CN106909307B - Method and device for managing double-active storage array - Google Patents

Abstract

The present invention relates to a method and device for managing a dual-active storage array. The method includes: counting the total amount of data sent by a host received by a first storage array. Receive the total amount of data sent by the second storage array receiving host from the second storage array. It is determined that the total amount of data sent by the host received by the first storage array is smaller than the total amount of data sent by the host received by the second storage array. When it is determined that the link between the first storage array and the second storage array is faulty, an arbitration request is sent to the arbitration server after a period of delay, and the arbitration server is respectively connected to the first storage array and the second storage array. Receive the arbitration result sent by the arbitration server. When it is determined that the second storage array is disconnected from the arbitration server, the storage array that receives the data data is determined through the storage array. During split-brain, the storage array will designate the storage array with the most total host data to receive host data.

Description

Method and device for managing active-active storage array

technical field

The present invention relates to the technical field of memory, and in particular, to a method and device for managing a dual-active storage array.

Background technique

With the advancement of digitalization in various industries, data has gradually become the core of the operation of enterprises and institutions. Users have higher and higher requirements for the stability of storage systems that carry data, including the reliability of stored data and business continuity. More and more enterprises are processing applications, data, and systems in a centralized manner. at risk. Although many storage manufacturers can provide users with highly stable storage devices, they cannot prevent emergencies such as natural disasters.

When a catastrophic emergency occurs, how to ensure the online linearity of the core business of the enterprise, that is, the uninterrupted operation of the core business 7*24 hours, has become the primary concern of the enterprise. Unplanned downtime caused by emergencies is inevitable, and irreversible factors such as data center power outages cause irreversible damage to production systems. In order to ensure the reliability of data storage and business continuity, the active-active technology emerged as the times require. The active-active technology can ensure the continuity of services to the greatest extent within a certain range of failure modes.

Active-active technology is oriented to application access. It requires that it provides uninterrupted access response for storage across devices/data centers. Under the fault model within the system's fault tolerance range, its IO reset time cannot cause service interruption. Therefore, the core requirements of the active-active technology are: storage access view consistency (a host accesses an active-active LUN (logical unit number, logical unit number) from the port of any controller, and its access rights and data are consistent) and faults The post-service switching delay cannot cause service interruption.

Since active-active LUNs are deployed on two arrays, only one LUN on one array is required to provide services when a split-brain occurs between the two active-active arrays.

In the prior art, a third-party arbitration mechanism is usually introduced: when a split-brain occurs, the storages at both ends submit arbitration to the arbitration server respectively, the LUN that wins the arbitration is allowed to be accessed by the host, and the LUN that fails the arbitration is not allowed to access.

However, in the prior art, the arbitration scheme has very high requirements on the reliability of the arbitration server. Once the arbitration server fails or the link between the arbitration server and the two storages fails, the split-brain of the active-active array will result in two The active-active array fails to provide access even if it fails to submit arbitration, resulting in service interruption of the host.

SUMMARY OF THE INVENTION

This application describes a method for managing an active-active storage array. With the method, after the storage array in the active-active array is split-brained, the method avoids as much as possible the purpose of replacing the storage array that receives data sent by the host.

In one aspect, an embodiment of the present application provides a method for managing a dual-active storage array. The method includes counting the total amount of host data received by the first storage array in the active-active array, and receiving the total amount of received host data counted by the second storage array in the active-active array. A comparison is made of how much of the total amount of host data received by the first storage array and the second storage array. When the link between the first storage array and the second storage array is disconnected, the storage array that receives the most data from the host continues to receive the data sent by the host. In the method, the storage array that receives more host data continues to receive data sent by the host, and the storage array that receives more host data is usually the storage array that is currently receiving host data. In this way, the storage system can be connected normally, and the replacement of the storage array that receives the data sent by the host can be avoided.

In one possible design, the first storage array and the second storage array are also each connected to an arbiter server. When the link between the first storage array and the second storage array is disconnected, an arbitration request is respectively sent to the arbitration server, and the data of the receiving host is determined through the arbitration request. Storage arrays that receive a large amount of host data will send the arbitration request immediately after detecting that they are disconnected from the arbitration server. Storage arrays that receive a small amount of host data will send the arbitration request after a preset time delay after detecting that they are disconnected from the arbitration server. The arbitration server determines the storage array that receives the host data according to the sequence of receiving the arbitration request. The storage array that first sends the arbitration request to the arbitration server obtains the data of the receiving host, and then the storage array that sends the arbitration request to the arbitration server does not receive the data sent by the host. In this way, the storage array that receives more host data will preferentially receive the storage array with less host data to continue to receive host data for a certain period of time. The storage array receives host data. This avoids that when the storage array that receives more host data fails and cannot receive host data, the storage array that receives less host data also cannot receive host data.

In one possible design, the delay for a preset period of time may be any time from 3 seconds to 10 seconds. If the delay time is too short, the purpose of giving priority to the storage array receiving more host data cannot be achieved. If the delay time is too long, the continuity of the system receiving host data will be affected.

In one possible design, the delay for a preset period of time may be a delay of 5 seconds. The time can not only effectively ensure the priority of the storage array that receives more host data, but also does not affect the continuity of the system receiving host data.

