CN106375427A - A distributed SAN storage system link redundancy optimization method - Google Patents

Abstract

The invention discloses a link redundancy optimization method for a distributed SAN storage system. The method comprises the steps of configuring an SAN service on each storage node; adding high-available configuration to each storage node; randomly allocating multiple connection requests to the storage nodes through utilization of the high-available configuration when the multiple connection requests are received, thereby finishing the butt joint between a client and the storage nodes; and providing a new connection link for the client through adoption of the high-available configuration when a fault occurs at a storage node. According to the method, the high-available configuration is added to original multipath link redundancy, the reliability of the SAN is ensured, network loads of the storage nodes are balanced, and a bandwidth utilization rate of a storage cluster is improved.

Description

A distributed SAN storage system link redundancy optimization method

technical field

The invention relates to the technical field of storage systems, in particular to a method for optimizing link redundancy of a distributed SAN storage system.

Background technique

Currently, existing distributed storage systems have high requirements for reliability and throughput. When the storage system has a large number of storage nodes, the probability of failure of the storage nodes will be greatly increased. After the failure of the storage nodes, it is very important to ensure that the client can continue to read and write data to the storage system. When a large number of client hosts access the storage system, the storage system needs to provide sufficient bandwidth for the client hosts to make full use of the bandwidth of each storage node, so as to improve the bandwidth utilization of the entire storage system is very important. SAN (Storage Area Network, storage area network) uses mesh channel technology to connect storage arrays and server hosts through FC switches to establish an area network dedicated to data storage.

The general SAN storage system design scheme is to use a few nodes as the service provider, and configure the multipath strategy on the client. However, in the large-scale distributed SAN storage system scenario, the link redundancy provided by the traditional multipath strategy is limited. There are few redundant links, and two service nodes are not enough to support the bandwidth of the entire distributed storage cluster, and the bandwidth usage is unbalanced, so the link reliability of the distributed SAN storage system is low, and the bandwidth utilization rate of the storage cluster is low.

Contents of the invention

The purpose of the present invention is to provide a distributed SAN storage system link redundancy optimization method, so as to improve the link reliability of the distributed SAN storage system and improve the bandwidth utilization rate of the storage cluster.

In order to solve the above technical problems, the present invention provides a distributed SAN storage system link redundancy optimization method, including:

Configure SAN services on each storage node;

Add high availability configuration to each storage node;

When multiple connection requests are received, the high-availability configuration is used to randomly distribute the multiple connection requests to each storage node to complete the connection between the client and the storage node;

When a storage node fails, a high-availability configuration is used to provide a new connection link for the client.

Preferably, the SAN service is used for processing client connection requests; the high availability configuration includes a high availability mechanism and a high availability allocation strategy.

Preferably, after configuring the SAN service on each storage node, the method further includes: initializing each storage node to the same state.

Preferably, said configuring SAN services on each storage node includes:

Set up a user-oriented SAN storage interface on each storage node.

Preferably, when multiple connection requests are received, the high availability configuration is used to randomly assign the multiple connection requests to each storage node, including:

When multiple clients access the storage system and receive connection requests from multiple clients, use the high-availability configuration to evenly distribute the connection requests of multiple clients to the SAN storage interfaces of each storage node according to the high-availability allocation strategy.

Preferably, after adding the high-availability configuration to each storage node, the method further includes: configuring a multi-path policy on the client to assign storage links to different virtual IP addresses.

Preferably, the high-availability configuration is used to provide a new connection link for the client, including:

The high-availability configuration is used to find a new storage node to replace the failed storage node, and a connection link between the client and the new storage node is established.

A distributed SAN storage system link redundancy optimization method provided by the present invention configures SAN services on each storage node; adds high availability configuration to each storage node; when receiving multiple connection requests, utilizes high The available configuration randomly distributes the multiple connection requests to each storage node to complete the connection between the client and the storage node; when a storage node fails, the high-availability configuration is used to provide a new connection link for the client. It can be seen that when a large number of connection requests arrive, the high-availability configuration will randomly allocate multiple connection requests to each storage node, complete the connection between the client and the storage node, provide services for the client host, and realize the load of each storage node at the bandwidth level Balanced to improve link reliability of the distributed SAN storage system. When a storage node fails, the high-availability configuration provides a new connection path for the client to achieve a more reliable redundancy mechanism than the traditional multi-path. The high-availability configuration is added to the multi-path link redundancy, which can balance the network load of each storage node while ensuring the reliability of the SAN, and improve the bandwidth utilization of the storage cluster.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only It is an embodiment of the present invention, and those skilled in the art can also obtain other drawings according to the provided drawings without creative work.

Fig. 1 is the flowchart of a kind of distributed SAN storage system link redundancy optimization method provided by the present invention;

FIG. 2 is a schematic diagram of a storage system link redundancy structure.

detailed description

The core of the present invention is to provide a distributed SAN storage system link redundancy optimization method, so as to improve the link reliability of the distributed SAN storage system and improve the bandwidth utilization rate of the storage cluster.

In order to enable those skilled in the art to better understand the solutions of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Please refer to Fig. 1, Fig. 1 is the flowchart of a kind of distributed SAN storage system link redundancy optimization method provided by the present invention, and this method comprises:

S11: configure SAN service on each storage node;

S12: Add high availability configuration to each storage node;

S13: When multiple connection requests are received, use high-availability configuration to randomly assign multiple connection requests to each storage node to complete the connection between the client and the storage node;

S14: When a storage node fails, adopt a high-availability configuration to provide a new connection link for the client.

It can be seen that when a large number of connection requests arrive, the high-availability configuration will randomly allocate multiple connection requests to each storage node, complete the connection between the client and the storage node, provide services for the client host, and realize the load of each storage node at the bandwidth level Balanced to improve link reliability of the distributed SAN storage system. When a storage node fails, the high-availability configuration provides a new connection path for the client to achieve a more reliable redundancy mechanism than the traditional multi-path. The high-availability configuration is added to the multi-path link redundancy, which can balance the network load of each storage node while ensuring the reliability of the SAN, and improve the bandwidth utilization of the storage cluster.

Based on the above method, specifically, the SAN service is used to process the connection request of the client. The high-availability configuration includes a high-availability mechanism and a high-availability allocation strategy, that is, when receiving multiple connection requests, use the high-availability mechanism and high-availability strategy to randomly distribute multiple connection requests to each storage node to complete the connection between the client and the storage node. butt. The high-availability allocation strategy can balance the bandwidth load of each storage node, which is convenient for the unified configuration and management of SAN services of each storage node.

Further, after step S11, the method further includes: initializing each storage node to the same state.

Wherein, the process of step S11 is specifically: setting a SAN storage interface facing the user layer in each storage node.

In step S13, when multiple connection requests are received, the process of randomly assigning multiple connection requests to each storage node using the high-availability configuration is specifically: when multiple clients access the storage system, receiving multiple client requests For connection requests, use the high availability configuration to evenly distribute the connection requests of multiple clients to the SAN storage interfaces of each storage node according to the high availability allocation strategy.

Wherein, after step S12, it further includes: configuring a multipath policy on the client, and assigning storage links to different virtual IP addresses.

In step S14, the process of using the high-availability configuration to provide a new connection link for the client is specifically: use the high-availability configuration to find a new storage node to replace the failed storage node, and establish a connection between the client and the new storage node link.

Specifically, in a large-scale distributed storage system, each storage node provides a user-oriented SAN storage interface. When a large number of external clients access the storage system, the high-availability configuration is used to send the client's access connection to the The SAN storage interface of each storage node, when the SAN service of a certain storage node fails, the high availability mechanism provides a new connection path for the client, so as to achieve a more reliable redundancy mechanism than the traditional multipath.

When a large number of connection requests arrive, the high-availability mechanism will first eliminate faulty nodes according to a predetermined allocation plan, and then preferentially select relatively idle storage nodes to provide services for clients, so as to achieve load balancing of each storage node at the bandwidth level. This method solves the shortcomings of the original SAN deployment scheme, and adds a layer of high-availability strategy after the original multi-path link redundancy, which can balance the network load of each storage node and improve the storage cluster while ensuring the reliability of the SAN. bandwidth utilization.

This method mainly aims at the problems of less redundant links and unbalanced bandwidth usage in traditional SAN storage schemes, so as to improve the reliability of distributed SAN storage links and optimize the load balance of each storage node at the bandwidth level. Specifically, deploy SAN services on each node of the cluster and initialize them to the same state; configure multiple sets of high-availability solutions for the storage system as required to provide SAN services to clients; when the high-availability configuration receives multiple connections When requesting, the connection request is randomly assigned to each storage node according to a certain strategy, that is, the high-availability allocation strategy, and the connection between the client and the storage node SAN storage node is completed; when a storage node fails, the high-availability configuration will be the new The client requests to provide a fault-free path connection, and gives priority to relatively idle storage nodes; at the same time, for the connected service client, on the one hand, multi-path links are enabled to provide services, and on the other hand, the faulty path is replaced by a high-availability strategy is a valid path.

The link redundancy structure of the storage system is shown in Figure 2. First, configure SAN services on each storage node to process client requests; then, configure high availability on the storage nodes. When the client accesses the highly available virtual IP At the same time, the virtual IP evenly distributes the client connection to each storage node through the equal distribution strategy; finally, configure the multipath strategy on the client to assign the storage links to different virtual IPs, and finally realize the redundancy between the client and the storage node. Configuration of remaining links. The equal share strategy is a high-availability allocation strategy.

Taking the scenario in Figure 2 as an example, when client 1 needs SAN storage services, first configure the multipath policy, specifying two virtual IPs, namely high-availability virtual IP 1 and high-availability virtual IP 2, and then assign them to high-availability The strategy allocates storage nodes for clients. This allocation strategy will first eliminate faulty nodes, and then prioritize relatively idle storage nodes to provide services for client hosts, such as storage node 1 and storage node 3. In the process of providing services, when storage node 1 fails, due to the existence of multipath, storage node 3 will provide services, and at the same time, the high-availability strategy will find new storage nodes to replace the failed nodes. Compared with the traditional SAN link redundancy scheme, the present invention has a more reliable link redundancy strategy, uses a high-availability allocation strategy to balance the bandwidth load of each storage node, and is convenient for unified configuration and management of SAN services of each storage node. The client in this article is the client host.

In summary, a distributed SAN storage system link redundancy optimization method provided by the present invention configures SAN services on each storage node; adds high-availability configuration to each storage node; when receiving multiple connection requests , use the high availability configuration to randomly assign multiple connection requests to each storage node to complete the connection between the client and the storage node; when a storage node fails, use the high availability configuration to provide a new connection link for the client. It can be seen that when a large number of connection requests arrive, the high-availability configuration will randomly allocate multiple connection requests to each storage node, complete the connection between the client and the storage node, provide services for the client host, and realize the load of each storage node at the bandwidth level Balanced to improve link reliability of the distributed SAN storage system. When a storage node fails, the high-availability configuration provides a new connection path for the client to achieve a more reliable redundancy mechanism than the traditional multi-path. The high-availability configuration is added to the multi-path link redundancy, which can balance the network load of each storage node while ensuring the reliability of the SAN, and improve the bandwidth utilization of the storage cluster.

A method for optimizing link redundancy of a distributed SAN storage system provided by the present invention has been introduced in detail above. In this paper, specific examples are used to illustrate the principle and implementation of the present invention, and the descriptions of the above embodiments are only used to help understand the method and core idea of the present invention. It should be pointed out that for those skilled in the art, without departing from the principles of the present invention, some improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims (7)

1. A distributed SAN storage system link redundancy optimization method, is characterized in that, comprising: Configure SAN services on each storage node; Add high availability configuration to each storage node; When multiple connection requests are received, the high-availability configuration is used to randomly distribute the multiple connection requests to each storage node to complete the connection between the client and the storage node; When a storage node fails, a high-availability configuration is used to provide a new connection link for the client. 2. The method according to claim 1, wherein the SAN service is used to process the connection request of the client; the high availability configuration includes a high availability mechanism and a high availability allocation strategy. 3. The method according to claim 1, further comprising: after configuring the SAN service on each storage node, initializing each storage node to a same state. 4. The method according to claim 1, wherein the configuring the SAN service on each storage node comprises: Set up a user-oriented SAN storage interface on each storage node. 5. The method according to claim 4, wherein when multiple connection requests are received, randomly assigning the multiple connection requests to each storage node using a high-availability configuration comprises: When multiple clients access the storage system and receive connection requests from multiple clients, use the high-availability configuration to evenly distribute the connection requests of multiple clients to the SAN storage interfaces of each storage node according to the high-availability allocation strategy. 6. The method according to claim 1, characterized in that, after adding the high availability configuration to each storage node, further comprising: configuring a multipath policy on the client side, and assigning storage links to different virtual IP addresses . 7. The method according to any one of claims 1 to 6, wherein the use of a high-availability configuration to provide a new connection link for the client comprises: The high-availability configuration is used to find a new storage node to replace the failed storage node, and a connection link between the client and the new storage node is established.