Isolation method of remote replication relationship and electronic device

Abstract

The application discloses a method for isolating remote replication relations and an electronic device, relates to the technical field of remote replication, and comprises the following steps: determining the state types of a plurality of remote replication relations in a target data stream, wherein the target data stream comprises a plurality of mutually unrelated remote replication relations; determining a preset data stream to which a target remote replication relation is to be migrated according to the state types of the remote replication relations; migrating the target remote replication relation to the corresponding preset data stream, wherein the preset data stream is a first preset data stream or a second preset data stream; and if the preset data stream is the second preset data stream, performing resource recovery and cleaning operations on the target remote replication relation in the second preset data stream. The application classifies remote replication relations according to state types and puts them into different preset data streams, thereby significantly reducing the influence of resource cleaning triggered by state changes on the data replication efficiency of other remote replication relations.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of remote replication, in particular to a method for isolating remote replication relationship and an electronic device. BACKGROUND

[0002] In a centralized storage system, a remote replication relationship is a synchronization relationship between volumes of two storage systems, which is divided into three types: synchronous remote replication relationship, asynchronous remote replication relationship and periodic asynchronous remote replication relationship; a remote replication consistency group is a logical grouping, that is, to ensure that the states of remote replication relationships in the group are consistent, and a remote replication relationship can be selectively added or not added to a remote replication consistency group.

[0003] In related technologies, when a remote replication relationship is created between two systems, a partnership between the two systems is first established, multiple independent data streams are generated in the background for data transmission, multiple remote replication consistency groups are distributed to each data stream, and remote replication relationships in a remote replication consistency group are all uniformly distributed to the same data stream, that is, the remote replication relationships in the data stream are not associated with each other, and the states are inconsistent. However, when the data stream cleans up the resources of the remote replication relationship with a state change, all data packets are released indiscriminately; it is not possible to determine whether the remote replication relationship has data transmission. For the data stream in which all the remote replication relationships are not associated with each other, such indiscriminate cleaning can cause a large number of other normally running remote replication relationships to be easily affected by the state change of a single remote replication relationship, resulting in synchronization pause and reducing data replication efficiency. SUMMARY

[0004] The present application provides a method for isolating remote replication relationship and an electronic device to at least solve the problem of indiscriminate release of all data resources on the data stream in related technologies, which causes the resources of normally running remote replication relationships to be incorrectly recycled, affecting data synchronization efficiency.

[0005] In a first aspect, the present application provides a method for isolating remote replication relationship, comprising:

[0006] determining the state types of multiple remote replication relationships in a target data stream, the target data stream being a data stream established for data transmission between a first storage system and a second storage system, and the target data stream including multiple remote replication relationships that are not associated with each other;

[0007] determining a target remote replication relationship to be migrated in each remote replication relationship according to the state type of each remote replication relationship, and determining a preset data stream to which the target remote replication relationship is to be migrated; the target remote replication relationship being a remote replication relationship that affects the data replication efficiency of other remote replication relationships in the target data stream when resource recycling and cleaning operations are performed in the target data stream;

[0008] Migrate the target remote replication relationship to a corresponding preset data stream, the preset data stream being a first preset data stream or a second preset data stream; the first preset data stream is used to carry a remote replication relationship in a warning state; and the second preset data stream is used to carry a remote replication relationship in a fault state or a stop state.

[0009] If the preset data stream is the second preset data stream, perform resource recycling and cleaning operations on the target remote replication relationship in the second preset data stream.

[0010] In a second aspect, the present application provides an electronic device, comprising:

[0011] a memory for storing a computer program;

[0012] a processor for executing the computer program to implement the steps of the method of any one of the first aspect.

[0013] The remote replication relationship isolation method and the electronic device provided by the present application can determine which remote replication relationships may or have state changes by determining the state types of the remote replication relationships, define the remote replication relationships as target remote replication relationships, and then migrate the target remote replication relationships to different preset data streams according to different states of the target remote replication relationships, so that when resource recycling and cleaning operations are performed on the data streams due to state changes of the remote replication relationships, other remote replication relationships that are still in normal operation are not affected, and the impact on data replication efficiency is significantly reduced. BRIEF DESCRIPTION OF DRAWINGS

[0014] To make the application embodiments clearer, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described in the following are only some of the application, and other drawings can be obtained by those skilled in the art without creative effort.

[0015] Figure 1 The application provides a remote replication relationship isolation method and an electronic device.

[0016] Figure 2 The application provides a remote replication relationship isolation method and an electronic device.

[0017] Figure 3 The application provides a remote replication relationship isolation method and an electronic device.

[0018] Figure 4 The application provides a remote replication relationship isolation method and an electronic device.

[0019] Figure 5 A flowchart for migrating a remote replication relationship from a target data stream to a second preset data stream is provided for an embodiment of the present application;

[0020] Figure 6 A flowchart for migrating a remote replication relationship from a second preset data stream to a target data stream is provided for an embodiment of the present application;

[0021] Figure 7 A structural diagram of an isolation device for a remote replication relationship is provided for an embodiment of the present application;

[0022] Figure 8 A structural diagram of an electronic device is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, any other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0024] In order to enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0025] In a centralized storage system, a remote replication relationship undertakes the function of data synchronization between volumes of two storage systems, which can be divided into three types: synchronous, asynchronous and periodic asynchronous. As a logical grouping mechanism, the core role of a remote replication consistency group is to ensure that the states of all remote replication relationships in the group remain uniform (such as synchronous start and synchronous stop). In practical applications, remote replication relationships can be selected to join a certain remote replication consistency group according to business needs, or can be selected not to join any group. In related technical solutions, if a remote replication relationship needs to be established between two storage systems, a partnership between the systems needs to be built first. After the partnership is established, multiple independent data streams will be generated in the background to support subsequent data transmission; at the same time, multiple remote replication consistency groups will be allocated to different data streams, and those remote replication relationships that have not joined any remote replication consistency group will be allocated to the same data stream. The remote replication relationships in the data stream are independent of each other and have no association, and their states also do not need to be consistent. However, this solution has obvious problems: when a data stream detects a state change of a remote replication relationship and triggers resource cleaning, all data packets in the stream will be released without distinction, and it is impossible to identify whether the remote replication relationship is in data transmission. For such a data stream that aggregates a large number of unrelated remote replication relationships, the state change of a single remote replication relationship may cause other normally running remote replication relationships to be affected, causing data synchronization to stop, and ultimately reducing the overall data replication efficiency.

