CN107018163B - Resource allocation method and device - Google Patents
Resource allocation method and device Download PDFInfo
- Publication number
- CN107018163B CN107018163B CN201610060283.8A CN201610060283A CN107018163B CN 107018163 B CN107018163 B CN 107018163B CN 201610060283 A CN201610060283 A CN 201610060283A CN 107018163 B CN107018163 B CN 107018163B
- Authority
- CN
- China
- Prior art keywords
- server
- data block
- cloud classroom
- preset
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
Abstract
The embodiment of the invention discloses a resource configuration method, which comprises the steps of acquiring the load capacity and the user response time of each cloud classroom in a cloud classroom system every other first preset time, and respectively determining the load capacity change values of two continuous time points of each cloud classroom; determining the cloud classroom with the load change value being greater than or equal to a first preset value as a first cloud classroom, and calling a load balancing application programming interface to acquire calling information when the user response time change values of two continuous time points of the first cloud classroom are greater than or equal to a second preset value; and carrying out load balancing on the servers of the first cloud classroom according to the calling information. The embodiment of the invention also discloses a resource allocation device.
Description
Technical Field
The present invention relates to the field of big data technologies, and in particular, to a resource allocation method and apparatus.
Background
Big data and cloud computing technologies have emerged as the fundamental platforms for various business systems. Since the big data architecture is developed based on the requirement of accessing and processing the huge files in the streaming data mode, the big data technology architecture is particularly suitable for education service systems, such as education clouds, cloud classrooms and the like.
In the prior art, when load balancing is required for sudden load increase of a cloud classroom, a Distributed File System (HDFS) cannot sense changes of an external System, and can only passively wait for calling of an Application Programming Interface (API) of a System, and the load balancing performed by increasing and decreasing servers causes resource waste. Because the HDFS is not suitable for processing small pieces of data, when a large amount of small pieces of data are read and written in a cloud classroom, the Input/Output (I/O) performance of the cloud classroom is seriously affected. The HDFS keeps high fault tolerance through backup data when data is written, and basically keeps a 1:2 backup mode, so that when data is written frequently in a plurality of cloud classrooms, the simultaneous implementation of data backup can cause a plurality of servers to write data frequently at the same time, and the performance of reading operation is influenced. Therefore, the problem of I/O performance reduction exists when the cloud classroom based on big data is used for load balancing, data reading and writing and data backup processing.
Disclosure of Invention
In order to solve the above technical problems, embodiments of the present invention are expected to provide a resource configuration method and device, which effectively improve the I/O performance of a cloud classroom and have high throughput, high scalability, and high fault tolerance.
The technical scheme of the invention is realized as follows:
the embodiment of the invention provides a resource allocation method, which comprises the following steps:
acquiring the load capacity and the user response time of each cloud classroom in a cloud classroom system every other first preset time, and respectively determining the load capacity change values of two continuous time points of each cloud classroom;
determining the cloud classroom with the load change value being greater than or equal to a first preset value as a first cloud classroom, and calling a load balancing application programming interface to acquire calling information when the user response time change values of two continuous time points of the first cloud classroom are greater than or equal to a second preset value;
and carrying out load balancing on the servers of the first cloud classroom according to the calling information.
In the above solution, the calling information at least includes: a first number of each server in the first cloud classroom, a second number of data blocks in each server, wherein each server in the first cloud classroom includes at least one data block;
the load balancing of the servers of the first cloud classroom according to the calling information includes:
determining a duty cycle value of each server according to the input/output load total amount in unit time, the read-write operand in unit time and the data block read-write total amount in unit time of each server in the first cloud classroom, determining the server with the largest duty cycle value as a first server, and determining the server with the smallest duty cycle value as a second server;
determining a data block with the maximum total number of read-write operations in a unit time in the first server as a first data block;
copying the first data block to a second server, and deleting the first data block in the first server;
a third server and a fourth server for obtaining the backup data storage of the first data block;
when the third server or a fourth server is the same as the second server, storing the backup data of the first data block in any server except the first server, the second server, the third server and the fourth server in the first cloud classroom.
The embodiment of the invention provides a resource allocation method, which comprises the following steps:
acquiring write operation commands, and arranging the same write operation commands in a cloud classroom in the same queue, wherein the write operation commands carry a first data block;
when the first data block is smaller than a preset data block multiplied by a first preset value and the number of write operations in the cloud classroom at a first preset time is larger than the average number of write operations per minute in the cloud classroom multiplied by a second preset value, acquiring a write operation command at intervals of a second preset time;
delaying the obtained write operation commands for a third preset time to obtain a third preset number of the write operation commands, wherein the product of the first preset value and the third preset value is 1;
and acquiring the server number of the third preset value of write operation commands, and executing the third preset value of write operation commands.
The embodiment of the invention provides a resource allocation method, which comprises the following steps:
the method comprises the steps of obtaining read operation commands, arranging the same read operation commands in a cloud classroom in the same queue, wherein the read operation commands carry a first data block;
when the first data block is smaller than a preset data block, determining a data block adjacent to the first data block in a queue where the first data block is located;
and when the data block adjacent to the first data block is an unread data block, increasing the read data block to the preset data block which is multiplied by a first preset value, and executing the read operation.
The embodiment of the invention provides a resource allocation method, which comprises the following steps:
the method comprises the steps of obtaining a write operation command, and determining a backup operation command according to the write operation command when the number of write operations in a cloud classroom in a first preset time is larger than the average number of the write operations in the cloud classroom per minute, wherein the average number is multiplied by a first preset value;
arranging the unexecuted backup operation commands in a back-up queue list;
when the number of write operations in second preset time is larger than a second preset value, delaying the backup operation command by the second preset time;
when the number of write operations in a second preset time is less than or equal to a second preset value, acquiring a first backup operation command in the backup queue list, wherein the first backup operation command carries a first data block;
and determining data information of the first data block on the main server, and writing the data information into any two different servers except the main server.
