CN103997537A - Method and device for remote operation of multiple nodes - Google Patents

Abstract

The invention provides a multi-node remote operation method and device. It involves computing applications and cluster management; it solves the problem of inefficiency in multi-node remote operations. The method includes: obtaining information on nodes requiring remote operation; when there are multiple nodes requiring remote operation, performing remote operations on the nodes in batches, and performing remote operations on at least two nodes in each batch. The technical solution provided by the invention is suitable for large-scale clusters, and realizes batch and efficient remote operation of nodes.

Description

Multi-node remote operation method and device

technical field

The invention relates to the field of server computing application and cluster management, in particular to a multi-node remote operation method and device.

Background technique

With the widespread use of clusters in the field of network services and high-performance computing, more and more content needs to be remotely operated in batches on each server in the cluster. At present, the scale of clusters used in various fields is also expanding year by year, including hundreds to over Clusters of thousands of servers have become necessary equipment for large Internet companies and supercomputing centers.

In the management of these large-scale clusters, it is inevitable to perform remote batch operations on each node server contained in the cluster. Generally, batch operations are performed using scripts. The general process of scripts is to remotely specify the nodes in the list sequentially through loops. server name, and do the same. Because this method executes remote operations sequentially, when the number of node servers in the cluster is hundreds, thousands or even tens of thousands, in this case, the time required for the sequential execution method will be several times that of a single server. Hundreds, thousands or even tens of thousands of times.

Therefore, it is very practical to provide a quick operation script tool.

Contents of the invention

The invention provides a multi-node remote operation method and device, which solves the problem of low efficiency of multi-node remote operation.

A multi-node remote operation method, comprising:

Obtain the node information that requires remote operation;

When the number of nodes requiring remote operation is multiple, remote operations are performed on the nodes in batches, and at least two nodes are remotely operated in each batch.

Preferably, obtaining node information requiring remote operation includes:

A preset node name list is obtained, and information of one or more nodes is stored in the node name list.

Preferably, when the number of nodes requiring remote operation is multiple, performing remote operations on the nodes in batches, and performing remote operations on at least two nodes in each batch includes:

dividing the nodes into one or more batches, each batch comprising at least two nodes;

Start a unique corresponding thread for each node that needs to be remotely operated in this batch, and check whether each node can log in remotely;

When all the nodes in this batch can log in remotely, use the threads corresponding to each node to perform remote operations on the corresponding nodes;

After the remote operation of the node is completed, end the thread corresponding to the node;

After all the threads in this batch are finished, start the remote operation on the next batch of nodes.

Preferably, when the number of nodes requiring remote operation is multiple, the remote operation is performed on the nodes in batches, and before the step of performing remote operation on at least two nodes in each batch, it also includes:

Set the maximum number of threads that can be started to a value greater than 1 and less than or equal to the number of processor cores of the current node.

Preferably, the nodes are divided into one or more batches, each batch includes at least two nodes including:

When the number of nodes does not exceed the maximum number of threads that can be started, divide the nodes into a batch;

When the number of nodes exceeds the maximum number of threads that can be started, divide the nodes into more than two batches, and each batch includes more than two nodes but less than the maximum number of threads.

Preferably, for each node that needs to be remotely operated in this batch, a unique corresponding thread is started, and after the step of detecting whether each node can log in remotely, it also includes:

When there is a node that cannot be logged in remotely, the remote operation to all nodes is stopped, and the information of the node that cannot be logged in is output.

The present invention also provides a multi-node remote operation device, including:

The node information obtaining module is used to obtain the node information requiring remote operation;

The batch processing module is configured to perform remote operations on the nodes in batches when there are multiple nodes requiring remote operations, and perform remote operations on at least two nodes in each batch.

Preferably, the batch processing module includes:

a batch division unit, configured to divide the nodes into one or more batches, each batch including at least two nodes;

The thread starting unit is used to start a unique corresponding thread for each node that needs to be remotely operated in this batch, and detect whether each node can log in remotely;

The operation execution unit is used to perform remote operations on the corresponding nodes using the threads corresponding to each node when all the nodes in this batch can log in remotely;

The thread end unit is used to end the thread corresponding to the node after the remote operation of the node is completed;

The execution management unit is used to start the remote operation on the next batch of nodes after all the threads in the current batch are finished.

Preferably, the device also includes:

The thread data limitation module is used to set the maximum number of threads that can be started to a value greater than 1 and less than or equal to the number of processor cores of the current node.

Preferably, the batch division unit is specifically used for:

When the number of nodes does not exceed the maximum number of threads that can be started, divide the nodes into a batch,

When the number of nodes exceeds the maximum number of threads that can be started, divide the nodes into more than two batches, and each batch includes more than two nodes but less than the maximum number of threads.

The present invention provides a multi-node remote operation method and device. When there are multiple nodes requiring remote operation, remote operations are performed on the nodes in batches, and at least two nodes are remotely operated in each batch. Operation, realizes batch and efficient remote operation of nodes, increases the number of nodes for remote operation within the same time-consuming time, improves the efficiency of remote operation of nodes, and solves the problem of low efficiency of remote operation of multiple nodes.