[0026] Therefore, in the face of the technical problems of the above related technologies, in order to accurately isolate abnormal remote replication relationships to avoid affecting normal data synchronization, first, for the problem that multiple remote replication relationships in a target data stream are not associated with each other but share resources, two types of preset data streams can be pre-set to clearly define the functional positioning of different preset data streams, providing a basis for subsequent migration operations, and then by judging the state type of each remote replication relationship, the target remote replication relationship that will affect other relationships can be accurately identified; second, in order to take differential processing according to the degree of abnormality to improve overall efficiency, a hierarchical migration strategy can be used. When the target remote replication relationship is identified, the corresponding preset data stream is determined according to the state type of the target remote replication relationship, the target relationship that needs to be temporarily isolated but continues to run is migrated to the first preset data stream to avoid interfering with the normal synchronization of the original data stream; at the same time, the target relationship that needs to be stopped and cleaned is migrated to the second preset data stream, and only resource recycling and cleaning operations are performed in this stream to ensure that the cleaning operation does not affect other normal relationships, thereby improving the stability and efficiency of data replication.

[0027] Figure 1 An application scenario diagram corresponding to a remote replication relationship isolation method provided by an embodiment of the present application is shown in FIG. 1. Figure 1As shown in the application scenario, the application scenario comprises: a first storage system 101, a second storage system 102, and a remote replication relationship isolation device 103, wherein the first storage system 101 and the second storage system 102 are communicatively connected through the remote replication relationship isolation device 103.

[0028] Optionally, the remote replication relationship isolation device 103 can be integrated alone, such as being integrated on a server, an electronic device, or integrated together with a storage system, and the present embodiment is not limited in this regard.

[0029] Specifically, when it is required to isolate the remote replication relationship between the first storage system 101 and the second storage system 102, the remote replication relationship isolation device 103 determines the state type of the remote replication relationship by monitoring the state of each remote replication relationship in the target data flow; then, according to the state type, the target remote replication relationship that will affect the synchronization of other relationships is screened out, and the preset data flow to be migrated is determined; subsequently, the remote replication relationship isolation device 103 migrates the target remote replication relationship to the corresponding preset data flow, and if the migration is to the second preset data flow, the remote replication relationship isolation device 103 performs resource recycling and cleaning operations on the target remote replication relationship in the data flow, so as to ensure that the remote replication relationship in the target data flow and the first preset data flow is not disturbed.

[0030] The technical solutions of the present application and how the technical solutions of the present application solve the above technical problems will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described again in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.

[0031] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic, for example, the division of the modules is only a logical function division, and actual implementation can have another division manner, for example, a plurality of modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed ones can be indirect coupling or communication connection through some interfaces, devices or modules, and can be electrical, mechanical or other forms.

[0032] Figure 2 A flowchart of a remote replication relationship isolation method provided by an embodiment of the present application is shown in FIG. 1. Figure 2As shown, the execution subject of the embodiment is a remote replication relationship isolation device, which can be realized by a computer program, a medium storing the related computer program, such as a U disk and / or an optical disk, or an entity device integrating or installing the related computer program, such as a chip or an electronic device. The electronic device can be a computer or a server, and the remote replication relationship isolation method provided by the embodiment includes the following steps:

[0033] S201, determining the state type of a plurality of remote replication relationships in a target data stream, the target data stream being a data stream established for data transmission between a first storage system and a second storage system, and the target data stream including a plurality of mutually independent remote replication relationships.

[0034] The target data stream refers to a data transmission channel established between the first storage system and the second storage system, which is specially used to carry remote replication relationships that are not joined in a remote replication consistency group, i.e., these relationships are independent of each other and their states do not need to be consistent, and all remote replication relationships of the non-consistency group are allocated to the data stream by default.

[0035] The remote replication relationship refers to the synchronization association between volumes in two storage systems, such as a primary volume and a secondary volume, which is divided into three types of synchronization, asynchronization, and periodic asynchronization according to the synchronization mode, and the core is to copy the data of the primary system to the secondary system.

[0036] Specifically, the remote replication relationship isolation device determines the state type of each remote replication relationship in the target data stream by real-time monitoring of the remote replication relationships and state changes.

[0037] Optionally, the time delay of read / write operations (Input / Output, IO) of each remote replication relationship in the target data stream is detected, i.e., the time from the entry of IO to the completion of processing. The ratio of the number of IOs with a time delay exceeding the normal range to the total number of IOs is periodically counted, and if it exceeds a preset threshold, it indicates that the current remote replication relationship state is poor. By recording the state of each remote replication relationship, the state type of the remote replication relationship is determined, such as determining that the state is normal if the state is poor but does not exceed the index threshold, and determining that the state is a fault state if the state is poor and exceeds the index threshold, so as to determine the state type of each remote replication relationship.

[0038] Optionally, it can also be that an instruction from a user is received, such as a user instruction to stop a certain remote replication relationship or to start a certain remote replication relationship, which is not limited in the embodiment.

[0039] S202, determine a target remote replication relationship to be migrated in each remote replication relationship according to a state type of each remote replication relationship, and determine a preset data stream to which the target remote replication relationship is to be migrated; the target remote replication relationship is a remote replication relationship that affects data replication efficiency of other remote replication relationships in a target data stream when resource recycling and cleaning operations are performed in the target data stream.

[0040] The target remote replication relationship refers to a remote replication relationship that needs to perform resource recycling and cleaning operations due to its own state change, and the operations affect the synchronization efficiency of other normal relationships in the target data stream. For example, the cleaning operation of a relationship in a warning, fault, or stop state indiscriminately releases data packets in the target data stream, resulting in interruption of synchronization of other normal relationships.

[0041] The preset data stream refers to a special data stream configured in advance and used to isolate the target remote replication relationship.

[0042] Specifically, after determining the state type of each remote replication relationship, which can include normal state, warning state, fault state, stop state, and start state, the remote replication relationship that can or will trigger resource recycling and cleaning operations is determined as the target remote replication relationship, such as the remote replication relationship in the warning state, fault state, stop state, and start state. Then, according to different state types and pre-set rules, such as migrating the warning state to a first preset data stream, migrating the fault state or stop state to a second preset data stream, and the like, the preset data stream to which each target remote replication relationship is to be migrated is determined.

[0043] S203, migrate the target remote replication relationship to the corresponding preset data stream, which is the first preset data stream or the second preset data stream; the first preset data stream is used to carry the remote replication relationship in the warning state; and the second preset data stream is used to carry the remote replication relationship in the fault state or stop state.

[0044] The first preset data stream can be used to carry the remote replication relationship that can have a state change, such as the target remote replication relationship in the warning state, and allows it to continue running and migrate back to the original target data stream after the state is restored.

[0045] The second preset data stream can be used to carry the remote replication relationship that will have a state change, such as the target remote replication relationship in the fault state or stop state, and is specially used for resource recycling and cleaning.