The embodiment of the invention also provides a resource allocation device, which comprises: a first obtaining unit, a first determining unit, a first calling unit and a first processing unit,
the first acquisition unit is used for acquiring the load capacity and the user response time of each cloud classroom in the cloud classroom system every other first preset time;
the first determining unit is used for respectively determining load change values of two continuous time points of each cloud classroom; the cloud classroom which is more than or equal to the load change value and is equal to or higher than a first preset value is determined as a first cloud classroom;
the first calling unit is used for calling a load balancing application programming interface to acquire calling information when the user response time change value of the first cloud classroom at two continuous time points is greater than or equal to a second preset value;
the first processing unit is used for carrying out load balancing on the servers of the first cloud classroom according to the calling information.
In the above solution, the calling information at least includes: a first number of each server in the first cloud classroom, a second number of data blocks in each server, wherein each server in the first cloud classroom includes at least one data block;
the first determining unit is used for determining a duty cycle value of each server according to the total input/output load in unit time, the number of read-write operations in unit time and the total read-write amount of the data block in unit time of each server in the first cloud classroom, determining the server with the maximum duty cycle value as a first server, and determining the server with the minimum duty cycle value as a second server; the data block which is the largest in total number of read-write operations in unit time in the first server is determined to be a first data block;
the first processing unit is used for copying the first data block to a second server and deleting the first data block in the first server; the server is further configured to store backup data of the first data block in any server except the first server, the second server, the third server and the fourth server in the first cloud classroom when the third server or the fourth server is the same as the second server;
the first obtaining unit is configured to obtain a third server and a fourth server where the backup data of the first data block is stored.
The embodiment of the invention also provides a resource allocation device, which comprises: a second acquisition unit, a second processing unit, wherein,
the second acquisition unit is used for acquiring write operation commands and arranging the same write operation commands in the cloud classroom in the same queue, wherein the write operation commands carry a first data block; the cloud classroom data processing device is also used for acquiring a write operation command every second preset time when the first data block is smaller than a preset data block multiplied by a first preset value and the number of write operations in the cloud classroom at a first preset time is larger than the average number of write operations per minute in the cloud classroom multiplied by a second preset value; the write operation control unit is further configured to obtain a third preset value of the write operation commands, where a product of the first preset value and the third preset value is 1; the server number of the third preset value write operation command is also obtained;
the second processing unit is used for delaying the acquired write operation command for a third preset time; and the processor is also used for executing the third preset value of write operation command.
The embodiment of the invention also provides a resource allocation device, which comprises: a third obtaining unit, a third determining unit, a third processing unit, wherein,
the third acquisition unit is used for acquiring the read operation commands and arranging the same read operation commands in the cloud classroom in the same queue, wherein the read operation commands carry the first data block;
the third determining unit is configured to determine, when the first data block is smaller than a preset data block, a data block adjacent to the first data block in a queue where the first data block is located;
and the third processing unit is used for increasing the read data block to the preset data block which is multiplied by a first preset value when the data block adjacent to the first data block is an unread data block, and executing a read operation.
The embodiment of the invention also provides a resource allocation device, which comprises: a fourth obtaining unit, a fourth determining unit, a fourth processing unit, wherein,
the fourth acquisition unit is used for acquiring a write operation command; the backup queue management device is further used for acquiring a first backup operation command in the backup queue list when the number of write operations in a second preset time is less than or equal to a second preset value, wherein the first backup operation command carries a first data block;
the fourth determining unit is used for determining a backup operation command according to the write operation command when the number of write operations in the cloud classroom in a first preset time is larger than the average number of write operations per minute of the cloud classroom multiplied by a first preset value; the data information of the first data block on the main server is also determined;
the fourth processing unit is configured to arrange the unexecuted backup operation command in a back-up queue list; the backup operation command is delayed for a second preset time when the number of write operations in the second preset time is larger than a second preset value; and the data information is written into any two different servers except the main server.
The embodiment of the invention provides a resource configuration method and a resource configuration device, wherein the load capacity and the user response time of each cloud classroom in a cloud classroom system are acquired every other first preset time, and the load capacity change values of two continuous time points of each cloud classroom are respectively determined; determining the cloud classroom with the load change value being greater than or equal to a first preset value as a first cloud classroom, and calling a load balancing application programming interface to acquire calling information when the user response time change values of two continuous time points of the first cloud classroom are greater than or equal to a second preset value; and carrying out load balancing on the servers of the first cloud classroom according to the calling information. The resource configuration method and the resource configuration device provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solve the problem of overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improve the I/O performance of the cloud classroom, and have high throughput, high expandability and high fault tolerance.
Drawings
Fig. 1 is a schematic flowchart of a resource allocation method according to a first embodiment;
fig. 2 is a schematic flowchart of a resource allocation method according to a second embodiment;
fig. 3 is a schematic flowchart of a resource allocation method according to a third embodiment;
fig. 4 is a schematic flowchart of a resource allocation method according to a fourth embodiment;
fig. 5 is a schematic flowchart of a resource allocation method according to the fifth embodiment;
fig. 6 is a schematic structural diagram of a resource allocation apparatus according to a sixth embodiment;
fig. 7 is a schematic structural diagram of a resource allocation apparatus according to a seventh embodiment;
fig. 8 is a schematic structural diagram of a resource allocation apparatus according to an eighth embodiment;
fig. 9 is a schematic structural diagram of a resource allocation apparatus according to a ninth embodiment.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
Example one
An embodiment of the present invention provides a resource allocation method, as shown in fig. 1, the method may include:
s101, the resource configuration device acquires the load and the user response time of each cloud classroom in the cloud classroom system every other first preset time, and the load change values of two continuous time points of each cloud classroom are respectively determined.