On the other hand, an embodiment of the present application provides a method for managing an active-active storage array. The method includes counting the total amount of host data received by the first storage array in the active-active array, and receiving the total amount of received host data counted by the second storage array in the active-active array. It is determined that the total amount of host data received by the first storage array is greater than the total amount of host data received by the second storage array. When the communication between the first storage array and the arbiter server fails, and the communication between the second storage array arbiter server fails, when the link transmission between the first storage array and the second storage array fails, the receiving host The first storage array with a large amount of data continues to receive the data sent by the data. This ensures that when the first storage array is disconnected from the arbiter server and the connection between the second storage array and the arbiter server is disconnected, when the link between the first storage array and the second storage array is disconnected, the storage array can Judge for yourself. The storage array that receives more data from the host continues to receive the data sent by the host, and the storage array that receives less data from the host does not receive the data sent by the host. It is avoided that when the storage array that receives more host data fails and cannot receive host data, the storage array that receives less host data also cannot receive host data.

In another aspect, embodiments of the present application provide a storage array in a dual-active storage array. The storage array includes methods for performing any one of the embodiments included in the methods of an aspect.

In yet another aspect, embodiments of the present application provide a storage array in a dual-active storage array. The storage array includes embodiments for performing the methods included in another aspect.

In yet another aspect, embodiments of the present application provide a storage array in a dual-active storage array. A storage array includes processors, memory, communication interfaces, and buses. Among them, the processor, memory and communication interface communicate through the bus. The communication interface is used to communicate with the second storage array and the host. Memory is used to store programs. When the apparatus is in operation, the processor is configured to execute a program stored in the memory to perform the above-described method.

In yet another aspect, an embodiment of the present application provides a method for managing an active-active storage array. The method includes determining whether the current first storage array receives data sent by the host. If the storage array is currently receiving data sent by the host, after the link between the first storage array and the second storage array is disconnected, it continues to receive data sent by the host. If the storage array does not currently receive the data sent by the host, after the link between the first storage array and the second storage array is disconnected, the data sent by the host will not be received. By adopting the aspect, it is directly determined that the storage array that receives the host data continues to receive the data sent by the host, thereby avoiding other methods and devices, and making the system more efficient. It also avoids the hassle of switching storage arrays that receive host data.

The present application provides a method and apparatus for managing an active-active storage array. This minimizes the need to replace the storage array receiving host data when the connection between the storage arrays is disconnected.

Description of drawings

FIG. 1 is a connection diagram of a dual-active storage array according to an embodiment of the present invention;

FIG. 2 is a flowchart of a method for managing a dual-active storage array provided by an embodiment of the present invention;

Fig. 3 is a system connection diagram after a failure of a dual-active storage array arbitration server provided by an embodiment of the present invention;

Fig. 4 is the system connection diagram after the disconnection of a kind of active-active array arbitration server connection provided by the implementation of the present invention;

5 is a flowchart of a method for managing an active-active storage array after the connection between the active-active storage array and the arbitration server is interrupted, provided by an embodiment of the present invention;

FIG. 6 is a flowchart of managing a dual-active storage array according to an embodiment of the present invention;

FIG. 7 is a structural diagram of a storage array for managing active-active arrays according to an embodiment of the present invention;

8 is a structural diagram of a management active-active storage array after an arbitration server failure according to an embodiment of the present invention;

FIG. 9 is a schematic diagram of a storage array device according to an embodiment of the present invention.

Detailed ways

The technical solutions of the present invention will be further described in detail below through the accompanying drawings and embodiments.

When the embodiment of the present invention is applied to a storage system, the storage system includes two storage arrays and one host, and both storage arrays can accept data sent by the host. The two storage arrays are also connected to each other so that when one storage array receives host data, the other storage array backs up the host data received by the storage array that receives the host data. When the link between the storage array receiving host data and the host fails, the backup storage array can continue to receive data sent by the host instead of the storage array currently receiving host data. A split-brain occurs when the link between two storage arrays goes down.

It should be noted that, in this embodiment of the present invention, the host may be a cluster, the cluster includes multiple devices, and any device in the cluster can send data to any storage array. For example, the host includes a first device and a second device, and connected to the host includes a first storage array and a second storage array. When the first storage array receives data sent by the host, the data is sent by the first device. When the second storage array receives data sent by the host, the data is sent by the second device.

Below, a more intuitive explanation will be given to the above-mentioned device of the present invention with reference to FIG. 1 . FIG. 1 is a connection diagram of a dual-active storage array according to an embodiment of the present invention. As shown in FIG. 1 , the storage array includes a host 104 , an arbitration server 103 , a first storage array 101 and a second storage array 102 . The first storage array 101 and the second storage array 102 respectively include a storage medium and a storage controller, and the storage medium is used for storing data sent by the host 104. The storage controller includes a controller and a storage unit, the storage unit is used to store the control instructions of the controller, and the controller is used to execute various operation instructions stored in the storage unit to complete various operations on the data.

The host 104 is configured to send an input/output (input/output, IO for short) request to the first storage array 101 or the second storage array 102.