[0046] Optionally, in the second preset data stream, a fault log analysis module is added to automatically record fault causes and generate repair suggestions (such as network delay, storage performance bottleneck). Through fault cause analysis, the user can quickly locate the problem, improve system maintenance efficiency, and reduce downtime.

[0047] The migration refers to an operation of transferring the target remote replication relationship from the original target data stream to the corresponding preset data stream.

[0048] Specifically, after determining that the target remote replication relationship needs to be migrated to the preset data stream, the target remote replication relationship is safely transferred to the corresponding preset data stream without interrupting data synchronization and losing data, so as to realize isolation from the normal relationship in the original target data stream.

[0049] For example, the isolation device of the remote replication relationship sends a silence instruction to the target remote replication relationship, so that the target remote replication relationship enters a silence state: at this time, the relationship stops receiving new host IOs, and temporarily stores the new IOs in a waiting queue. The isolation device of the remote replication relationship queries the unfinished IOs of the target remote replication relationship on the original target data stream through an IO log; if there are unfinished IOs, the isolation device will continue to wait until all IOs are processed. When all residual IOs on the original target data stream are processed, the target remote replication relationship is switched from the silence state to a silence end state. The target remote replication relationship is migrated from the target data stream to the corresponding preset data stream, such as registration in the preset data stream.

[0050] S204, if the preset data stream is a second preset data stream, a resource recovery and cleaning operation is performed on the target remote replication relationship in the second preset data stream.

[0051] The resource recovery and cleaning operation refers to an operation of releasing various system resources occupied by the target remote replication relationship in the data transmission process and cleaning residual data, including releasing cache space, recovering network link resources, deleting invalid logs, and temporary data, so as to ensure that the resources can be reallocated to other normal relationships.

[0052] For example, it is first confirmed that the target remote replication relationship is in a complete silence or cleaning state and has no unfinished IOs, and the communication link of the second preset data stream with other streams is closed and the resource permission is locked. Then, the resource monitoring module identifies three types of resources, namely storage, network, and log, and generates a list by counting the occupied amount. The resources are recovered by type, the temporary data block is cleaned, the cache is released, the connection and bandwidth are closed, the logs are cleaned and the configuration records are deleted. It is verified whether the resources are released, the unreleased resources are retried, an alarm containing the result is sent to the administrator after cleaning, and finally the communication of the second preset data stream and the permission are restored, the recovered resources are included in the global resource pool, and if there are other relationships to be cleaned, the process is repeated.

[0053] The isolation method of the remote replication relationship provided by the embodiments of the present application can provide a basis for subsequent targeted processing of each remote replication relationship by accurately identifying the state type, avoid blind operation, and improve overall processing efficiency and accuracy. According to the state type of each remote replication relationship, the target remote replication relationship to be migrated in each remote replication relationship is determined, which can accurately find out the target remote replication relationship that will affect the data replication efficiency of other remote replication relationships in the target data stream when performing resource recycling and cleaning operations. The determination of the target remote replication relationship to be migrated to the preset data stream provides a clear direction for subsequent migration operations, and the target remote replication relationship can be migrated to the most suitable environment according to the characteristics and purposes of different preset data streams, which creates conditions for subsequent resource recycling, cleaning operations and optimization of overall data replication efficiency. The target remote replication relationship that affects the data replication efficiency of other remote replication relationships in the target data stream is migrated out, realizing the isolation of the problem relationship. The data replication process of the remaining remote replication relationships in the target data stream can be stably and efficiently performed, improving the reliability and efficiency of overall data replication.

[0054] As an optional implementation, on the basis of any one of the above embodiments, the state type of the plurality of remote replication relationships in the target data stream is determined, including:

[0055] In response to receiving a remote replication relationship stop request initiated by a user, the state type of the corresponding remote replication relationship is determined as a first state type according to the remote replication relationship stop request.

[0056] If it is determined that the remote replication relationship meets the preset fault detection standard, the state of the remote replication relationship meeting the preset fault detection standard is determined as a second state type.

[0057] The remote replication relationship stop request refers to an instruction initiated by a user through a management interface or a command line (such as a stoprcrelationship command) to terminate data synchronization of a specific remote replication relationship.

[0058] The preset fault detection standard refers to a rule for judging the abnormality of a remote replication relationship, such as an index based on IO latency, bad IO proportion, error count value, etc.

[0059] The first state type refers to the state of a remote replication relationship after a user initiatively initiates a stop request, such as a user stop state.

[0060] The second state type refers to the state of a remote replication relationship when it meets a fault detection standard, such as a fault state, a warning state, etc.

[0061] Specifically, the isolation device of the remote replication relationship receives a user-initiated remote replication relationship stop request through a management interface or a command line terminal; parses the request content to extract key information such as the ID of the remote replication relationship and the identity of the initiating user. The isolation device of the remote replication relationship queries the corresponding remote replication relationship according to the parsed ID and marks it as a first state type. It can be understood that fault detection is a continuous action from the start of the operation of the isolation device of the remote replication relationship. The isolation device of the remote replication relationship monitors all running remote replication relationships in the target data stream, judges whether the remote replication relationship meets the preset fault detection standard according to the preset fault detection rule, and marks the corresponding remote replication relationship as a second state type when it meets the standard.

[0062] The isolation method of the remote replication relationship provided by the embodiments of the present application sets the corresponding remote replication relationship to different state types for different situations, explicitly identifies the remote replication relationship in a specific state due to a user stop request, and can perform resource recycling and release operations on these relationships. The remote replication relationship is monitored in real time through the preset fault detection standard, and the relationship state is determined as the second state type as soon as the standard is met, which can timely discover potential faults that may affect data replication. The remote replication relationship that meets the fault detection standard is explicitly marked as a specific state, which provides a clear guide for the fault handling process.

[0063] As an optional implementation, on the basis of any one of the above embodiments, the second state type includes a first fault type and a second fault type, and if it is determined that the remote replication relationship meets the preset fault detection standard, the state of the remote replication relationship that meets the preset fault detection standard is determined as the second state type, including:

[0064] Determine the error count value of each remote replication relationship.

[0065] If the error count value is equal to the first preset threshold, the remote replication relationship with the error count value equal to the first preset threshold is determined as the first fault type.

[0066] If the error count value is equal to the second preset threshold, the remote replication relationship with the error count value equal to the second preset threshold is determined as the second fault type.

[0067] The first fault type refers to a slight abnormality of the remote replication relationship, such as a large IO latency fluctuation but still running.

[0068] The second fault type refers to a severe abnormality of the remote replication relationship, such as a large number of IO failures and unable to normally synchronize.

[0069] The error count value, i.e. bad_count, is used to quantify the abnormality degree of the remote replication relationship IO, and the higher the value is, the more serious the abnormality is.