A cloud classroom system is composed of an HDFS and a plurality of servers, and a plurality of cloud classrooms are borne on the cloud classroom system. The HDFS enables all the servers to form a virtual resource pool, and the required servers are extracted from the whole virtual resource pool when each cloud classroom is established to form a cloud classroom.
The HDFS comprises a metanode NameNode, a data node DataNode and a Client. The metanode stores and manages various metadata such as a file name, an index, a location, and the like. And the data nodes are responsible for storing and managing various data contents. The client provides various API interfaces, allowing external program calls.
The cloud classroom load detection module monitors the load and the user response time of each cloud classroom in the cloud classroom system every other first preset time to obtain load change values of each cloud classroom at two continuous time points. The cloud classroom load monitoring module runs on a cloud classroom Web end server and directly interacts with a plurality of cloud classrooms. The cloud classroom Web system is a Web management platform for managing cloud classrooms, is independently operated on an independent server system, and is a platform for a foreground user interface and a background teacher interface.
Illustratively, the load and the user response time of each cloud classroom in a cloud classroom system are monitored every 1s by a cloud classroom system comprising 100 cloud classrooms, and the load change value between the last second and the next second of each cloud classroom in the cloud classroom system is obtained.
And S102, the resource configuration device determines that the cloud classroom with the load change value larger than or equal to a first preset value is a first cloud classroom, and when the user response time change value of two continuous time points of the first cloud classroom is larger than or equal to a second preset value, the resource configuration device calls a load balancing application programming interface to obtain calling information.
When the load capacity of one cloud classroom in the cloud classroom system sharply increases at two continuous time points, the user response time change values of the cloud classroom at the two continuous time points are measured. And if the user response time of the cloud classroom at two continuous time points is obviously prolonged, the resource configuration device calls the load balancing application programming interface to call the load balancing command. It should be noted here that the load balancing command is not a load balancing adjustment command existing inside the HDFS, and the load balancing command inside the HDFS is mainly balanced according to the data block distribution capacity on the server. The load balancing method of this embodiment is to process according to the busy/idle condition of the server being used, and needs to implement a load balancing adjustment module in the HDFS system, which processes according to the busy/idle condition of the server. Calling the load balancing adjustment module of this embodiment requires adding a new API to the HDFS client to allow the application calls of the upper layer. In this embodiment, the load balancing command inside the HDFS is not called.
For example, if the load of one of the cloud classrooms in the cloud classroom system is increased by 30% or more at two consecutive time points, the user response time change values of the cloud classroom at the two consecutive time points are measured. And when the user response time increment value of two continuous time points of the cloud classroom is more than or equal to 20%, the resource configuration device calls the load balancing application programming interface to call the load balancing command.
S103, the resource configuration device performs load balancing on the servers of the first cloud classroom according to the calling information.
And the resource configuration device calls the load balancing command and performs load balancing processing on the first cloud classroom according to the busy and idle condition of the server in use. In a simulation test of an actual system, the load balancing method of the embodiment can effectively improve the I/O performance by about 129% under the conditions of the same user, the same load and the same HDFS server.
The resource configuration method provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
Example two
An embodiment of the present invention provides a resource allocation method, as shown in fig. 2, the method may include:
s201, the resource configuration device acquires the load and the user response time of each cloud classroom in the cloud classroom system every other first preset time, and the load change values of two continuous time points of each cloud classroom are respectively determined.
The cloud classroom system comprises a plurality of cloud classrooms, the load capacity and the user response time of each cloud classroom in the cloud classroom system are monitored every other first preset time, and the load capacity change value of each cloud classroom at two continuous time points is obtained.
And S202, the resource configuration device determines that the cloud classroom with the load change value larger than or equal to a first preset value is a first cloud classroom, and when the user response time change value of two continuous time points of the first cloud classroom is larger than or equal to a second preset value, the resource configuration device calls a load balancing application programming interface to obtain calling information.
When the load capacity of one cloud classroom in the cloud classroom system sharply increases at two continuous time points, the user response time change values of the cloud classroom at the two continuous time points are measured. And if the user response time of the cloud classroom at two continuous time points is obviously prolonged, the resource configuration device calls the load balancing application programming interface to call the load balancing command.
S203, the resource allocation device determines the duty cycle value of each server according to the input/output load total amount in unit time, the read-write operand in unit time and the data block read-write total amount in unit time of each server in the first cloud classroom, determines the server with the maximum duty cycle value as a first server, and determines the server with the minimum duty cycle value as a second server.
Each cloud classroom is composed of a plurality of servers, and the resource configuration device calculates the duty cycle value of each server in the first cloud classroom. And calculating and obtaining the duty cycle value B of each server according to the following formula:
B＝M*(M
2/M
1)*(N
2/N
1)
wherein, M is the total input/output load of the server in unit time; the M is
1Reading operation number of the server in unit time; the M is
2Writing operation number of the server in unit time; said N is
1Reading out the total amount of the data blocks in unit time for the server; said N is
2Writing a total amount for the server for the data blocks in the unit time.
According to the formula, the larger the duty cycle value B of the server is, the larger the load capacity of the server is, the more busy the server is; conversely, the smaller the duty cycle value B of the server is, the smaller the load of the server is, the more idle the server is.
