instruction to the current main control module so that the current main control module forwards at least one currently unprocessed service to a plurality of target main control modules based on the service forwarding instruction.

According to the device for improving the availability of the system, the real-time performance of each main control module is monitored, when the overload risk of the current main control module is monitored, the corresponding balanced load strategy is executed based on the determined load state of the main control module, so that the main control module with the overload risk is adjusted to the load balanced state, and therefore, when the overload risk of the main control module exists, part of unprocessed business in the current main control module is forwarded to a plurality of main control modules, other main control modules except the current main control module balance the performances of the current main control module and other main control modules, overload of the current main control module is avoided, load unbalance among the main control modules is caused, and the system efficiency is affected; meanwhile, because the service among the plurality of main control modules can be mutually forwarded, the idle main control modules can be utilized even in an asymmetric service scene, and the utilization rate of the whole machine is improved.

Example five

Referring to fig. 7, fig. 7 is a schematic structural diagram of a storage system according to an embodiment of the present invention, as shown in fig. 7, the system includes:

the main control modules are in communication connection with each other, and are used for processing or forwarding service;

the balancing load management module is connected with a plurality of the main control modules, and is used for executing the method for improving the availability of the system according to any embodiment; the balanced load management module is a CIF management module and mainly applied to programmable logic devices such as ARM.

In some embodiments, the system further comprises:

the wireless module is connected with the serial port module and is used for converting signals of the serial port module into wireless signals such as WIFI and the like so that external equipment can interact information with the balanced load management module through the wireless signals.

In some embodiments, the system further comprises:

The front-end module is in communication connection with the main control module and is used for receiving service data sent by the service server and sending the service data to a plurality of main control modules.

In some embodiments, the system further comprises:

the indication module is in communication connection with the serial port module and is used for externally indicating the real-time state of the main control module.

According to the storage system provided by the embodiment of the invention, the real-time performance of each main control module is monitored, when the overload risk of the current main control module is monitored, the corresponding balanced load strategy is executed based on the determined load state of the main control module, so that the main control module with the overload risk is adjusted to the load balanced state, therefore, when the overload risk of the main control module exists, the unprocessed part of business in the current main control module is forwarded to a plurality of main control modules, and other main control modules except the current main control module balance the performance of the current main control module and the performance of other main control modules, so that the overload of the current main control module is avoided, the load among the plurality of main control modules is unbalanced, and the system efficiency is influenced; meanwhile, because the service among the plurality of main control modules can be mutually forwarded, the idle main control modules can be utilized even in an asymmetric service scene, and the utilization rate of the whole machine is improved. Based on the same inventive concept, the embodiment of the present invention further provides an electronic device, which may include a memory, a processor, and a computer program stored on the memory and capable of running on the processor, where the processor is configured to implement the method for improving usability of a system according to any of the foregoing embodiments.

Based on the same inventive concept, the embodiment of the present invention further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor executes the method for improving the usability of the system according to any of the foregoing embodiments.

Based on the same inventive concept, embodiments of the present invention also provide a computer-readable storage medium storing a computer program for executing the method of improving system usability as described in any of the above embodiments.

In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by differences from other embodiments, and identical and similar parts between the embodiments are all enough to be referred to each other.

It will be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the invention may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiment and all such alterations and modifications as fall within the scope of the embodiments of the invention.

Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or terminal device comprising the element.

The above detailed description of a method, apparatus, system, device and storage medium for improving usability of a system, provided by the present invention, has described the principles and embodiments of the present invention with specific examples applied therein, the description of the above examples is only for helping to understand the method and core idea of the present invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.

Claims

1. A method for improving availability of a system, the method being applied to a storage system, the storage system including a plurality of master control modules, the method comprising:

2. The method of claim 1, wherein determining, based on the current performance parameter, a current load state of the current master control module comprises:

3. The method of improving system availability of claim 2, wherein the executing a load balancing policy corresponding to the load status comprises:

4. A method for improving availability of a system according to any one of claims 1-3, wherein forwarding at least one service currently not processed in the master control module to a plurality of target master control modules comprises:

5. The method for improving system availability according to claim 1, wherein forwarding at least one service currently unprocessed in the master control module to a plurality of target master control modules comprises:

6. The method for improving system availability according to claim 1, wherein forwarding at least one service currently unprocessed in the master control module to a plurality of target master control modules comprises:

7. The method of improving system availability of claim 4, wherein the forwarding step comprises:

8. The method of improving system availability of claim 3, wherein after said forwarding at least one service currently unprocessed in said current master module to a plurality of said target master modules, said method further comprises:

9. The method for improving availability of system according to claim 4, wherein the services processed by the main control module have a plurality of service models; under the condition of processing the services of different service models, the preset performance parameters of the main control module are different; the method for acquiring the preset performance parameters comprises the following steps:

10. An apparatus for improving system availability, the apparatus comprising:

11. A storage system, the system comprising:

the balancing load management module is connected with a plurality of the main control modules, and is used for executing the method for improving the availability of the system according to any one of claims 1-9;

12. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor, when executed, implements a method of improving system usability as claimed in any of claims 1-9.

13. A computer readable storage medium storing a computer program for performing the method of improving the usability of a system according to any of claims 1-9.