[0070] The first preset threshold refers to a bad_count critical value for determining the remote replication relationship as the first fault type, and is set as 20, for example, representing a mild abnormality of the relationship that needs to be temporarily isolated.

[0071] The second preset threshold refers to a bad_count critical value for determining the remote replication relationship as the second fault type, and is set as 40, for example, representing a severe abnormality of the relationship that needs to be stopped for cleaning.

[0072] Specifically, the isolation device of the remote replication relationship periodically counts the number of error IOs for all started remote replication relationships in the target data stream, and if the proportion of bad IOs exceeds the preset index value, the error count value is increased by one; if the proportion of bad IOs does not exceed the preset index value, the error count value is decreased by one. The isolation device of the remote replication relationship maintains an independent error count record table for each remote replication relationship, and stores the current value of bad_count, the time stamp of each increase or decrease, and the trigger cause, etc. The isolation device of the remote replication relationship periodically scans the bad_count of all remote replication relationships, and compares it with the first preset threshold and the second preset threshold. If the bad_count of a certain remote replication relationship is equal to the first preset threshold, the relationship is determined as the first fault type. If the bad_count of a certain remote replication relationship is equal to the second preset threshold, the relationship is determined as the second fault type.

[0073] The isolation method of the remote replication relationship provided by the embodiments of the present application can subdivide the faults of the remote replication relationship into the first fault type and the second fault type by setting different preset thresholds, can more accurately identify the nature and severity of the faults, can identify and handle remote replication relationships of different fault types in time, can accurately distinguish the severity of the faults through error count hierarchical thresholds and multi-dimensional verification, can provide clear basis for subsequent differential processing of mild isolation and severe cleaning, can guarantee that normal relationships are not disturbed, and can avoid fault diffusion.

[0074] As an optional implementation, on the basis of any one of the above embodiments, determining the error count value of each remote replication relationship comprises:

[0075] Periodically calculating the normal distribution value of a plurality of read-write operations in each remote replication relationship by using a preset normal distribution algorithm according to a first preset time length.

[0076] If it is determined that the normal distribution value of the read-write operation is within the preset threshold range, the read-write operation with the normal distribution value within the preset threshold range is determined as an error operation.

[0077] The proportion of the error read-write operation in each remote replication relationship to the total read-write operation in the respective remote replication relationship is calculated to determine the proportion of the error read-write operation in each remote replication relationship.

[0078] If it is determined that the proportion of the current error read-write operation exceeds the preset proportion value, the error count value corresponding to the remote replication relationship exceeding the preset proportion value is incremented by one to obtain the current error count value corresponding to the remote replication relationship exceeding the preset proportion value.

[0079] If it is determined that the proportion of the current error read-write operation does not exceed the preset proportion value, the error count value corresponding to the remote replication relationship not exceeding the preset proportion value is decremented by one to obtain the current error count value corresponding to the remote replication relationship not exceeding the preset proportion value.

[0080] The first preset time length refers to the period of calculating the proportion of error read-write operation, such as 30 seconds, i.e. statistics is performed once every 30 seconds.

[0081] The preset normal distribution algorithm refers to calculating the mean (μ) and standard deviation (σ) of the read-write operation time delay based on the statistical normal distribution principle to determine the normal distribution value of each remote replication relationship.

[0082] For example, the formula corresponding to the preset normal distribution algorithm is shown in formula (1):

[0083] (1)

[0084] In the formula, x is the current time delay detection result value, μ is the mean, and σ is the standard deviation.

[0085] Optionally, a machine learning model (such as LSTM) can be used instead of the normal distribution algorithm to perform time series prediction on the IO delay of the remote replication relationship and dynamically adjust the threshold value.

[0086] The preset threshold range refers to the normal read-write operation time delay interval determined based on the normal distribution, such as being set to [μ-2σ, μ+2σ], and the operation exceeding the range is determined as an error operation.

[0087] The preset proportion value refers to the critical proportion for determining whether to change the error count value, such as being set to 33% (1 / 3).

[0088] For example, the isolation device of the remote replication relationship calculates the normal distribution value of the plurality of IOs in the remote replication relationship every 30 seconds using a preset normal distribution algorithm, that is, the time delay detection result of each IO is obtained every 30 seconds, the normal distribution value corresponding to each IO is calculated, and when the normal distribution value exceeds the preset threshold range, such as [μ-2σ, μ+2σ], it is determined that the IO is an error IO. The number of error IOs in each remote replication relationship is counted, the proportion of the number of error IOs to the total number of IOs in the respective remote replication relationship is calculated, and the proportion of error IOs in each remote replication relationship is obtained. If the proportion of error IOs in a remote replication relationship exceeds the preset proportion value, such as 1 / 3, it is determined that this period is a bad period, and the error count value corresponding to the remote replication relationship is increased by one. If the proportion of error IOs in a remote replication relationship does not exceed the preset proportion value, such as 1 / 3, it is determined that this period is a good period, and the error count value corresponding to the remote replication relationship is decreased by one. According to the above steps, the error count value of each remote replication relationship in the target data stream is determined in turn.

[0089] Optionally, a limit value of the error count value can be set, such as being increased by one and being maximized to 40, and being decreased by one and being minimized to 0.

[0090] The isolation method of the remote replication relationship provided by the embodiments of the present application can sensitively detect the changes, so as to determine the read-write operation with the normal distribution value within the preset threshold range as an error operation, thereby improving the accuracy of fault detection. The error count value is dynamically adjusted according to the comparison result of the proportion of the current error read-write operation and the preset proportion value, so as to more accurately judge the authenticity and severity of the fault, thereby avoiding the influence of misjudgment on the performance and guaranteeing the stable operation of the data replication service.

[0091] As an optional implementation, on the basis of any one of the above embodiments, before periodically calculating the normal distribution value of the plurality of read-write operations in each remote replication relationship using the preset normal distribution algorithm according to the first preset time length, the method further comprises:

[0092] Periodically collecting the time delay detection result of each read-write operation in each remote replication relationship in the target data stream according to a second preset time length.

[0093] Calculating the mean and standard deviation of each time delay detection result, so as to calculate the normal distribution value of each read-write operation based on the mean and standard deviation using the preset normal distribution algorithm.

[0094] The second preset time length refers to the period length of collecting the time delay detection result of the read-write operation, such as 10 minutes.

[0095] The time delay detection result refers to the time consumption record of a single read-write operation from entering the remote replication to being processed and completed.

[0096] Specifically, the isolation device of the remote replication relationship automatically triggers the time delay collection task at a second preset interval time length, collects time delay detection results of each read-write operation in each remote replication relationship in the target data stream, and calculates a mean value and a standard deviation based on the collected data, so that the pre-calculated mean value and standard deviation can be used to calculate a normal distribution value when determining an error count value.