The server with the maximum duty cycle value is the first server, namely the most busy server; the server with the lowest duty cycle value is the second server, i.e., the most idle server. When the duty cycle of each server in a cloud classroom is calculated, the situations that the duty cycles of several servers are the same and are all maximum values or the duty cycles of several servers are the same and are all minimum values may occur. In this case, only one server with the largest duty cycle is selected as the first server, and similarly, only one server with the smallest duty cycle is selected as the second server.
S204, the resource allocation device determines the data block with the maximum total number of read-write operations in the first server in unit time as a first data block.
Specifically, in the most busy server, the data block that the user has the most read and write operations per unit time is the most busy data block. And the resource allocation device positions the most busy data block, migrates the most busy data block and performs load balancing.
The granularity of cloud classroom management is servers, not data blocks. I.e., increasing performance by increasing the number of servers when I/O performance is problematic, and decreasing the number of servers when not needed. The method has too coarse management granularity and too large resource utilization. Therefore, the embodiment processes the busy data block of the busy server, and reasonably adjusts and controls the available resources.
S205, the resource allocation device copies the first data block to a second server, and deletes the first data block in the first server.
Specifically, after the resource allocation device copies the busy data block to the idle server, the busy data block on the busy server is deleted.
S206, the resource configuration device acquires a third server and a fourth server of the backup data storage of the first data block.
In the HDFS according to the embodiment of the present invention, each data block is default to store 2 backup data blocks in addition to the data block itself. The backup data of the first data block is stored in any two servers other than the first server.
The number of the backup data blocks can also be set according to needs, and the invention is not particularly limited.
And S207, when the third server or the fourth server is the same as the second server, the resource configuration device stores the backup data of the first data block in any server except the first server, the second server, the third server and the fourth server in the first cloud classroom.
Because the backup data and the main data cannot be in the same server, when one of the backup data of the first data block is stored in the second server, the main data and the backup data of the first data block are simultaneously stored in the second server, and the backup data of the first data block needs to be adjusted to migrate the backup data. And storing the backup data of the first data block in any server except the first server, the second server, the third server and the fourth server in the first cloud classroom.
The resource configuration method provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
EXAMPLE III
An embodiment of the present invention provides a resource allocation method, as shown in fig. 3, the method may include:
s301, the resource configuration device obtains the write operation commands, and arranges the same write operation commands in the cloud classroom in the same queue, wherein the write operation commands carry the first data block.
Different cloud classrooms have different cloud classroom numbers, and the write operation commands of the same cloud classroom are arranged in the same write command queue. The write command has a buffer to which a pointer points, the buffer storing specific data.
Cloud classrooms based on big data technology are very sensitive to reading and writing of a large amount of small data, which can greatly reduce the user performance of the cloud classrooms, because big data mainly handles big files and streaming data.
In various complex cloud classroom environments, frequent small file read-write operations in a certain classroom, or frequent small file read operations in a plurality of cloud classrooms, or frequent small file write operations in a plurality of cloud classrooms are frequently encountered, which may affect the user performance of the whole cloud classroom platform. Therefore, aiming at the cloud classroom characteristics of the big data technology, the embodiment provides a delayed writing method under the condition of small file writing.
And S302, when the first data block is smaller than a preset data block multiplied by a first preset value and the number of write operations in the cloud classroom at a first preset time is larger than the average number of write operations per minute in the cloud classroom multiplied by a second preset value, the resource configuration device acquires a write operation command every second preset time.
The default basic unit of storage in HDFS is a data block size of 64M, i.e. the standard data block size is 64M. In this embodiment, the predetermined data block is a standard data block. Illustratively, the frequency of cloud classroom write operations is calculated if the first data chunk is less than 1/5 times the standard data chunk size. The average number of write operands per minute in cloud classroom is X, the last 1 minute write operand in cloud classroom is Y, and when Y >3X, the write operations in cloud classroom are frequent. When the write operation of the cloud classroom is frequent, the resource configuration device acquires the next write operation command in the write command queue every 2 ms.
S303, the resource allocation device delays the obtained write operation command by a third preset time to obtain a third preset number of the write operation commands, wherein the product of the first preset value and the third preset value is 1.
The resource configuration device delays the obtained write operation commands each time for 2ms, and if the write operation commands are all smaller than 1/5 times of the size of a standard data block and the write operation of a cloud classroom is frequent, 5 write operation commands are obtained, and the 5 write operation commands are integrated to form a complete write operation command.
In this embodiment, the first preset value multiple may also be set to other values, and the present invention is not limited specifically, but the product of the first preset value and the third preset value is required to be 1.
S304, the resource configuration device acquires the server number of the third preset value of write operation commands, and executes the third preset value of write operation commands.
In step S303, 5 write operation commands are integrated to form a complete write operation command, the resource allocation device obtains the server number of the integrated command from the meta node, executes the write operation command, and writes the integrated command into the corresponding server to complete the deferred write operation.
The resource configuration method provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
Example four
An embodiment of the present invention provides a resource allocation method, as shown in fig. 4, the method may include:
s401, the resource configuration device obtains the read operation commands, the same read operation commands in the cloud classroom are arranged in the same queue, and the read operation commands carry the first data block.
And arranging the read operation commands of the same cloud classroom in the read command queue of the same cloud classroom according to the serial number of the cloud classroom.
Large and frequent small data read operations in a cloud classroom based on big data technology can significantly reduce user performance. Aiming at the characteristics of a cloud classroom of a big data technology, the embodiment provides a pre-reading method under the condition of reading a small file.
S402, when the first data block is smaller than a preset data block, the resource configuration device determines a data block adjacent to the first data block in a queue where the first data block is located.