The arbitration server 103 is configured to arbitrate the IO operation between the host 104 and the first storage array 101 or the IO operation between the host 104 and the second storage array 102.

The first storage array 101 and the second storage array 102 are respectively connected to the host 104 and can receive IO requests sent by the host 104. In the embodiment of the present invention, when the storage system starts to work, one of the first storage array and the second storage array is set as the main storage array, and the IO request sent by the host 104 is directly received. The other is set up as a backup storage array, which is used to back up the contents of the host's data received by the primary storage array. In another possible situation, one storage array is set as the main storage array, which mainly receives IO requests sent by the host 104. The other storage array is a secondary storage array, which is used to receive some IO requests sent by the host 104, and the primary storage array and the secondary storage array back up each other. The embodiments of the present invention are mainly discussed for the above-mentioned first use case, but the control methods and devices used in the embodiments of the present invention are also applicable to the second use case.

In one example, the first storage array 101 and the second storage array 102 are configured differently. One storage array has more processing power, and the other storage array has less processing power. When the active-active array is working, the storage array with stronger processing capability is the primary storage array, and the storage array with weaker processing capability is the secondary storage array.

An internal connection is included between the first storage array 101 and the second storage array 102, and the internal connection can obtain the connection status of the other storage array and be used for backing up the contents of the opposite storage array.

The arbitration server 103 is respectively connected to the first storage array 101 and the second storage array 102, so that each storage array can send the arbitration file to the arbitration server 103, and the arbitration server 103 sends the arbitration result to each storage array.

When the first storage array 101 and the second storage array 102 are connected to the host 104, they also count the total amount of received host data in real time. The total amount of received host data can also be compared between the first storage array 101 and the second storage array 102, and the data comparison method can be performed on any device, which is not limited in the present invention. For example, the first storage array 101 or the second storage array 102 sends the statistics of the total amount of received host data to the main storage array, and compares the total amount of received host data on the main storage array. The storage array also saves each comparison result separately.

In one example, the comparison of the total amount of received host data is performed on the first storage array, and the second storage array will send the counted total amount of received host data to the primary storage array. The first storage array will receive the total amount of host data that the second storage array will receive from the second storage array. For example, the first storage array is a storage array that initially receives host data, so the total amount of host data received by the first storage array is greater than the total amount of host data received by the second storage array.

Those skilled in the art know that the comparison of the total amount of received host data can also be performed in the second storage array, and the comparison result is sent to the first storage array. In addition, the two storage arrays can send and receive the statistical results of the total host data to each other, compare the two storage arrays respectively, and send the comparison results to each other.

In the embodiment of the present invention, the first storage array and the second storage array are compared at regular intervals. The first storage array and the second storage array update the comparison result.

A split brain occurs when the connection between the first storage array 101 and the second storage array 102 is disconnected. In the case of split-brain, each storage array cannot obtain the information that the other storage array receives host data. Therefore, the arbitration server needs to arbitrate the receiving host data of each storage array. The quorum server determines which storage array receives the host's data.

Each storage array sends an arbitration request to the arbitration server 103 for obtaining the arbitration result. Storage arrays that receive a large amount of host data submit an arbitration request to the arbitration server 103 in real time, and storage arrays that receive a small amount of host data submit an arbitration request to the arbitration server 103 after a preset period of time. Therefore, according to the aforementioned assumed total amount of received host data in this embodiment of the present invention, the first storage array 101 immediately sends an arbitration request to the arbitration server 103 during split-brain, and the second storage array 102 receives a predetermined amount of data after the split-brain. A request for arbitration is sent after the time period.

In one example, a storage array that receives a small amount of host data will send an arbitration request to the arbitration server within 3 seconds to 10 seconds after a split brain is detected. In a more specific example, a storage array with a small amount of data from the receiving host will send an arbitration request to the arbitration server with a delay of 5 seconds.

The arbitration server 103 will perform arbitration according to the arbitration request and the arbitration principle, and send the arbitration result to the first storage array and the second storage array. In the embodiment of the present invention, the arbitration principle may be first-come, first-served, that is, the arbitration server 103 determines which storage array wins the arbitration request sent by which storage array first receives the data sent by the host. Under normal requests, the storage array that sends the arbitration request in real time will win the arbitration, so that the storage array will continue to receive data sent by the host.

When a split-brain occurs in an active-active array, the storage system is configured to set the primary storage array to receive data sent by the host when it starts working. When the main storage array and the arbitration server 103 do not fail, the main storage array will win the arbitration, and the main storage array will continue to receive data sent by the host. The primary storage array wins the arbitration, so that the storage array that receives the host's data in split-brain wins the arbitration, and there is no need to switch the storage array of the receiving host. Therefore, the first storage array 101 will continue to be connected to the host, so as to achieve the effect that the storage array does not need to be replaced.

In the embodiment of the present invention, the total amount of host data received by the first storage array and the second storage array is compared above, and the storage array that receives more host data is given priority to continue to receive the data sent by the host, which is only the implementation of the present invention. One of the specific implementation methods in the example cannot be used to limit the present invention.

In an example, a direct arbitration method can also be used. When split brain occurs, one of the storage arrays that is receiving the data sent by the host continues to receive the data sent by the host, thereby avoiding switching the storage that receives the host data when split brain occurs. possible effects of the array.