[0097] Optionally, the sampling specimen is up to 1000, and sampling is stopped after enough sampling, up to 1 minute of sampling, if the sampling data is less than 80% (less than 800) in 1 minute, the sampling data is discarded. Starting from the third successful sampling, after each sampling, the mean μ and σ are calculated using the latest 3 sampling data (up to 3000). Up to 3 sampling data is retained, if sampling fails for 1 hour, all sampling data is discarded, and the mean μ and standard deviation σ are reset to 0.

[0098] It should be noted that only the time delay detection results of the read-write operation in the remote replication relationship on the target data stream are collected, and those in the preset data stream are not collected.

[0099] The isolation method of the remote replication relationship provided by the embodiment of the application periodically collects time delay detection results of each read-write operation in each remote replication relationship in the target data stream according to a second preset interval time length, which can obtain more detailed and meticulous basic data, and provide rich data support for subsequent fault analysis. By calculating the mean value and the standard deviation of each time delay detection result, the distribution characteristics of the read-write operation time delay can be more accurately determined, the mean value reflects the average level of the read-write operation time delay, and the standard deviation reflects the dispersion degree of the time delay. Based on the two accurately calculated parameters, the normal distribution value calculated by the preset normal distribution algorithm can more truly reflect the distribution of the read-write operation time delay, and thus the accuracy of fault detection is improved.

[0100] As an optional implementation, on the basis of any one of the above embodiments, the target remote replication relationship to be migrated to a preset data stream is determined, including:

[0101] If the target remote replication relationship is of the first fault type, the target remote replication relationship is determined to be migrated to the first preset data stream.

[0102] If the target remote replication relationship is of the second fault type and the first state type, the target remote replication relationship is determined to be migrated to the second preset data stream.

[0103] Specifically, when the target remote replication relationship is determined as the first failure type, it indicates that the remote replication relationship is likely to fail and needs to be migrated to the first preset data stream specially set for carrying such relationship. When the target remote replication relationship is determined as the second failure type and the first state type, it indicates that the remote replication relationship has failed or needs to be stopped, and needs to be migrated to the second preset data stream specially set for carrying the stopped remote replication relationship, so that the target remote replication relationship can perform resource recycling and cleaning operations therein, avoiding affecting other normal remote replication relationships.

[0104] For example, when the A remote replication relationship is determined as the first failure type, the A remote replication relationship is migrated from the target data stream to the first preset data stream. The specific process can be as shown in Figure 3 , including:

[0105] S301, changing the A remote replication relationship to a silent state.

[0106] S302, entering the host IO into a waiting queue.

[0107] S303, waiting for the IO processing of the A remote replication relationship on the target data stream to be completed.

[0108] S304, changing the A remote replication relationship to a silent end state.

[0109] S305, switching the A remote replication relationship from the target data stream to the first preset data stream.

[0110] S306, setting the A remote replication relationship to a normal state.

[0111] S307, processing all IOs in the waiting queue.

[0112] S308, the host IO no longer enters the waiting queue.

[0113] It can be understood that since the error count value is immediately migrated from the target data stream to the first preset data stream after reaching the first preset threshold, the second failure state is only possible to be migrated from the first preset data stream to the second preset data stream, and only when a user-initiated remote replication relationship stop request is received, the remote replication relationship is migrated from the target data stream to the second preset data stream.

[0114] For example, when the B remote replication relationship is determined as the second failure type, the B remote replication relationship is migrated from the first preset data stream to the second preset data stream. The specific process can be as shown in Figure 4 , including:

[0115] S401, changing the B remote replication relationship to a silent state.

[0116] S402, the host IO enters a waiting queue.

[0117] S403, waiting for the IO processing of the B remote replication relationship on the first preset data stream to be completed.

[0118] S404, changing the B remote replication relationship to a silent end state.

[0119] S405, switching the B remote replication relationship from the first preset data stream to the second preset data stream.

[0120] S406, setting the state of the B remote replication relationship to stop.

[0121] S407, triggering an alarm to inform the user.

[0122] It can be understood that when the user initiates a remote replication relationship stop request, the stop command is not executed first, but the stop command is executed first, that is, the corresponding target remote replication relationship is migrated to the second preset data stream first, and then the stop command is executed.

[0123] For example, when it is determined that the C remote replication relationship is of the first state type, the C remote replication relationship is migrated from the target data stream to the second preset data stream, and the specific process can be as shown in Figure 5 , including:

[0124] S501, changing the C remote replication relationship to a silent state.

[0125] S502, the host IO enters a waiting queue.

[0126] S503, waiting for the IO processing of the C remote replication relationship on the target data stream to be completed.

[0127] S504, changing the C remote replication relationship to a silent end state.

[0128] S505, switching the C remote replication relationship from the target data stream to the first preset data stream.

[0129] S506, setting the C remote replication relationship to a normal state.

[0130] S507, processing all IOs in the waiting queue.

[0131] S508, the host IO no longer enters the waiting queue.

[0132] S509, executing the stop command to immediately stop the C remote replication relationship.

[0133] Optionally, when initiating the remote replication relationship stop request, the user can first initiate a pre-stop request and then initiate a stop request, or can issue them together.

[0134] The isolation method of the remote replication relationship provided by the embodiments of the present application can, when the target remote replication relationship is of the first fault type, migrate it to the first preset data stream. For the target remote replication relationship of the second fault type and the first state type, migrating it to the second preset data stream can prevent the fault from further spreading. Through this step, the preset data stream can be accurately matched based on the state type, which not only ensures that the slightly abnormal relationship does not affect the original data stream, but also centrally manages the relationships that need to be stopped and cleaned up, thereby maximizing the reduction of interference on normal synchronization.

[0135] As an optional implementation, on the basis of any one of the above embodiments, the method further comprises:

[0136] If it is determined that the error count value of the remote replication relationship in the first preset data stream is equal to the second preset threshold, the remote replication relationship with the error count value equal to the second preset threshold is migrated to the second preset data stream.

[0137] If it is determined that the error count value of the remote replication relationship in the first preset data stream is less than the first preset threshold, the remote replication relationship with the error count value less than the first preset threshold is migrated to the target data stream.

[0138] It can be understood that, after the remote replication relationship is migrated to the first preset data stream, the change of the error count value thereof still needs to be monitored.

[0139] Specifically, when the error count value of the remote replication relationship in the first preset data stream is equal to the second preset threshold, it indicates that the fault degree of the relationship is aggravated, and the relationship needs to be immediately migrated to the second preset data stream; if the error count value of the remote replication relationship in the first preset data stream is less than the first preset threshold, it indicates that the fault degree of the relationship is alleviated, and the relationship is migrated back to the target data stream. The specific error count value determination method is the same as described above.