As in step S302, the preset data block of the present embodiment is a standard data block, i.e., 64M. When the first data block is smaller than the standard data block size, the resource configuration device acquires the data block number and the server number of the first data block and a next read command in the cloud classroom read command queue from the meta node. And comparing the data block of the next read command in the cloud classroom read command queue with the physical address of the first data block, and if the physical addresses of the data blocks are adjacent, determining that the data block of the next read command in the cloud classroom read command queue is adjacent to the first data block.
And S403, when the data block adjacent to the first data block is an unread data block, the resource allocation device increases the read data block to the preset data block which is the first preset value times, and executes the read operation.
When the data block of the next read command in the cloud classroom read command queue is the unread data block, the size of the first data block to be read is increased to 2 times of the size of the standard data block, and the data block of the next read command in the cloud classroom read command queue is included. After the data blocks are increased, an extended read command is formed, and the read command is executed to complete the pre-read operation. In the cloud classroom mode, the HDFS system independently adopts the delayed writing and pre-reading methods of the third embodiment and the fourth embodiment, and the I/O performance of the cloud classroom can be effectively improved by about 42%. The delayed writing and pre-reading method is a new method realized aiming at frequent small data block reading and writing commands, is presented in a software form in specific realization and is stored on a server where an HDFS client is positioned.
The resource configuration method provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
EXAMPLE five
An embodiment of the present invention provides a resource allocation method, as shown in fig. 5, the method may include:
s501, the resource configuration device obtains a write operation command, and when the number of write operations in the cloud classroom in a first preset time is larger than the average number of write operations per minute of the cloud classroom multiplied by a first preset value, the resource configuration device determines a backup operation command according to the write operation command.
Because the big data HDFS system improves the high fault tolerance of data through 2 backups when writing data, the data backup is realized when writing data under the condition of infrequent writing operation in a cloud classroom, and the fault tolerance of the system can be effectively improved. However, when the cloud classrooms write frequently, especially when a plurality of cloud classrooms write frequently at the same time, the data backup is realized at the same time, which brings great influence on the I/O performance of the cloud classroom system, so the embodiment provides the backup method under the condition that the write operation of the cloud classrooms is frequent.
The HDFS client receives a write operation command sent by a cloud classroom, judges the frequent condition of the write operation of the cloud classroom, and adopts a backup copy technology if the write operation of the cloud classroom is frequent. The frequent cloud classroom write operation condition is determined according to the write operation number of the cloud classroom in the first preset time and the average cloud classroom write operation per minute which is twice as long as the first preset time, and the specific cloud classroom write operation frequent condition judgment method is the same as that in the step S302. The backup operation command needs to be determined according to the write operation command of the cloud classroom before the backup operation.
S502, the resource configuration device arranges the unexecuted backup operation commands in a rear backup queue list.
And the HDFS client acquires the storage position of the data block in the data node and the storage position of the backup data from the meta node. And the HDFS client writes data into the main server and arranges the unexecuted backup operation commands in the back backup queue list.
And S503, when the number of write operations in the second preset time is greater than a second preset value, the resource configuration device delays the backup operation command by the second preset time.
And if an unexecuted backup operation command exists when the backup queue list is inquired, judging the frequent condition of the write operation of the HDFS. And when the number of the write operations in the second preset time is greater than the second preset value, the HDFS write operations are frequent. Illustratively, the write operation number of the HDFS in the last 1 minute is N, and if N >5, the write operation of the HDFS is determined to be frequent. And when the HDFS write operation is determined to be frequent, delaying the backup operation command for 2ms, and inquiring the unexecuted backup operation command in the backup queue list again.
And S504, when the number of write operations in the second preset time is less than or equal to a second preset value, the resource configuration device acquires a first backup operation command in the backup queue table, wherein the first backup operation command carries a first data block.
When the HDFS write operation is determined to be infrequent, the backup operation is started to be executed. And acquiring a first backup operation command in the backup queue list, wherein the backup operation command carries a first data block.
S505, the resource allocation device determines the data information of the first data block on the main server, and writes the data information into any two different servers except the main server.
Since the data information on the primary server and the two backup data information cannot be stored on the same server, the backup operation can be completed by writing the data information into any two different servers except the primary server after the resource allocation device obtains the data information of the first data block on the primary server.
The resource configuration method provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
EXAMPLE six
An embodiment of the present invention provides a resource allocation apparatus 6, as shown in fig. 6, the apparatus includes: a first obtaining unit 601, a first determining unit 602, a first calling unit 603, and a first processing unit 604, wherein,
the first obtaining unit 601 is configured to obtain a load amount and a user response time of each cloud classroom in the cloud classroom system every first preset time;
the first determining unit 602 is configured to determine load change values of two consecutive time points of each cloud classroom respectively; the cloud classroom which is more than or equal to the load change value and is equal to or higher than a first preset value is determined as a first cloud classroom;
the first calling unit 603 is configured to call a load balancing application programming interface to obtain calling information when a user response time variation value at two consecutive time points in the first cloud classroom is greater than or equal to a second preset value;
the first processing unit 604 is configured to perform load balancing on the servers of the first cloud classroom according to the call information.
Further, the calling information at least comprises: a first number of each server in the first cloud classroom, a second number of data blocks in each server, wherein each server in the first cloud classroom includes at least one data block;
the first determining unit 602 is configured to determine a duty cycle value of each server according to an input/output load total amount per unit time, a read-write operand per unit time, and a data block read-write total amount per unit time of each server in the first cloud classroom, determine that a server with the largest duty cycle value is a first server, and determine that a server with the smallest duty cycle value is a second server; the data block which is the largest in total number of read-write operations in unit time in the first server is determined to be a first data block;
the first processing unit 604 is configured to copy the first data block to a second server, and delete the first data block in the first server; the server is further configured to store backup data of the first data block in any server except the first server, the second server, the third server and the fourth server in the first cloud classroom when the third server or the fourth server is the same as the second server;
the first obtaining unit 601 is configured to obtain a third server and a fourth server of the backup data storage of the first data block.