Description of drawings

FIG. 1 is a flowchart of a multi-node remote operation method provided by Embodiment 1 of the present invention;

Fig. 2 is the specific flowchart of step 103 in Fig. 1;

FIG. 3 is a schematic structural diagram of a multi-node remote operation device provided by Embodiment 2 of the present invention;

FIG. 4 is a schematic structural diagram of the batch processing module 302 in FIG. 3 .

Detailed ways

In the management of large-scale clusters, it is inevitable to perform remote batch operations on each node server contained in the cluster. Generally, batch operations are performed using scripts. The general process of scripts is through loops, and the names of node servers in the list are specified remotely in sequence. , and do the same. Because this method executes remote operations sequentially, when the number of node servers in the cluster is hundreds, thousands or even tens of thousands, in this case, the time required for the sequential execution method will be several times that of a single server. Hundreds, thousands or even tens of thousands of times.

Embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined arbitrarily with each other.

First, Embodiment 1 of the present invention will be described with reference to the accompanying drawings.

An embodiment of the present invention provides a multi-node remote operation method. The process of using this method to perform batch remote operations on multiple nodes is shown in Figure 1, including:

Step 101, obtaining node information requiring remote operation;

In the embodiment of the present invention, the entire remote operation process is controlled by running scripts.

In this step, specifically, the server in the cluster (can be any server in the cluster designated to complete the remote operation, preferably, it can be the management server) obtains a preset node name list, and stores the following in the node name list. Information about one or more nodes.

Step 102, setting the maximum number of threads that can be started to a value greater than 1 and less than or equal to the number of processor cores of the current node;

In this step, the maximum number of threads that can be started by the script is set. Considering the processing capacity of the processor, it can be set to a value greater than 1 and less than or equal to the number of processor cores of the current node; preferably, it can be set to the current The value of the number of processor cores on the node.

Step 103, when there are multiple nodes requiring remote operations, perform remote operations on the nodes in batches, and perform remote operations on at least two nodes in each batch;

This step is specifically shown in Figure 2, including:

Step 1031, divide the nodes into one or more batches, each batch includes at least two nodes;

In this step, first calculate the total number of nodes to be operated through the obtained node list to count the total number of nodes; then determine whether the total number of nodes exceeds the maximum number of threads that can be started by the script. Specific divisions are made according to the judgment results, including:

When the number of nodes does not exceed the maximum number of threads that can be started, divide the nodes into a batch;

When the number of nodes exceeds the maximum number of threads that can be started, divide the nodes into more than two batches, and each batch includes more than two nodes but less than the maximum number of threads.

Then, remote operations can be performed on the above-mentioned multiple nodes in batches.

Step 1032, start a unique corresponding thread for each node that needs remote operation in this batch, and detect whether each node can log in remotely;

In this step, the started threads are used to detect in parallel whether the nodes in the node list can log in remotely. According to the detection results, when all the nodes in this batch have successfully logged in remotely, go to step 1033 , otherwise go to step 1034 .

Step 1033 , when all the nodes in this batch can log in remotely, use the threads corresponding to each node to perform remote operations on the corresponding nodes, and then go to step 1035 .

Step 1034, when there are nodes that cannot be logged in remotely, stop the remote operation of all nodes, and output the information of the nodes that cannot be logged in;

In this step, if it is detected that a node cannot be logged in remotely, the output prompts that the corresponding node cannot be logged in remotely, prompts the user to modify the node name list or check the corresponding node, and exits the script to stop the remote operation of all nodes.

Step 1035, after the remote operation of the node is completed, end the thread corresponding to the node;

Step 1036, after all the threads in this batch are finished, start the remote operation on the next batch of nodes;

In this step, if the total number of nodes exceeds the maximum number of threads that can be started by the script, the batch operations on different nodes are executed in parallel with the number of started threads each time. After all the threads started at one time are finished, the next batch of threads is started. The batch operation is completed in batches until the last node in the node name list.

Embodiment 2 of the present invention will be described below with reference to the accompanying drawings.

An embodiment of the present invention provides a multi-node remote operation device, the structure of which is shown in Figure 3, including:

Node information acquisition module 301, configured to acquire node information requiring remote operation;

The batch processing module 302 is configured to perform remote operations on the nodes in batches when there are multiple nodes requiring remote operations, and perform remote operations on at least two nodes in each batch.

Preferably, the structure of the batch processing module 302 is as shown in Figure 4, including:

A batch division unit 3021, configured to divide the nodes into one or more batches, each batch including at least two nodes;

The thread starting unit 3022 is used to start a unique corresponding thread for the nodes that need to be remotely operated in this batch, and detect whether each node can log in remotely;

The operation execution unit 3023 is configured to use the thread corresponding to each node to remotely operate the corresponding node when all the nodes in this batch can log in remotely;

The thread end unit 3024 is used to end the thread corresponding to the node after the remote operation of the node is completed;

The execution management unit 3025 is configured to start remote operations on the next batch of nodes after all the threads in the current batch are finished.

Preferably, the device also includes:

The thread data limitation module 303 is configured to set the maximum number of threads that can be started to a value greater than 1 and less than or equal to the number of processor cores of the current node.