FIG. 2 is a flowchart of a method for managing a dual-active storage array according to an embodiment of the present invention. The method is applied to a storage system, and the storage system includes a first storage array, a second storage array, an arbitration server and a host. The first storage array and the second storage array can receive data sent by the host. The first storage array is connected to the second storage array, so as to back up the data of the other storage array and obtain the state of the host data received by the other storage array and the total amount of the host data received. The first storage array and the second storage array are also respectively connected with the arbitration server, so that when the first storage array and the second storage array are disconnected, the arbitration server arbitrates to determine which storage array receives the data sent by the host. The execution body of the method is the first storage array. As shown in Figure 2, the method includes:

S201, the total amount of data sent by the receiving host is counted.

The first storage array counts the total amount of received host data in real time, and records the statistical results in the storage array.

S202: Receive the total amount of received host data counted by the second storage array.

The second storage array counts the total amount of data sent by the receiving host in real time, and records the statistical result in the storage array. A comparison module is set in one of the storage arrays, and the storage array where the comparison module is located will obtain the total amount of data sent from the host by the other storage array. In the embodiment of the present invention, the comparison module is arranged in the first storage array. The second storage array also sends the counted total amount of received host data to the first storage array.

In one example, the comparison module may be disposed in the second storage array, and the first storage array sends the total amount of the received host data to the second storage array. The comparison module can also be arranged in the first storage array and the second storage array at the same time, and the first storage array and the second storage array send each statistical total amount of the received host data. The first storage array and the second storage array send statistical results to each other.

S203. Determine that the total amount of data sent by the host received by the first storage array is smaller than that of the second storage array.

The comparison unit compares the total amount of host data received by each storage array, and sends the comparison results to the first storage array and the second storage array respectively. In one example, the first storage array and the second storage array also store the results of comparing the total amount of host data received by the first storage array and the second storage array, respectively, so as to determine the amount of host data received by their own storage arrays when necessary. Whether the total amount is greater than that of the other storage array.

In one example, the first storage array and the second storage array store the comparison result of the total amount of received host data by marking. According to the comparison results, compared to another storage array, the storage array that receives more host data will be marked as a priority site, and the storage array that receives less host data will be marked as a non-priority site. Determine the amount of data received by the own storage array and the other storage array through the priority site and the non-priority site.

In the embodiment of the present invention, the comparison of the total amount of the received host data is continuous, and the duration may be a fixed time to perform a comparison. The storage array will be marked based on the results of each comparison.

In the embodiment of the present invention, it is assumed that the comparison unit determines that the first storage array receives less data from the host, and the second storage array receives more data from the host.

S204, when the first storage array and the second storage array are split-brained, the first storage array sends an arbitration request to the arbitration server after a preset period of time.

When the link transmission between the first storage array and the second storage array fails, the first storage array and the second storage array are split-brained. At this time, the first storage array and the second storage array need to send arbitration requests to the arbitration server respectively. Because the total amount of host data received by the first storage array is small, and the total amount of host data received by the second storage array is large, the second storage array will send an arbitration request to the arbitration server in real time, and the first storage array will delay Send an arbitration request to the arbitration server for a preset period of time. In the embodiment of the present invention, the preset delay time may be any length from 3 seconds to 10 seconds.

In one example, the preset delay time is 5 seconds.

S205: Receive the arbitration result sent by the arbitration server.

The arbitration server will make the arbitration result according to the arbitration request and arbitration principle. Preferably, the arbitration principle may be first-come, first-served, that is, the arbitration server first receives the arbitration request sent by which storage array, and determines which storage array wins.

Since the first storage array delays sending the arbitration request to the arbitration server for a preset time, when the link between the second storage array and the arbitration server does not fail, the arbitration server will first receive the arbitration request sent by the second storage array . The arbitration result received by the first storage array indicates that the second storage array continues to receive the data sent by the host. According to the arbitration result, the first storage array will not receive the data sent by the host, and the second storage array will receive the data sent by the host.

Fig. 3 is a system connection diagram after a failure of a dual-active storage array arbitration server according to an embodiment of the present invention. Fig. 4 is a system connection diagram after disconnection of the arbitration server of a dual-active storage array provided by the implementation of the present invention. As shown in FIG. 3 and FIG. 4 , it includes a host 104 , an arbitration server 103 , a first storage array 101 and a second storage array 102 . The arbiter servers are respectively connected with the first storage array and the second storage array. Both the first storage array and the second storage array are connected to at least one host, and both the first storage array and the second storage array are capable of receiving data from the host.

Each of the first storage array 101 and the second storage array 102 counts the total amount of data sent by the receiving host in real time. One of the first storage array 101 or the second storage array 102 may also compare the total amount of data sent by the receiving host, for example, the first storage array 101 may compare the total amount of data sent by the receiving host. The second storage array 102 will send the real-time statistics of the total amount of received host data to the first storage array 101, so as to determine the amount of data received by the first storage array and the second storage array by comparison. In this embodiment, it is assumed that the first storage array 101 receives a large amount of data from the host, and the second storage array 102 receives a small amount of data sent from the host.

During the working process of the storage array, it will also communicate with the arbitration server 103 to ensure that the connection between the storage array and the arbitration server 103 is normal, so that the arbitration server 103 can accept the arbitration request and send the arbitration result to the storage array when arbitration is required. .

If one of the storage arrays, for example, after the first storage array 101 detects that it is disconnected from the arbitration server 103, the first storage array 101 will determine which subject the current active-active array uses to determine the storage array that receives host data.

If the storage array currently receiving the host data is determined by the arbitration server, the first storage array 101 will send a message to the second storage array 102, so that the second storage array 102 can detect (for example, using a heartbeat mechanism to detect) whether the second storage array is compatible with Arbitration server 103 is disconnected. If the second storage array 102 is disconnected from the arbitration server 103 (including the situations shown in 3 and 4), the second storage array 102 will send the query result to the first storage array 101. In the embodiment of the present invention, the query result returned by the second storage array to the first storage array determines that the link between the second storage array and the arbitration server is faulty. When the first storage array determines that both the first storage array 101 and the second storage array 102 are disconnected from the arbitration server 103, the first storage array 101 will negotiate with the second storage array 102. Change the storage array that receives host data from the arbiter server to the storage array. The negotiation result will be sent to the first storage array 101 and the second storage array 102 respectively, and the first storage array 101 and the second storage array 102 will switch according to the negotiation result.

If the currently received host data is determined by the storage array, it will be determined whether the total amount of host data received by the first storage array 101 is the largest. If the first storage array 101 receives the largest amount of host data, the first storage array receives the host data. If the first storage array 101 is not the storage array that receives the most host data, it does not receive host data. In this embodiment, since it is assumed in advance that the first storage array 101 receives the most data from the host, the first storage array 101 will receive the data sent by the host.

If the connection between the second storage array 102 and the arbitration server 103 is normal, no operation will occur (the connection structure shown in FIG. 1 ), and the storage array receiving the host data will continue to receive the data sent by the host.

If a split brain occurs between the first storage array and the second storage array, the storage array that is normally connected to the arbitration server 103 will send the arbitration file to the arbitration server 103. The storage array that is disconnected from the arbitration server 103 will not be able to send the arbitration request to the arbitration server 103, so the storage array that is normally connected to the arbitration server 103 will receive the data sent by the host.

The foregoing negotiation through the first storage array 101, etc., is only a specific implementation manner of the embodiment of the present invention, and is not used to limit the embodiment of the present invention. In the above-mentioned embodiment of the present invention, it can also be set to negotiate through the second storage array 102.

In order to make the description of the disconnection between the active-active storage array and the arbitration server clearer, FIG. 5 is a flowchart of a method for managing the active-active storage array after the connection between the active-active storage array and the arbitration server is interrupted, provided by this embodiment of the present invention. The method is applied to a storage system, including a first storage array, a second storage array, an arbitration server and a host. The first storage array and the second storage array are capable of receiving data sent by the host. The first storage array and the second storage array are connected to each other, so as to back up the data of the other storage array and obtain the status of the data received from the host by the other storage array. The first storage array and the second storage array are also respectively connected to the arbitration server, so that when the first storage array and the second storage array are disconnected, the arbitration server arbitrates to determine which storage array receives the data of the host. The execution body of the method is the first storage array. As shown in Figure 5, the shown method specifically includes:

S501: Count the total amount of data sent by the host received by the first storage array.

The first storage array counts the total amount of received host data in real time, and records the statistical results in the storage array.

S502: Receive the total amount of received host data counted by the second storage array.

The second storage array counts the total amount of received host data in real time, and records the statistical results in the storage array. A comparison module is set in one of the storage arrays, and the comparison module will obtain the total amount of host data received by the storage array. In one example, the comparison module is provided in the first storage array. The second storage array also sends the total amount of received host data to the first storage array.

S503: Determine that the total amount of data sent by the host received by the first storage array is greater than the total amount of data sent by the host received by the second storage array.

The comparison unit compares the total amount of host data received by each storage array, and sends the comparison results to the first storage array and the second storage array respectively. In one example, the first storage array and the second storage array also store the results of comparing the total amount of host data received by the first storage array and the second storage array, respectively, so as to determine the amount of host data received by their own storage arrays when necessary. Whether the total amount is more than the other storage array.

In the embodiment of the present invention, it is assumed that the comparison unit determines that the first storage array receives a large amount of data from the host, and the second storage array receives a small amount of data from the host.

S504, when the communication between the first storage array and the second storage array and the arbitration server fails, and the link between the first storage array and the second storage array fails, continue to receive the data sent by the host.

After the first storage array detects that it is disconnected from the arbiter server, the first storage array will determine who determines the storage array that receives the host data in the current active-active array.

If the storage array currently receiving host data is determined by the arbitration server, a verification message is sent to the second storage array, so that the second storage array detects (for example, using a heartbeat mechanism) whether the connection with the arbitration server is disconnected. If the second storage array returns to the first storage array that the connection with the arbiter server is normal, no operation is performed. A storage array that is currently receiving data from the host will continue to receive data from the host.

At this time, if the link between the first storage array and the second storage array in the storage array is disconnected, the storage array that is normally connected to the quorum server will send an arbitration request to the quorum server, while the storage array that is disconnected from the quorum server will send an arbitration request to the quorum server. The arbitration request cannot be sent to the arbitration server, and the arbitration server will issue the arbitration result according to the received arbitration request. The quorum result indicates that the storage array that is properly connected to the quorum server will receive the data sent by the host.

If the first storage array receives the message that the second storage array is disconnected from the arbitration server, the first storage array will negotiate with the second storage array, and the storage array determined to receive the data sent by the host will be changed by the arbitration server to the storage array determined by the storage array. Sure. The negotiation result will be sent to the first storage array 101 and the second storage array 102 respectively, and the first storage array 101 and the second storage array 102 will switch according to the negotiation result.

If the currently received host data is determined by the storage array, it will be determined whether the total amount of host data received by the first storage array 101 is the largest. If the first storage array 101 receives the largest amount of host data, the first storage array receives the host data. If the first storage array 101 is not the storage array that receives the most host data, it does not receive host data. In this embodiment, since it is assumed in advance that the first storage array 101 receives the most data from the host, the first storage array 101 will receive the data sent by the host.

In the following, the management of the active-active storage array according to the embodiment of the present invention will be described in detail through a specific example. FIG. 6 is a flowchart of managing a dual-active storage array according to an embodiment of the present invention. As shown in Figure 6, the method process takes the main storage array in the storage array as the execution body. The process of the embodiment of the present invention is based on the hardware structure shown in FIG. 1 as a whole, including a first storage array, a second storage array, an arbitration server, and a host, and the functions of each part are the same as those shown in FIG. 1 . . Here, the main storage array is defined as the first storage array (it may also be defined as the second storage array, which is not limited here), and the main storage array is also used to perform the first storage array and the second storage array for receiving host data. Compare and save the comparison results. Under normal circumstances, when the active-active array starts working, the primary storage array is set to receive the data sent by the host, and the other storage array is used for backup.

When the storage system starts to work, step 601 is executed.

The main storage array will periodically count the total amount of received host data.

The main storage array will cyclically count and update the total amount of data received from the host. After the main storage array has counted the total amount of data received from the host, step 602 will be executed.

The first storage array 101 receives the total amount of received host data counted by the second storage array 102, and the first storage array compares the counted data of the first storage array and the second storage array with the total amount of data sent by the host.

In a specific embodiment of the present invention, since the first storage array is set to take over the storage array when the business first starts to work, the total amount of host data received by the first storage array is greater than that of the second storage array.

During the working process of the storage array, it will also communicate with the arbitration server 103 to ensure that the connection between the storage array and the arbitration server is normal, so that the arbitration server can accept the arbitration request and send the arbitration result to the storage array when the arbitration server is required to arbitrate.

When the main storage array detects that the connection with the arbitration server is normal, step 613 is executed. If the primary storage array detects that the connection with the arbiter server is disconnected, step 604 is performed.

During the working process of the main storage array, communication between the first storage array and the second storage array will be maintained at any time to determine whether the connection between the first storage array and the second storage array is disconnected.

If it is detected that the connection between the first storage array and the second storage array is normal, that is, when there is no split-brain, no split-brain occurs, then step 663 is performed, and no other operations are performed.

If a split brain occurs between the first storage array and the second storage array, step 623 is performed.

Determine whether the current storage array receives the most host data in total, and if the current storage array receives the most host data, perform step 643.

Immediately send an arbitration request to the arbitration server.

If the current storage array does not receive the most data sent by the host, perform step 633.

Delays sending an arbitration request to the arbitration server for a preset period of time.

After step 633 or 643 is performed, step 653 is performed to obtain the arbitration result sent by the arbitration server, and determine the storage array that receives the data sent by the host according to the arbitration result.

In the embodiment of the present invention, the arbitration principle of the arbitration server is first-come, first-served, so the storage array that receives the most data sent by the host wins the arbitration and continues to receive the data sent by the host.

In executing step 604, it will be determined who currently determines the storage array of the storage array that receives the data sent by the host. The first storage array will send a confirmation message to the second storage array, and the second storage array will, according to the confirmation message, inquire about the storage array to which the receiving host of the second storage array sends data. The second storage array returns the verification result to the first storage array after the verification is completed. If the storage array to receive the host data is determined by the storage array, then step 614 is performed. If it is determined by the arbitration server that the current storage array receives the data sent by the host, then step 605 is performed.

In the process of executing step 614, it will be determined whether the current storage array is the storage array that receives the most host data. If the current storage array is the storage array that receives the most host data, go to step 624 . If the current storage array is not the storage array that receives the most host data, step 634 is performed.

In performing step 624, the current storage array will receive the data sent by the host.

In step 634, the current storage array will not receive the data sent by the host, and the current storage array will be used to back up the storage array that receives the data sent by the host, so that when the storage array receiving the host data fails, seamless switching can be performed to the currently backed-up storage array.

During the execution of step 605, the first storage array will send a verification message to the second storage array to determine whether the second storage array is currently connected to the arbitration server normally. If the connection between the second storage array and the arbitration server is normal, step 615 is performed. If the second storage array is disconnected from the arbitration server, step 606 is performed.

In executing step 615, it is first determined whether the first storage array is disconnected from the second storage array.

If the connection between the first storage array and the second storage array is normal, step 645 is performed. No other operations are performed.

If the first storage array is disconnected from the second storage array, step 625 is performed.

The storage array connected to the quorum server will send the quorum request to the quorum server. The arbitration server conducts arbitration according to the received arbitration request. Since the first storage array is disconnected from the arbitration server, the arbitration server can only receive the arbitration request sent by the second storage array at this time.

After the arbitration request is sent, step 635 is performed.

Obtain the arbitration result, and determine the storage array that receives the host data according to the arbitration result. Since the arbitration server can only receive the arbitration request sent by the second storage array, the second storage array will obtain the data sent by the receiving host.

Before executing step 606, the second storage array is disconnected from the arbitration server, and the result is returned to the first storage array.

Therefore, in step 606, the first storage array will negotiate with the second storage array to change the storage array determined by the arbitration server to receive host data to the storage array determined by the storage array to receive host data according to the returned result. The first storage array and the second storage array will be replaced according to the negotiation result.

After the storage array is switched to the storage array determined by the storage array to receive the data of the host, step 607 will be executed.

If the current storage array receives the most host data, step 609 is performed. If the current storage array does not receive the most host data, step 608 is performed.

In performing step 624, the current storage array will receive the data sent by the host.

In step 634, the current storage array will not receive the data sent by the host, and the current storage array will be used to back up the storage array that receives the data from the host, and can seamlessly switch to the current backup when the storage array that receives the data sent by the host fails storage array.

In a specific embodiment of the present invention, a structure diagram of a storage array in an active-active array is also included, and the storage array may be 101 or 102 shown in FIG. 1 . The following takes the storage array 101 as an example for detailed description. FIG. 7 is a structural diagram of a dual-active array storage array provided by an embodiment of the present invention. As shown in FIG. 7 , the storage array includes a counting unit 711 , a processing unit 712 , and a receiving unit 713 .

A counting unit 711, configured to count the total amount of data sent by the first storage array and received by the host; and configured to receive data sent by the second storage array and received by the second storage array from the host total amount.

The processing unit 712 is configured to determine that the total amount of data received by the first storage array and sent by the host is less than the total amount of data received by the second storage array and sent by the host; When a link between the second storage arrays fails, the arbitration request is sent to an arbitration server after a preset time period, and the arbitration server is respectively connected to the first storage array and the second storage array.

The receiving unit 713 is configured to receive an arbitration result sent by the arbitration server, where the arbitration result is used to instruct the second storage array to continue to accept the data sent by the host.

The statistical unit in the embodiment of the present invention may be executed through S201 and S202 in the method steps, the processing unit may be executed through S202 and S203 in the method steps, and the receiving unit may be executed through S205 in the method steps.

In the embodiment of the present invention, it also includes a structure diagram of the storage array in a storage array in which a dual-active array arbitration unit failure is shown in FIG. 1 . The first memory array 101 is taken as an example for description below. Fig. 8 is a structural diagram of an active-active storage array for managing an arbitration server after a failure according to an embodiment of the present invention, as shown in Fig. 8: the storage array shown is used to issue a query command. It includes a statistical unit 811 and a processing unit 812 .

A statistical unit 811, configured to count the total amount of data sent by the host received by the first storage array. And, receive the total amount of data sent by the second storage array and received by the host from the second storage array.

The processing unit 813 is configured to determine that the total amount of data sent by the host received by the first storage array is greater than the total amount of data sent by the host received by the second storage array. and, when the communication between the first storage array and the arbiter server fails and the communication between the second storage array and the arbiter server fails, when the first storage array and the second storage array fail When the link between the storage arrays fails, the data sent by the host continues to be received.

The statistical unit in the embodiment of the present invention may be executed by S501 in the method step, and the processing unit may be executed by S502 in the method step.

FIG. 9 is a schematic diagram of a storage array device according to an embodiment of the present invention.

As shown in FIG. 9 , the device includes: a processor 901, a memory 902, a communication interface 903 and a bus 904. The processor 901, the memory 902 and the communication interface 903 are connected to each other through the bus 904 for communication.

The processor 901 may use a general-purpose central processing unit (Central Processing Unit, CPU), a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, for executing relevant programs to The technical solutions provided by the foregoing method embodiments of the present invention are implemented.

The memory 902 may be a read only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). Memory 902 may store operating systems and other applications. When the technical solutions provided by the embodiments of the present invention are implemented through software or firmware, the program code for implementing any optional technical solutions provided by the foregoing method embodiments of the present invention is stored in the memory 902 and executed by the processor 901 . The communication interface 903 is used to communicate with other switches or control servers.

The bus 904 may include a path to communicate information between the various components of the device (e.g., the processor 901, the memory 902, and the communication interface 903).

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

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be implemented in hardware, a software module executed by a processor, or a combination of the two. A software module can be placed in random access memory (RAM), internal memory, read only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other device in the art. in any other known form of storage medium.

The specific embodiments described above further describe the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above descriptions are only specific embodiments of the present invention, and are not intended to limit the scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims (8)

1. A method for managing an active-active storage array, the method being applied to a storage system, the storage system comprising a first storage array and a second storage array, the first storage array and the second storage array being at least A host is connected, and both the first storage array and the second storage array can receive data sent by the host, the method is executed by the first storage array, and it is characterized in that, it includes: Counting the total amount of data sent by the host received by the first storage array; Receive the total amount of data sent by the second storage array and received by the host from the second storage array; determining that the total amount of data sent by the host received by the first storage array is less than the total amount of data sent by the host received by the second storage array; When it is determined that the link between the first storage array and the second storage array is faulty, delaying a preset period of time to send an arbitration request to the arbitration server, and the arbitration server communicates with the first storage array and the second storage array respectively. connecting the second storage array; receiving an arbitration result sent by the arbitration server, where the arbitration result is used to indicate that the second storage array continues to accept data sent by the host. 2 . The method according to claim 1 , wherein the preset time period is any duration from 3 seconds to 10 seconds. 3 . 3. A method for managing an active-active storage array, the method being applied to a storage system, the storage system comprising a first storage array, a second storage array, and an arbitration server, the arbitration server being connected to the first storage array and the arbitration server respectively. A second storage array is connected; both the first storage array and the second storage array are connected to at least one host, and both the first storage array and the second storage array can receive data sent by the host, and the method comprises: The first storage array execution is characterized in that it includes: Counting the total amount of data sent by the host received by the first storage array; Receive the total amount of data sent by the second storage array and received by the host from the second storage array; determining that the total amount of data sent by the host received by the first storage array is greater than the total amount of data sent by the host received by the second storage array; When the communication between the first storage array and the arbiter server fails, and the communication between the second storage array and the arbiter server fails, when the first storage array and the second storage array fail When the link between the arrays fails, the data sent by the host continues to be received. 4. A first storage array in an active-active storage array, the first storage array is applied in a storage system, the storage system further includes a second storage array and at least one host, the arbitration server is respectively connected with the first storage system. The storage array and the second storage array are connected; the first storage array and the second storage array are both connected to at least one host, and both the first storage array and the second storage array can receive data sent by the host, and It is characterized in that, the first storage array further includes: a processor, a memory, a communication interface and a bus, wherein the processor, the memory and the communication interface communicate through the bus; the communication interface is used to communicate with the second storage array and the host; the memory is used to store programs; When the active-active storage array device is running, the processor is configured to execute the program stored in the memory, so as to execute the method of any one of claims 1 or 2. 5. A first storage array in an active-active storage array, the first storage array is applied in a storage system, the storage system further includes a second storage array and at least one host, the arbitration server is respectively connected with the first storage system. The storage array and the second storage array are connected; the first storage array and the second storage array are both connected to at least one host, and both the first storage array and the second storage array can receive data sent by the host, and It is characterized in that, the first storage array further includes: a processor, a memory, a communication interface and a bus, wherein the processor, the memory and the communication interface communicate through the bus; the communication interface is used to communicate with the second storage array and the host; the memory is used to store programs; When the active-active storage array device is running, the processor is configured to execute the program stored in the memory to execute the method of claim 3 . 6. An active-active storage array, the storage array is applied in a storage system, the storage system comprises a first storage array, a second storage array and an arbitration server, the arbitration server is respectively connected with the first storage array and the second storage array. Storage array connection; both the first storage array and the second storage array are connected to at least one host, and both the first storage array and the second storage array can receive data sent by the host, and a method for managing active-active storage arrays Performed by the first storage array, wherein the storage array includes: A statistical unit, configured to count the total amount of data sent by the host received by the first storage array; and configured to receive data sent by the second storage array and received by the host from the second storage array total; a processing unit, configured to determine that the total amount of data received by the first storage array and sent by the host is smaller than the total amount of data received by the second storage array and sent by the host; and, configured to determine that the first storage array When the link between the array and the second storage array fails, delay a preset period of time to send an arbitration request to the arbitration server, and the arbitration server communicates with the first storage array and the second storage array respectively. connect; A receiving unit, configured to receive an arbitration result sent by the arbitration server, where the arbitration result is used to instruct the second storage array to continue to receive data sent by the host. 7 . The active-active storage array according to claim 6 , wherein the processing unit delays a preset period of time for an arbitrary duration ranging from 3 seconds to 10 seconds. 8 . 8. An active-active storage array, the storage array is applied in a storage system, the storage system comprises a first storage array, a second storage array and an arbitration server, the arbitration server is respectively connected with the first storage array and the second storage array. Storage array connection; both the first storage array and the second storage array are connected to at least one host, and both the first storage array and the second storage array can receive data sent by the host, and a method for managing active-active storage arrays Performed by the first storage array, wherein the storage array includes: A counting unit, configured to count the total amount of data sent by the first storage array and received by the host; and configured to receive the total amount of data sent by the second storage array and received by the host from the second storage array quantity; a processing unit, configured to determine that the total amount of data received by the first storage array and sent by the host is greater than the total amount of data received by the second storage array and sent by the host; and, configured to determine that the first storage array When the communication between the array and the arbiter server fails, and the communication between the second storage array and the arbiter server fails, when the link between the first storage array and the second storage array fails When a failure occurs, continue to receive data sent by the host.

Images