Specifically, for the description of the resource allocation apparatus provided in the embodiment of the present invention, reference may be made to the description of the resource allocation method in the first embodiment to the second embodiment, and details of the embodiment of the present invention are not repeated herein.
In practical applications, the first obtaining Unit 601, the first determining Unit 602, the first calling Unit 603, and the first Processing Unit 604 may be implemented by a Central Processing Unit (CPU), a microprocessor Unit (MPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), or the like, which are located in the resource configuration device 6.
The resource configuration device provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
EXAMPLE seven
An embodiment of the present invention provides a resource allocation apparatus 7, as shown in fig. 7, the apparatus includes: a second acquisition unit 701, a second processing unit 702, wherein,
the second obtaining unit 701 is configured to obtain write operation commands, and arrange the same write operation commands in a cloud classroom in a same queue, where the write operation commands carry a first data block; the cloud classroom data processing device is also used for acquiring a write operation command every second preset time when the first data block is smaller than a preset data block multiplied by a first preset value and the number of write operations in the cloud classroom at a first preset time is larger than the average number of write operations per minute in the cloud classroom multiplied by a second preset value; the write operation control unit is further configured to obtain a third preset value of the write operation commands, where a product of the first preset value and the third preset value is 1; the server number of the third preset value write operation command is also obtained;
the second processing unit 701 is configured to delay the obtained write operation command by a third preset time; and the processor is also used for executing the third preset value of write operation command.
Specifically, for the description of the resource allocation apparatus provided in the embodiment of the present invention, reference may be made to the description of the resource allocation method in the third embodiment, and details of the embodiment of the present invention are not described herein again.
In practical applications, the second obtaining unit 701 and the second processing unit 702 can be implemented by a CPU, an MPU, a DSP, or an FPGA located in the resource configuration device 7.
The resource configuration device provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
Example eight
An embodiment of the present invention provides a resource allocation apparatus 8, as shown in fig. 8, the apparatus includes: a third acquisition unit 801, a third determination unit 802, a third processing unit 803, wherein,
the third obtaining unit 801 is configured to obtain read operation commands, arrange the same read operation commands in a cloud classroom in the same queue, where the read operation commands carry a first data block;
the third determining unit 802 is configured to determine, when the first data block is smaller than a preset data block, a data block adjacent to the first data block in a queue where the first data block is located;
the third processing unit 803 is configured to, when the data block adjacent to the first data block is an unread data block, increase the read data block to the preset data block that is twice the first preset value, and perform a read operation.
Specifically, for the description of the resource allocation apparatus provided in the embodiment of the present invention, reference may be made to the description of the resource allocation method in the fourth embodiment, and details of the embodiment of the present invention are not described herein again.
In practical applications, the third obtaining unit 801, the third determining unit 802, and the third processing unit 803 may be implemented by a CPU, an MPU, a DSP, an FPGA, or the like located in the resource configuration apparatus 8.
The resource configuration device provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
Example nine
An embodiment of the present invention provides a resource allocation apparatus 9, as shown in fig. 9, the apparatus includes: a fourth acquiring unit 901, a fourth determining unit 902, a fourth processing unit 903, wherein,
the fourth obtaining unit 901 is configured to obtain a write operation command; the backup queue management device is further used for acquiring a first backup operation command in the backup queue list when the number of write operations in a second preset time is less than or equal to a second preset value, wherein the first backup operation command carries a first data block;
the fourth determining unit 902 is configured to determine a backup operation command according to the write operation command when the number of write operations in the cloud classroom at the first preset time is greater than the average number of write operations per minute of the cloud classroom multiplied by the first preset value; the data information of the first data block on the main server is also determined;
the fourth processing unit 903 is configured to arrange the unexecuted backup operation command in a back-up queue list; the backup operation command is delayed for a second preset time when the number of write operations in the second preset time is larger than a second preset value; and the data information is written into any two different servers except the main server.
Specifically, for the description of the resource allocation apparatus provided in the embodiment of the present invention, reference may be made to the description of the resource allocation method in the fifth embodiment, and details of the embodiment of the present invention are not described herein again.
In practical applications, the fourth obtaining unit 901, the fourth determining unit 902, and the fourth processing unit 903 may be implemented by a CPU, an MPU, a DSP, or an FPGA located in the resource configuration device 9.
The resource configuration device provided by the embodiment of the invention can avoid the problem of overlarge resource usage caused by the increase and decrease of servers in the existing load balancing method, simultaneously solves the problem of the overall performance reduction of a cloud classroom caused by frequent data reading and writing, effectively improves the I/O performance of the cloud classroom, and has high throughput, high expandability and high fault tolerance.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention.
Claims (10)
1. A method for resource allocation, the method comprising:
acquiring the load capacity and the user response time of each cloud classroom in a cloud classroom system every other first preset time, and respectively determining the load capacity change values of two continuous time points of each cloud classroom;
determining the cloud classroom with the load change value being greater than or equal to a first preset value as a first cloud classroom, and calling a load balancing application programming interface to acquire calling information when the user response time change values of two continuous time points of the first cloud classroom are greater than or equal to a second preset value;
determining a server with the maximum load capacity in the first cloud classroom as a first server, and determining a server with the minimum load capacity in the first cloud classroom as a second server;
determining a data block with the maximum total number of read-write operations in a unit time in the first server as a first data block;
copying the first data block to a second server, and deleting the first data block in the first server;
storing backup data of the first data block in any two servers in the first cloud classroom other than the first server.