Preferably, the batch division unit 3021 is specifically used for:

When the number of nodes does not exceed the maximum number of threads that can be started, divide the nodes into a batch,

When the number of nodes exceeds the maximum number of threads that can be started, divide the nodes into more than two batches, and each batch includes more than two nodes but less than the maximum number of threads.

The above-mentioned multi-node remote operation device can be integrated in the server in the cluster, and combined with a multi-node remote operation method provided by Embodiment 1 of the present invention, when there are multiple nodes requiring remote operation, batches Perform remote operations on the nodes in batches, and perform remote operations on at least two of the nodes in each batch, realizing batch and efficient remote operation of nodes, and increasing the number of nodes for remote operations within the same time-consuming time, improving the efficiency of remote operation of nodes Efficiency, which solves the problem of low efficiency of multi-node remote operation.

Those of ordinary skill in the art can understand that all or part of the steps of the above-mentioned embodiments can be implemented using a computer program flow, the computer program can be stored in a computer-readable storage medium, and the computer program can be run on a corresponding hardware platform (such as system, device, device, device, etc.), and when executed, includes one or a combination of the steps of the method embodiment.

Optionally, all or part of the steps in the above embodiments can also be implemented using integrated circuits, and these steps can be fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module accomplish. As such, the present invention is not limited to any specific combination of hardware and software.

The devices/functional modules/functional units in the above embodiments can be realized by general-purpose computing devices, and they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices.

When each device/functional module/functional unit in the above-mentioned embodiments is realized in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, and the like.

Any person familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present invention, and all should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope described in the claims.

Claims (10)

1. A multi-node remote operation method, characterized in that, comprising: Obtain the node information that requires remote operation; When the number of nodes requiring remote operation is multiple, remote operations are performed on the nodes in batches, and at least two nodes are remotely operated in each batch. 2. The multi-node remote operation method according to claim 1, wherein obtaining the node information requiring remote operation comprises: A preset node name list is obtained, and information of one or more nodes is stored in the node name list. 3. The remote batch operation method according to claim 1, wherein when there are multiple nodes requiring remote operations, remote operations are performed on the nodes in batches, and each batch is for at least two nodes. The remote operation of the above nodes includes: dividing the nodes into one or more batches, each batch comprising at least two nodes; Start a unique corresponding thread for each node that needs to be remotely operated in this batch, and check whether each node can log in remotely; When all the nodes in this batch can log in remotely, use the threads corresponding to each node to perform remote operations on the corresponding nodes; After the remote operation of the node is completed, end the thread corresponding to the node; After all the threads in this batch are finished, start the remote operation on the next batch of nodes. 4. The multi-node remote operation method according to claim 3, wherein when there are multiple nodes requiring remote operation, remote operations are performed on the nodes in batches, and each batch pairs at least two Before the step of remote operation of the node, it also includes: Set the maximum number of threads that can be started to a value greater than 1 and less than or equal to the number of processor cores of the current node. 5. The multi-node remote operation method according to claim 4, wherein the nodes are divided into one or more batches, and each batch includes at least two nodes including: When the number of nodes does not exceed the maximum number of threads that can be started, divide the nodes into a batch; When the number of nodes exceeds the maximum number of threads that can be started, divide the nodes into more than two batches, and each batch includes more than two nodes but less than the maximum number of threads. 6. The multi-node remote operation method according to claim 3, characterized in that, each start a unique corresponding thread for the nodes that need to be remotely operated in this batch, and after detecting whether each node can log in remotely, it also includes : When there is a node that cannot be logged in remotely, the remote operation to all nodes is stopped, and the information of the node that cannot be logged in is output. 7. A multi-node remote operation device, characterized in that, comprising: The node information obtaining module is used to obtain the node information requiring remote operation; The batch processing module is configured to perform remote operations on the nodes in batches when there are multiple nodes requiring remote operations, and perform remote operations on at least two nodes in each batch. 8. The multi-node remote operation device according to claim 7, wherein the batch processing module comprises: a batch division unit, configured to divide the nodes into one or more batches, each batch including at least two nodes; The thread starting unit is used to start a unique corresponding thread for each node that needs to be remotely operated in this batch, and detect whether each node can log in remotely; The operation execution unit is used to perform remote operations on the corresponding nodes using the threads corresponding to each node when all the nodes in this batch can log in remotely; The thread end unit is used to end the thread corresponding to the node after the remote operation of the node is completed; The execution management unit is used to start the remote operation on the next batch of nodes after all the threads in the current batch are finished. 9. The multi-node remote operation device according to claim 8, characterized in that the device further comprises: The thread data limitation module is used to set the maximum number of threads that can be started to a value greater than 1 and less than or equal to the number of processor cores of the current node. 10. The multi-node remote operation device according to claim 9, wherein the batch division unit is specifically used for: When the number of nodes does not exceed the maximum number of threads that can be started, divide the nodes into a batch, When the number of nodes exceeds the maximum number of threads that can be started, divide the nodes into more than two batches, and each batch includes more than two nodes but less than the maximum number of threads.