[0140] Optionally, in order to prevent the remote replication relationship from being frequently moved between the target data stream and the first preset data stream, the remote replication relationship is not allowed to be moved again within 5 minutes after being moved to the first preset data stream or the target data stream.

[0141] The isolation method of the remote replication relationship provided in the embodiments of the present application can avoid the high error rate relationship from continuing to run in the original data stream when the error count value of the remote replication relationship in the first preset data stream is equal to the second preset threshold value, indicating that the relationship has relatively frequent errors. If the error count value of the remote replication relationship is less than the first preset threshold value, it indicates that the relationship runs relatively stably and has fewer errors. Migrating the relationship to the target data stream can make the low error rate relationship run in a more suitable environment, further reducing the risk of data errors. Through this step, the first preset data stream can dynamically adjust the relationships it carries, avoiding the deterioration of mild abnormalities into severe failures, and returning the normal relationships to the conventional running environment, thereby maximizing the overall remote replication efficiency.

[0142] As an optional implementation, on the basis of any one of the above embodiments, the preset data stream further includes a third preset data stream, and the method further includes:

[0143] If it is determined that the error count value of the remote replication relationship in the first preset data stream is equal to the third preset threshold value, the remote replication relationship with the error count value equal to the third preset threshold value is migrated to the third preset data stream.

[0144] If it is determined that the error count value of the remote replication relationship in the third preset data stream is equal to the second preset threshold value, the remote replication relationship with the error count value equal to the second preset threshold value is migrated to the second preset data stream.

[0145] If it is determined that the error count value of the remote replication relationship in the first preset data stream is less than the first preset threshold value, the remote replication relationship with the error count value less than the first preset threshold value is migrated to the target data stream.

[0146] The third preset threshold value is a critical value of the error count for determining that the remote replication relationship is upgraded from a mild abnormality to a moderate abnormality, and is set to 30, for example, between the first preset threshold value and the second preset threshold value.

[0147] The third preset data stream is between the first preset data stream and the second preset data stream, and is specially used to carry the remote replication relationship with the error count value reaching the third preset threshold value, so as to realize three-level isolation of mild, moderate and severe abnormalities, avoid rapid spread of abnormalities, and reserve buffer time for troubleshooting.

[0148] It can be understood that after the remote replication relationship is migrated to the first preset data stream and the third preset data stream, the change of the error count value of the remote replication relationship still needs to be monitored.

[0149] Specifically, when the error count value of the remote replication relationship in the first preset data stream is equal to the third preset threshold value, it indicates that the fault degree of the relationship is aggravated, and the relationship needs to be migrated to the third preset data stream; if the error count value of the remote replication relationship in the third preset data stream is equal to the second preset threshold value, it indicates that the fault degree of the relationship is further aggravated, and the relationship needs to be migrated to the second preset data stream; if the error count value of the remote replication relationship in the first preset data stream is less than the first preset threshold value, it indicates that the fault degree of the relationship is alleviated, and the relationship needs to be migrated back to the target data stream. The specific error count value determination method is consistent with the above.

[0150] It can be understood that if the error count value of the remote replication relationship in the third preset data stream is less than the third preset threshold value, the corresponding remote replication relationship can be migrated back to the first preset data stream.

[0151] Optionally, an automatic recovery mechanism can be introduced in the second preset data stream. When the remote replication relationship is sampled as a good period for three times in succession, the remote replication relationship is automatically moved back to the target data stream. No manual intervention of the user is needed, the self-healing capability of the system is improved, and the resource occupation of the second preset data stream is reduced.

[0152] The isolation method of the remote replication relationship provided in the embodiment of the application divides special data streams according to the abnormal degree, and avoids mixed flow operation of relationships in different states. The slightly abnormal relationship is isolated in the first preset data stream, and does not occupy the target data stream resource. The moderately abnormal relationship is monitored in the third preset data stream, and is prevented from directly deteriorating to the severe fault to affect the whole system, and the fault propagation path is cut off from the physical layer. Resources are allocated according to needs in different data streams, the target data stream focuses on efficient synchronization of normal relationships, the first and third preset data streams only provide necessary resources for abnormal relationships, and the second preset data stream concentrates on cleaning up resources and releasing redundant occupation.

[0153] As an optional implementation, on the basis of any one of the above embodiments, the method further includes:

[0154] In response to receiving a remote replication relationship start request or a remote replication relationship creation request initiated by the user, a corresponding remote replication relationship is searched for or created in the second preset data stream according to the remote replication relationship start request or the remote replication relationship creation request.

[0155] If it is determined that the state of the remote replication relationship is a silent end, the remote replication relationship in the silent end state is migrated from the second preset data stream to the target data stream, so that the remote replication relationship starts data replication.

[0156] The remote replication relationship start request or the remote replication relationship creation request refers to an instruction initiated by a user through a management interface, a command line, or the like, and is used for newly creating a remote replication relationship or restarting a stopped relationship, and can include target primary volume, secondary volume information, synchronization parameters, and the like.

[0157] The silent end state refers to a final state of the remote replication relationship after completing all residual IO processing and resource state verification, indicating that the relationship has been completely cleaned up and has no data consistency risk, and can be safely migrated to the target data stream to start synchronization.

[0158] Specifically, since all stopped remote replication relationships have been migrated to the second preset data stream, starting will not affect other normal remote replication relationships. For example, if a certain remote replication relationship is to be started, the specific relationship is queried according to the information in the request, and the relationship is migrated from the second preset data stream to the target data stream, so that the relationship starts data replication. Since the newly created remote replication relationship is in a stopped state, in order to unify the processing, the newly created remote replication relationship is placed in the second preset data stream by default. After the silent state, the replication relationship is migrated to the target data stream, so that the relationship starts data replication.

[0159] For example, when a D remote replication relationship start request initiated by a user is received, the D remote replication relationship is migrated from the second preset data stream to the target data stream. The specific process can be as shown in Figure 6 , including:

[0160] S601, change the D remote replication relationship to a silent state.

[0161] S602, enter the host IO into a waiting queue.

[0162] S603, change the D remote replication relationship to a silent end state.

[0163] S604, start the D remote replication relationship.

[0164] S605, switch the D remote replication relationship from the second preset data stream to the target data stream.

[0165] S606, set the D remote replication relationship to a normal state.

[0166] S607, process all IOs in the waiting queue.

[0167] S608, the host IO no longer enters the waiting queue.

[0168] It can be understood that, since there is no IO on the data stream path, the D remote replication relationship directly enters the silent end state.

[0169] The method for isolating a remote replication relationship provided in the embodiments of the present application completes relationship creation or start configuration in the second preset data stream, avoids configuration conflicts caused by direct operation in the target data stream, and reduces interference on other remote replication relationships in the target data stream. The silent end state ensures that the relationship has no residual data and complete configuration, and can be directly started after migration to the target data stream, thereby reducing the risk of data inconsistency. The user can pre-configure and modify parameters in the second preset data stream without interrupting normal synchronization of other relationships in the target data stream, thereby improving operation flexibility.

[0170] Figure 7 A structural diagram of an isolation device for a remote replication relationship provided in an embodiment of the present application is shown in FIG. 1. The isolation device for a remote replication relationship provided in the embodiment is located in an electronic device. The isolation device for a remote replication relationship provided in the embodiment includes a determination module 71, a migration module 72, and an execution module 73. Figure 7

[0171] Specifically, the determination module 71 is configured to determine state types of a plurality of remote replication relationships in a target data stream, the target data stream being a data stream established for data transmission between a first storage system and a second storage system, and the target data stream including a plurality of remote replication relationships that are not associated with each other; and determine, according to the state types of the remote replication relationships, a target remote replication relationship to be migrated in the remote replication relationships, and a preset data stream to which the target remote replication relationship is to be migrated; the target remote replication relationship being a remote replication relationship that affects data replication efficiency of other remote replication relationships in the target data stream when resource recycling and cleaning operations are performed in the target data stream; the migration module 72 is configured to migrate the target remote replication relationship to the corresponding preset data stream, the preset data stream being a first preset data stream or a second preset data stream; the first preset data stream being used to carry a remote replication relationship in a warning state; and the second preset data stream being used to carry a remote replication relationship in a fault state or a stop state; and the execution module 73 is configured to perform, if the preset data stream is the second preset data stream, resource recycling and cleaning operations on the target remote replication relationship in the second preset data stream.

[0172] Optionally, in the determination of the state types of the plurality of remote replication relationships in the target data stream, the determination module 71 is specifically configured to: in response to receiving a remote replication relationship stop request initiated by a user, determine, according to the remote replication relationship stop request, a state type of a corresponding remote replication relationship as a first state type; and if it is determined that the remote replication relationship meets a preset fault detection standard, determine a state of the remote replication relationship that meets the preset fault detection standard as a second state type.

[0173] ​Optionally, the second state type comprises a first fault type and a second fault type, and the determining module 71 is configured to: determine an error count value of each remote replication relationship; determine a remote replication relationship with an error count value equal to a first preset threshold as the first fault type; and determine a remote replication relationship with an error count value equal to a second preset threshold as the second fault type, when the determining module 71 determines that the remote replication relationship meets the preset fault detection criterion.

[0174] Optionally, when determining the error count value of each remote replication relationship, the determining module 71 is configured to: periodically calculate a normal distribution value of a plurality of read-write operations in each remote replication relationship according to a first preset time period using a preset normal distribution algorithm; determine a read-write operation with a normal distribution value within a preset threshold range as an error operation when the determining module 71 determines that the normal distribution value of the read-write operation is within the preset threshold range; calculate a proportion of the number of error read-write operations in each remote replication relationship to the total number of read-write operations in the respective remote replication relationship to determine the proportion of error read-write operations in each remote replication relationship; and increase the error count value of a remote replication relationship that exceeds the preset proportion value by one to obtain a current error count value of the remote replication relationship that exceeds the preset proportion value when the determining module 71 determines that the current proportion of error read-write operations exceeds the preset proportion value; and decrease the error count value of a remote replication relationship that does not exceed the preset proportion value by one to obtain a current error count value of the remote replication relationship that does not exceed the preset proportion value when the determining module 71 determines that the current proportion of error read-write operations does not exceed the preset proportion value.

[0175] Optionally, the isolation device of the remote replication relationship further comprises a collecting module and a calculating module.

[0176] Correspondingly, the collecting module is configured to periodically collect a time delay detection result of each read-write operation in each remote replication relationship in the target data stream according to a second preset time period; and the calculating module is configured to calculate a mean value and a standard deviation of each time delay detection result to calculate a normal distribution value of each read-write operation based on the mean value and the standard deviation using a preset normal distribution algorithm.

[0177] Optionally, when determining the preset data stream to which the target remote replication relationship is to be migrated, the determining module 71 is configured to: determine that the target remote replication relationship is to be migrated to a first preset data stream when the target remote replication relationship is of the first fault type; and determine that the target remote replication relationship is to be migrated to a second preset data stream when the target remote replication relationship is of the second fault type and the first state type.

[0178] Optionally, the migration module 72 is further configured to: if it is determined that the error count value of the remote replication relationship in the first preset data flow is equal to the second preset threshold, migrating the remote replication relationship with the error count value equal to the second preset threshold to the second preset data flow; and if it is determined that the error count value of the remote replication relationship in the first preset data flow is less than the first preset threshold, migrating the remote replication relationship with the error count value less than the first preset threshold to the target data flow.

[0179] Optionally, the preset data flow further includes a third preset data flow, and the migration module 72 is further configured to: if it is determined that the error count value of the remote replication relationship in the first preset data flow is equal to a third preset threshold, migrating the remote replication relationship with the error count value equal to the third preset threshold to the third preset data flow; if it is determined that the error count value of the remote replication relationship in the third preset data flow is equal to the second preset threshold, migrating the remote replication relationship with the error count value equal to the second preset threshold to the second preset data flow; and if it is determined that the error count value of the remote replication relationship in the first preset data flow is less than the first preset threshold, migrating the remote replication relationship with the error count value less than the first preset threshold to the target data flow.

[0180] Optionally, the isolation apparatus of the remote replication relationship further includes a creation module and a search module.

[0181] Correspondingly, the creation module and the search module are configured to: in response to receiving a user-initiated remote replication relationship starting request or a remote replication relationship creation request, searching for or creating a corresponding remote replication relationship in the second preset data flow according to the remote replication relationship starting request or the remote replication relationship creation request; and the migration module 72 is further configured to: if it is determined that the state of the remote replication relationship is silent end, migrating the remote replication relationship with the state of silent end from the second preset data flow to the target data flow, so that the remote replication relationship starts data replication.

[0182] It should be noted that the technical effects of the isolation apparatus of the remote replication relationship in the present embodiment have been explained in the above-mentioned embodiments of the isolation method of the remote replication relationship, and therefore, the present embodiment will not be described again.

[0183] Figure 8 A structural schematic diagram of an electronic device provided by an embodiment of the present application is shown in FIG. 8. Figure 8 As shown in FIG. 8, the electronic device 80 provided by an embodiment of the present application includes a memory 81 and a processor 82.

[0184] The memory 81 stores a computer program, and the processor 82 is configured to run the computer program to perform the steps in any of the above-mentioned embodiments of the isolation method of the remote replication relationship.

[0185] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is configured to execute the steps in the above-mentioned any one of the isolation methods of remote copy relationship.

[0186] In an example embodiment, the above-mentioned computer readable storage medium can include, but is not limited to, a U disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0187] The embodiment of the present application further provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the steps in the above-mentioned any one of the isolation methods of remote copy relationship.

[0188] The embodiment of the present application further provides another computer program product, which comprises a non-volatile computer readable storage medium, and the non-volatile computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps in the above-mentioned any one of the isolation methods of remote copy relationship.

[0189] The skilled in the art can further realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized in electronic hardware, computer software or combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been described in general in the above description. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. The skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0190] The above provides a detailed introduction to the isolation method of remote copy relationship and the electronic device. The principle and implementation mode of the present application are described by applying specific examples in the present application. The above description of the examples is only used to help understand the method of the present application and its core idea. It should be pointed out that, for the ordinary skilled in the art, some improvements and modifications can be made to the present application without departing from the principle of the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims

1. A method for isolating a remote copy relationship, the method comprising: The method comprises the following steps: determining the state types of a plurality of remote replication relationships in a target data stream, the target data stream being a data stream established for data transmission between a first storage system and a second storage system, the target data stream comprising a plurality of mutually independent remote replication relationships; determining a target remote replication relationship to be migrated in each of the remote replication relationships according to the state types of the remote replication relationships, and determining a preset data stream to which the target remote replication relationship is to be migrated; the target remote replication relationship being a remote replication relationship that affects the data replication efficiency of other remote replication relationships in the target data stream when resource recycling and cleaning operations are performed on the target data stream; migrating the target remote replication relationship to a corresponding preset data stream, the preset data stream being a first preset data stream or a second preset data stream; the first preset data stream being used to carry remote replication relationships in a warning state; the second preset data stream being used to carry remote replication relationships in a fault state or a stop state; if the preset data stream is the second preset data stream, performing resource recycling and cleaning operations on the target remote replication relationship in the second preset data stream.

2. The method of claim 1, wherein, The method comprises the following steps: in response to receiving a user-initiated remote replication relationship stop request, determining the state type of the corresponding remote replication relationship as a first state type according to the remote replication relationship stop request; if it is determined that a remote replication relationship meets a preset fault detection standard, determining the state of the remote replication relationship meeting the preset fault detection standard as a second state type.

3. The method of claim 2, wherein, The second state type comprises a first fault type and a second fault type, and the step of determining the state of the remote replication relationship meeting the preset fault detection standard as the second state type comprises the following steps: determining an error count value of each of the remote replication relationships; if the error count value is equal to a first preset threshold value, determining the remote replication relationship with the error count value equal to the first preset threshold value as the first fault type; if the error count value is equal to a second preset threshold value, determining the remote replication relationship with the error count value equal to the second preset threshold value as the second fault type.

4. The method of claim 3, wherein, The method comprises the following steps: periodically calculating normal distribution values of a plurality of read-write operations in each of the remote replication relationships according to a first preset time period and using a preset normal distribution algorithm; if it is determined that the normal distribution values of the read-write operations are within a preset threshold value range, determining the read-write operations with the normal distribution values within the preset threshold value range as error operations; calculating the proportion of the number of error read-write operations in each of the remote replication relationships to the total number of read-write operations in the respective remote replication relationship to determine the proportion of error read-write operations in each of the remote replication relationships; if it is determined that the proportion of the current error read-write operations exceeds a preset proportion value, increasing the error count value corresponding to the remote replication relationship exceeding the preset proportion value by one to obtain the current error count value corresponding to the remote replication relationship exceeding the preset proportion value; If it is determined that the proportion of the current error read-write operation does not exceed the preset proportion value, the error count value of the remote replication relationship corresponding to the preset proportion value that does not exceed the preset proportion value is reduced by one to obtain the current error count value of the remote replication relationship corresponding to the preset proportion value that does not exceed the preset proportion value.

5. The method of claim 4, wherein, Before the normal distribution value of each read-write operation in each remote replication relationship is calculated by using the preset normal distribution algorithm periodically according to the first preset time length, the method further comprises the steps of: Periodically collecting the time delay detection results of each read-write operation in each remote replication relationship in the target data stream according to a second preset time length; Calculating the mean and standard deviation of each time delay detection result, and calculating the normal distribution value of each read-write operation based on the mean and standard deviation by using a preset normal distribution algorithm.

6. The method of claim 3, wherein, The method further comprises the steps of: If the target remote replication relationship is of a first fault type, it is determined that the target remote replication relationship is to be migrated into a first preset data stream; If the target remote replication relationship is of a second fault type and a first state type, it is determined that the target remote replication relationship is to be migrated into a second preset data stream.

7. The method of claim 6, wherein, The method further comprises the steps of: If it is determined that the error count value of the remote replication relationship in the first preset data stream is equal to a second preset threshold value, the remote replication relationship with the error count value equal to the second preset threshold value is migrated into the second preset data stream; If it is determined that the error count value of the remote replication relationship in the first preset data stream is less than a first preset threshold value, the remote replication relationship with the error count value less than the first preset threshold value is migrated into the target data stream.

8. The method of claim 6, wherein, The preset data stream further comprises a third preset data stream, and the method further comprises the steps of: If it is determined that the error count value of the remote replication relationship in the first preset data stream is equal to a third preset threshold value, the remote replication relationship with the error count value equal to the third preset threshold value is migrated into the third preset data stream; If it is determined that the error count value of the remote replication relationship in the third preset data stream is equal to the second preset threshold value, the remote replication relationship with the error count value equal to the second preset threshold value is migrated into the second preset data stream; If it is determined that the error count value of the remote replication relationship in the first preset data stream is less than the first preset threshold value, the remote replication relationship with the error count value less than the first preset threshold value is migrated into the target data stream.

9. The method of claim 1, wherein, The method further comprises the steps of: In response to receiving a user-initiated remote replication relationship start request or a remote replication relationship creation request, searching for or creating a corresponding remote replication relationship in the second preset data stream according to the remote replication relationship start request or the remote replication relationship creation request; If it is determined that the state of the remote replication relationship is silent end, the remote replication relationship with the state of silent end is migrated from the second preset data stream to the target data stream, so that the remote replication relationship starts data replication.

10. An electronic device, comprising: The method further comprises the steps of: a memory for storing a computer program; a processor for executing the computer program to implement the steps of the remote replication relationship isolation method according to any one of claims 1 to 9.

Images