2. The resource allocation method according to claim 1,
the calling information at least comprises: a first number of each server in the first cloud classroom, a second number of data blocks in each server, wherein each server in the first cloud classroom includes at least one data block;
the determining that the server with the maximum load capacity in the first cloud classroom is the first server and the determining that the server with the minimum load capacity in the first cloud classroom is the second server includes:
determining a duty cycle value of each server according to the input/output load total amount in unit time, the read-write operand in unit time and the data block read-write total amount in unit time of each server in the first cloud classroom, determining the server with the largest duty cycle value as a first server, and determining the server with the smallest duty cycle value as a second server; the duty cycle value is used for representing the load capacity of the server;
the storing the backup data of the first data block in any two servers except the first server in the first cloud classroom includes:
a third server and a fourth server for obtaining the backup data storage of the first data block;
when the third server or a fourth server is the same as the second server, storing the backup data of the first data block in any server except the first server, the second server, the third server and the fourth server in the first cloud classroom.
3. A method for resource allocation, the method comprising:
acquiring write operation commands, and arranging the same write operation commands in a cloud classroom in the same queue, wherein the write operation commands carry a first data block;
when the size of the first data block is smaller than a first preset value times of a preset data block, and the number of write operations in the cloud classroom at a first preset time is larger than the average number of write operations in the cloud classroom per minute at a second preset time, acquiring a write operation command at intervals of a second preset time;
delaying the obtained write operation commands for a third preset time to obtain a third preset number of the write operation commands, wherein the product of the first preset value and the third preset value is 1;
and acquiring the server number of the third preset value of write operation commands, and executing the third preset value of write operation commands.
4. A method for resource allocation, the method comprising:
the method comprises the steps of obtaining read operation commands, arranging the same read operation commands in a cloud classroom in the same queue, wherein the read operation commands carry a first data block;
when the first data block is smaller than a preset data block, determining a data block adjacent to the first data block in a queue where the first data block is located;
when the data block adjacent to the first data block is an unread data block, increasing the size of the data block to be read to be a first preset value times of the preset data block to form an expanded read operation command; the expanded read operation command is used for reading the first data block and a data block adjacent to the first data block; and executing the extended read operation command.
5. A method for resource allocation, the method comprising:
the method comprises the steps of obtaining a write operation command, and determining a backup operation command according to the write operation command when the number of write operations in a cloud classroom in a first preset time is larger than the average number of the write operations in the cloud classroom per minute, wherein the average number is multiplied by a first preset value;
arranging the unexecuted backup operation commands in a back-up queue list;
when the number of write operations in second preset time is larger than a second preset value, delaying the backup operation command by the second preset time;
when the number of write operations in a second preset time is less than or equal to a second preset value, acquiring a first backup operation command in the backup queue list, wherein the first backup operation command carries a first data block;
and determining data information of the first data block on the main server, and writing the data information into any two different servers except the main server.
6. An apparatus for resource configuration, the apparatus comprising: a first obtaining unit, a first determining unit, a first calling unit and a first processing unit,
the first acquisition unit is used for acquiring the load capacity and the user response time of each cloud classroom in the cloud classroom system every other first preset time;
the first determining unit is used for respectively determining load change values of two continuous time points of each cloud classroom; the cloud classroom which is more than or equal to the load change value and is equal to or higher than a first preset value is determined as a first cloud classroom; the server with the maximum load capacity in the first cloud classroom is determined to be a first server, and the server with the minimum load capacity in the first cloud classroom is determined to be a second server; the data block which is the largest in total number of read-write operations in unit time in the first server is determined to be a first data block;
the first calling unit is used for calling a load balancing application programming interface to acquire calling information when the user response time change value of the first cloud classroom at two continuous time points is greater than or equal to a second preset value;
the first processing unit is used for copying the first data block to a second server and deleting the first data block in the first server; storing backup data of the first data block in any two servers in the first cloud classroom other than the first server.
7. The apparatus according to claim 6, wherein the call information at least includes: a first number of each server in the first cloud classroom, a second number of data blocks in each server, wherein each server in the first cloud classroom includes at least one data block;
the first determining unit is used for determining a duty cycle value of each server according to the total input/output load in unit time, the number of read-write operations in unit time and the total read-write amount of the data block in unit time of each server in the first cloud classroom, determining the server with the maximum duty cycle value as a first server, and determining the server with the minimum duty cycle value as a second server; the duty cycle value is used for representing the load capacity of the server;
the first obtaining unit is further configured to obtain a third server and a fourth server where the backup data of the first data block is stored;
the first processing unit is further configured to store the backup data of the first data block in any server except the first server, the second server, the third server, and the fourth server in the first cloud classroom when the third server or the fourth server is the same as the second server.
8. An apparatus for resource configuration, the apparatus comprising: a second acquisition unit, a second processing unit, wherein,
the second acquisition unit is used for acquiring write operation commands and arranging the same write operation commands in the cloud classroom in the same queue, wherein the write operation commands carry a first data block; the data processing device is further used for acquiring a write operation command every second preset time when the size of the first data block is smaller than a first preset value of a preset data block and the number of write operations in the cloud classroom at a first preset time is larger than the average number of write operations per minute in the cloud classroom at a second preset time; the write operation control unit is further configured to obtain a third preset value of the write operation commands, where a product of the first preset value and the third preset value is 1; the server number of the third preset value write operation command is also obtained;
the second processing unit is used for delaying the acquired write operation command for a third preset time; and the processor is also used for executing the third preset value of write operation command.
9. An apparatus for resource configuration, the apparatus comprising: a third obtaining unit, a third determining unit, a third processing unit, wherein,
the third acquisition unit is used for acquiring the read operation commands and arranging the same read operation commands in the cloud classroom in the same queue, wherein the read operation commands carry the first data block;
the third determining unit is configured to determine, when the first data block is smaller than a preset data block, a data block adjacent to the first data block in a queue where the first data block is located;
the third processing unit is configured to, when the data block adjacent to the first data block is an unread data block, increase the size of the data block to be read to a first preset value multiple of the preset data block, and form an extended read operation command; and the expanded read operation command is used for reading the first data block and the data block adjacent to the first data block and executing the expanded read operation command.
10. An apparatus for resource configuration, the apparatus comprising: a fourth obtaining unit, a fourth determining unit, a fourth processing unit, wherein,
the fourth acquisition unit is used for acquiring a write operation command; the backup queue management method is also used for acquiring a first backup operation command in a back backup queue list when the number of write operations in a second preset time is less than or equal to a second preset value, wherein the first backup operation command carries a first data block;
the fourth determining unit is used for determining a backup operation command according to the write operation command when the number of write operations in the cloud classroom in a first preset time is larger than the average number of write operations per minute of the cloud classroom multiplied by a first preset value; the data information of the first data block on the main server is also determined;
the fourth processing unit is configured to arrange the unexecuted backup operation command in a back-up queue list; the backup operation command is delayed for a second preset time when the number of write operations in the second preset time is larger than a second preset value; and the data information is written into any two different servers except the main server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610060283.8A CN107018163B (en) | 2016-01-28 | 2016-01-28 | Resource allocation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610060283.8A CN107018163B (en) | 2016-01-28 | 2016-01-28 | Resource allocation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107018163A CN107018163A (en) | 2017-08-04 |
| CN107018163B true CN107018163B (en) | 2020-02-11 |
Family
ID=59439212
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610060283.8A Active CN107018163B (en) | 2016-01-28 | 2016-01-28 | Resource allocation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107018163B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108121312B (en) * | 2017-11-29 | 2020-10-30 | 南瑞集团有限公司 | ARV load balancing system and method based on integrated hydropower management and control platform |
| CN109936589B (en) * | 2017-12-15 | 2021-09-03 | 中国移动通信集团浙江有限公司 | Dynamic load balancing method, device and platform based on API (application program interface) heat value |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102594861A (en) * | 2011-12-15 | 2012-07-18 | 杭州电子科技大学 | Cloud storage system with balanced multi-server load |
| CN103167026A (en) * | 2013-02-06 | 2013-06-19 | 数码辰星科技发展（北京）有限公司 | Processing method, system and device for cloud storage environmental data |
| CN103188346A (en) * | 2013-03-05 | 2013-07-03 | 北京航空航天大学 | Distributed decision making supporting massive high-concurrency access I/O (Input/output) server load balancing system |
| CN104462432A (en) * | 2014-12-15 | 2015-03-25 | 成都英力拓信息技术有限公司 | Self-adaptive distributed computing method |
-
2016
- 2016-01-28 CN CN201610060283.8A patent/CN107018163B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102594861A (en) * | 2011-12-15 | 2012-07-18 | 杭州电子科技大学 | Cloud storage system with balanced multi-server load |
| CN103167026A (en) * | 2013-02-06 | 2013-06-19 | 数码辰星科技发展（北京）有限公司 | Processing method, system and device for cloud storage environmental data |
| CN103188346A (en) * | 2013-03-05 | 2013-07-03 | 北京航空航天大学 | Distributed decision making supporting massive high-concurrency access I/O (Input/output) server load balancing system |
| CN104462432A (en) * | 2014-12-15 | 2015-03-25 | 成都英力拓信息技术有限公司 | Self-adaptive distributed computing method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107018163A (en) | 2017-08-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7138126B2 (en) | Timeliness resource migration to optimize resource placement | |
| KR102281739B1 (en) | Resource scheduling method, scheduling server, cloud computing system, and storage medium | |
| US10635323B2 (en) | Managing storage system | |
| US10146469B2 (en) | Dynamic storage tiering based on predicted workloads | |
| US20190220315A1 (en) | Dynamic Adjustment Of Application Resources In A Distributed Computing System | |
| US9182923B2 (en) | Controlling throughput of processing units associated with different load types in storage system | |
| US10078639B1 (en) | Cluster file system comprising data mover modules having associated quota manager for managing back-end user quotas | |
| US20190220195A1 (en) | Efficient modification of storage system metadata | |
| CN106339386B (en) | Database flexible scheduling method and device | |
| US10356150B1 (en) | Automated repartitioning of streaming data | |
| Xu et al. | {SpringFS}: Bridging Agility and Performance in Elastic Distributed Storage | |
| US20200201763A1 (en) | Memory hierarchy-aware processing | |
| WO2020134364A1 (en) | Virtual machine migration method, cloud computing management platform, and storage medium | |
| CN112988332B (en) | Virtual machine live migration prediction method and system and computer readable storage medium | |
| CN112148430A (en) | Method for online safe migration of virtual machine with virtual network function | |
| CN109582649A (en) | A kind of metadata storing method, device, equipment and readable storage medium storing program for executing | |
| CN107018163B (en) | Resource allocation method and device | |
| CN109426439A (en) | The method and device of dilatation is carried out to distributed memory system | |
| KR101662173B1 (en) | Distributed file management apparatus and method | |
| US20230376357A1 (en) | Scaling virtualization resource units of applications | |
| US10594620B1 (en) | Bit vector analysis for resource placement in a distributed system | |
| GB2585543A (en) | Data migration in a hierarchical storage management system | |
| US10824640B1 (en) | Framework for scheduling concurrent replication cycles | |
| WO2017122263A1 (en) | Management computer and management method | |
| CN106202262A (en) | A kind of information processing method and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |