WO2015131310A1 - 提供数据服务的方法和网络设备 - Google Patents
提供数据服务的方法和网络设备 Download PDFInfo
- Publication number
- WO2015131310A1 WO2015131310A1 PCT/CN2014/072796 CN2014072796W WO2015131310A1 WO 2015131310 A1 WO2015131310 A1 WO 2015131310A1 CN 2014072796 W CN2014072796 W CN 2014072796W WO 2015131310 A1 WO2015131310 A1 WO 2015131310A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- node
- data center
- user equipment
- priority
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 59
- 230000005540 biological transmission Effects 0.000 claims abstract description 278
- 238000005457 optimization Methods 0.000 claims abstract description 26
- 235000008694 Humulus lupulus Nutrition 0.000 claims description 29
- 238000004891 communication Methods 0.000 claims description 11
- 238000005265 energy consumption Methods 0.000 abstract description 5
- 230000002068 genetic effect Effects 0.000 description 18
- 238000010586 diagram Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 5
- 230000001934 delay Effects 0.000 description 5
- 238000012546 transfer Methods 0.000 description 5
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
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
- H04L67/1095—Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes
Definitions
- Embodiments of the present invention relate to the field of information technology, and, more particularly, to a method and network device for providing a data service. Background technique
- the data center node divides the data into different heat levels according to the click rate of the stored data, and copies data of different heat levels to other data center nodes.
- the user equipment can select the source of the data based on the power consumption and the transmission delay.
- the number of network layers that the user equipment needs to obtain data is not effectively reduced. Therefore, the energy saving effect of such a data center system is not ideal. Summary of the invention
- the embodiments of the present invention provide a data service method and a network device, which can save energy required when acquiring data.
- an embodiment of the present invention provides a data center node, where the data center node includes: a storage unit, configured to save data; and a control unit, configured to use, according to the first time granularity statistics, each data saved by the storage unit a click rate, determining a heat level of each data according to the click rate of each data, and determining a mirror node corresponding to each heat level based on the energy saving optimization strategy; and a communication unit, configured to copy the data saved by the storage unit A mirror node corresponding to the heat level of the saved data.
- control unit is further configured to determine a hop count and a transmission delay of a transmission link of another data center node to the data center node, where the other data center node a data center node other than the data center node in the data center system where the data center node is located; the control unit is specifically configured to determine, according to the energy saving optimization strategy, the number of hops and the transmission delay of the transmission link, A mirror node corresponding to a heat level.
- control unit is specifically configured to optimize a minimum average hop count and a minimum average according to the hop count and the transmission delay of the transmission link.
- An algorithm for transmitting delays determines a mirror node corresponding to each heat level.
- the communication unit is further configured to use a hop count of a transmission link of the first user equipment to the data center node, The transmission delay of the first user equipment to the data center node and the load information of the data center node are sent to the first user equipment, where the first user equipment is a user equipment that requests data saved by the data center node.
- an embodiment of the present invention provides a regional network node, where the regional network node includes: a storage unit, configured to save data; a control unit, configured to determine a priority of each target data, and based on the target data Priority of updating the data held by the storage unit, wherein the target data is data requested by the user equipment and stored in the at least one data center node and not stored in the storage unit.
- control unit is specifically configured to determine a heat level of each target data and a hop count of a transmission link of each target data, according to each target The heat level of the data and the number of hops of the transmission link of the target data determine the priority of each of the target data.
- the each target data is saved in at least one data center based on an algorithm for optimizing a minimum average hop count and a minimum average transmission delay node.
- control unit is further configured to determine a priority of data held by the storage unit, where the control unit is specifically configured to be based on The priority of each target data, the remaining space of the storage unit, and the priority of the data held by the storage unit, update the data held by the storage unit.
- an embodiment of the present invention provides a method for providing a data service, where the method is performed by a data center node, where the method includes: counting, according to a first time granularity, a click rate of each data saved by the data center node; Determining the heat level of each data according to the click rate of each data; determining a mirror node corresponding to each heat level based on the energy saving optimization strategy; copying the data saved by the data center node to a heat level corresponding to the saved data Mirror node.
- the method further includes: determining a hop count and a transmission delay of a transmission link of another data center node to the data center node, where the other data center node is A data center node of the data center system in which the data center node is located, except the data center node corresponding to the data center node, where the determining the mirror node corresponding to each heat level includes: determining the number according to the hop count and the transmission delay of the transmission link A mirror node corresponding to each heat level.
- the energy-saving optimization policy determines, according to the hop count and the transmission delay of the transmission link, the mirror node corresponding to each heat level, including According to the hop count and transmission delay of the transmission link, an algorithm for optimizing the minimum average hop count and the minimum average transmission delay is used to determine the mirror node corresponding to each heat level.
- the method further includes: a hop count of the transmission link of the first user equipment to the data center node, the first The transmission delay of the user equipment to the data center node and the load information of the data center node are sent to the first user equipment, where the first user equipment is a user equipment that requests data saved by the data center node.
- an embodiment of the present invention provides a method for providing a data service, where the method is performed by a regional network node, where the method includes: determining a priority of each target data, where the target data is saved by the user equipment request at least a data center node and no data stored in the storage unit; updating data held by the area network node based on the priority of each of the target data.
- the method further includes: determining a heat level of each target data and a hop count of a transmission link of each target data, where the determining each target
- the priority of the data includes: determining a priority of each of the target data according to the heat level of each target data and the number of hops of the transmission link of each target data.
- the each target data is saved in at least one data center based on an algorithm for optimizing a minimum average hop count and a minimum average transmission delay node.
- the method further includes: determining a priority of data held by the network node of the area, where the target data is based on the target data Priority, updating data saved by the network node of the area, including: updating the area network node based on the priority of each target data, the remaining space of the network node of the area, and the priority of data held by the network node of the area The data.
- the user equipment can acquire data from any of the plurality of data center nodes. Furthermore, since these mirror nodes are determined according to the minimum average hop count and the minimum average transmission delay, it is possible to improve the efficiency with which the user equipment acquires data on these mirror nodes and reduce the power consumption when acquiring data.
- the regional network node can update the data stored in the cache device in real time according to the data requested by the user equipment, so that the saved data is the number of the user equipment with a high click rate. According to. The user equipment can quickly obtain the required data from the regional network node without acquiring data from the data center node. In this way, the speed at which the user device can acquire data can be accelerated, and the user experience is improved.
- FIG. 1 is a schematic diagram of a data center system provided in accordance with an embodiment of the present invention.
- FIG. 2 is a structural block diagram of a data center node according to an embodiment of the present invention.
- FIG. 3 is a structural block diagram of a regional network node according to an embodiment of the present invention.
- FIG. 4 is a structural block diagram of a network device according to an embodiment of the present invention.
- FIG. 5 is a structural block diagram of a network device according to an embodiment of the present invention.
- FIG. 6 is a schematic flowchart of a method for providing a data service according to an embodiment of the present invention.
- FIG. 7 is a schematic flowchart of a method for providing a data service according to an embodiment of the present invention.
- FIG. 8 is a schematic flowchart of a method for providing a data service according to an embodiment of the present invention.
- the embodiment of the present invention provides a data center system, where the data center system includes N data center nodes and M regional network nodes, wherein the regional network node has a cache device, and N and M are positive integers.
- the data center node and the regional network node may be servers of different functions, wherein the data center node may be a server located in a core network, and the regional network node may be located in a metropolitan area. The server in the network.
- a first data center node configured to calculate, according to a first time granularity, a click rate of each data saved by the first data center node, and determine a heat level of each data according to a click rate of each data, based on energy saving optimization Measuring, determining a mirror node corresponding to each heat level, the first number Copying data saved by the central node to the mirror node of the number of heat levels of the saved data, wherein the first data center node is any one of the N data center nodes, and the selection of the mirror node may be A data center node other than the first data center node among the N data center nodes.
- each of the N data center nodes can calculate the click rate of each data saved by itself according to the first time granularity, and determine the heat of each data according to the click rate of each data.
- Level and based on the energy saving optimization strategy, the algorithm for optimizing the minimum average hop count and the minimum average transmission delay is used to determine the mirror node corresponding to each heat level, and the saved data is copied to the number corresponding to the heat level of the data.
- the user equipment can retrieve data from any of the plurality of data center nodes.
- these mirror nodes are determined based on the minimum average hop count and the minimum average transmission delay, the efficiency with which the user equipment acquires data on these mirror nodes can be improved.
- a first regional network node configured to determine a priority of each target data, and based on a priority of each target data, update data stored in a cache device of the first regional network node, where the target data is a user equipment request There is no saved data stored in the at least one data center node and the cache device of the first regional network node, and the first regional network node is any one of the M regional network nodes.
- each target data is stored in at least one data center node based on an algorithm for optimizing the minimum average hop count and the minimum average transmission delay.
- the each target data is stored in at least one data center node, the at least one data center node being determined based on an algorithm for the user's minimum average hop count and minimum average transmission delay.
- each of the M regional network nodes can update the data stored in the cache device in real time according to the data requested by the user equipment, so that the data stored by the cache device of the regional network node is saved. It is data with high user device click-through rate.
- the user equipment can quickly obtain the required data from the regional network node without obtaining data from the data center node. In this way, the speed at which the user device can acquire data can be accelerated, and the user experience is improved.
- the data center system provided by the embodiment of the present invention enables the user equipment to quickly acquire high-heat data (that is, data with high click-through rate) from the regional network node.
- the user equipment can also obtain the required data from multiple data center nodes, and the efficiency of obtaining the required data is high. Therefore, the data center system provided by the embodiment of the present invention can save user equipment to acquire data. The energy required when it is needed.
- the user equipment may send a data request to the regional network node connected to the user equipment, where the data request is used to request to acquire data saved by the data center network.
- the regional network node will The data request sent by the device is forwarded to all data center nodes of all data center nodes in the data center node list, wherein the data center node list is saved in the regional network node.
- each data center node can send data corresponding to the data request to the user equipment according to the data request of the user equipment.
- the first data center node may calculate the click rate of the data requested by the user equipment based on the first time granularity, and then divide the data into different heat levels according to the click rate, and use the heat level to reflect the heat of the data.
- the first data center node may also determine a mirror node corresponding to each heat level based on the energy saving optimization policy. Different heat levels can correspond to different numbers of mirror nodes. The mirror nodes corresponding to the high heat level are more than the mirror nodes corresponding to the low heat level. In this way, the user equipment can have more ways to obtain data of high heat level.
- the security of the data of the high heat level is also improved due to the increase in the backup of the data of the high heat level.
- the energy consumption required by the user equipment to acquire data from the mirror nodes determined based on the energy saving optimization policy may be lower and more efficient than the randomly determined mirror nodes.
- the first data center node can also be used to receive and save data sent by other data center nodes in the data center system.
- the first data center node holds four data of Al, A2, A3, and A4.
- the first data center node can count the click rate of the four data within 2 hours. For example, within 2 hours, the first data center received 220 data requests for requesting A1, 250 requests for requesting A2, 1000 requests for requesting A3, and 2000 requests for requesting A4. .
- the four data rates of Al, A2, A3, and A4 are 200, 250, 500, and 2000 in 2 hours. If the hit rate for data within 0 hours is 0 to 199, the heat level is 1 for the data rate, the heat rating for data with a click rate of 201 to 500 is 2, and the heat rating for data with a click rate of 501 to 1000 is 3, and the click rate is 1001.
- the above data has a heat rating of 4, then A1 and A2 have a heat rating of 2, A3 has a heat rating of 3, and A4 has a heat rating of 4.
- the heat level of the data stored in the first data center can be reflected by the heat level.
- the first data center determines different mirror nodes according to the heat level, and copies corresponding data to the mirror nodes.
- the first data center may determine that all data center nodes in the data center system correspond to mirror nodes with a heat level of 4, and one-half of the data center nodes in the data center system correspond to a heat level of 3
- the mirror node, one third of the data center nodes in the data center system corresponds to a mirror node with a heat level of 2, and the heat level of 1 does not set a mirror node. That is, the first data center node can copy ⁇ 4 to all data center nodes in the data center system, and copy A3 to one-half of the data center nodes in the data center system, and A1 and ⁇ 2 Copy to one-third of the data center nodes in the data center system. If the first data center node also holds data ⁇ 5 and ⁇ 5 has a click rate of 50 within two hours, the first data center node does not copy ⁇ 5 to any other data center node.
- first time granularity and the relationship between the hit rate and the heat level.
- the relationship between the first time granularity, the click rate, and the heat level can be set as needed.
- the first time granularity can be longer or shorter.
- the range of click and heat levels can also be changed.
- a first data center node specifically configured to perform a hop (English: hop) number and transmission according to a transmission link of the data center node other than the first data center node to the first data center node of the data center node Delay, the algorithm for optimizing the minimum average hop count and the minimum average transmission delay is used to determine the mirror node corresponding to each heat level.
- a hop English: hop
- the algorithm for optimizing the minimum average hop count and the minimum average transmission delay may be a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the first data center node is specifically configured to determine a mirror node corresponding to each heat level by using a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the objective equation of the linear programming algorithm is aimed at energy saving, enabling the user equipment to efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during data transmission are optimized, so as to achieve energy saving and high efficiency. That is to say, the limit equation of the linear programming algorithm mainly includes the flow transfer equation and the limitation of the number of mirror nodes.
- the minimum average hop count and the minimum average transmission delay during data transmission are based on the hop count of the transmission link of the data center node other than the first data center node to the first data center node among the N data center nodes The average of the average and the transmission delay.
- the algorithm for optimizing the minimum average hop count and the minimum average transmission delay may be a genetic algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the first data center node is specifically used to optimize the minimum average hop count and A genetic algorithm with a minimum average transmission delay determines the mirror nodes corresponding to each heat level.
- the moderate calculation function of the genetic algorithm is determined by the minimum average hop count and the minimum average transmission delay of the data transmission.
- the genetic algorithm aims to save energy, enabling the user equipment to efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during data transmission are based on the hop count of the transmission link of the data center node other than the first data center node to the first data center node among the N data center nodes The average of the average and the transmission delay.
- the first data center node is further configured to use a hop count of the first user equipment to the first data center, a transmission delay of the first user equipment to the first data center, and the first data center.
- the load information of the node is sent to the first user equipment, where the first user equipment is a user equipment that requests data saved by the first data center node.
- the user equipment sends a data request to the regional network node in the data center system. If the data requested by the data request is not in the cache device of the regional network node that receives the data request, the regional network node forwards the data request sent by the user equipment to all data center nodes in the data center node list.
- the first data center node determines the hop count of the first user equipment to the first data center node transmission link, and the first user equipment to the first The transmission delay of the data center node and the load of the first data center node. Then, the first data center sends the hop count of the transmission link, the transmission delay, and the load information of the first data center to the first user equipment.
- the other data center nodes in the data center system may also receive the data request sent by the first user equipment and send the hop count and transmission of the transmission link to the first user equipment to the first user equipment. Delay and their respective load information.
- the first user equipment can receive the hop count, transmission delay, and load information of the transmission link sent by the multiple data center nodes.
- the first user equipment can select an appropriate data center node for data transmission according to the hop count, transmission delay, and load information of the transmission link. For example, the following formula can be used to determine the level of a data center node:
- L(i) hop (i) * a% + delay (i) * b% + workload (i) * c% , Equation 1.1
- L(i) represents the rank of the i-th data center node
- hop(i The number of hops of the transmission link of the i-th data center node to the first user equipment
- delay(i) represents the transmission delay of the i-th data center node to the first user equipment
- workload(i) indicates The load of i data center nodes, A%, b%, and c% represent the hop count of the transmission link, the transmission delay, and the weight of the load occupying the level of the data center node, respectively.
- a%, /?%, and ⁇ "% can be designed as needed. For example, if the number of hops of the transmission link is expected to have the greatest impact on the level of the data center node, the value of a% can be made. Greater than % and c% If you want the transmission delay to have the greatest impact on the level of the data center node, you can make the value of % greater than % and ⁇ >%.
- a first regional network node specifically configured to determine a heat level of each target data and a hop count of a transmission link of each target data, according to a heat level of each target data and a hop of a transmission link of each target data Number, determines the priority of each target data.
- the first area network node determines the hop count of the transmission link of each of the data center nodes storing the target data to the first area network node. Then, the priority of each target data is determined based on the heat level of each target data and the number of hops of the transmission link of each target data. For example, the following formula can be used to prioritize target data:
- P(i) pop (i) * x% + hop(i) * y% , Equation 1.2
- P(i) represents the priority of the i-th target data
- pop(i) represents the i-th target data
- the heat level, hop(i) represents the hop count of the transmission link of the i-th target data
- y% represents the hop count of the transmission link of the target data.
- the weight of the priority of the data The weight of the priority of the data.
- _1% and ⁇ can be designed as needed. For example, if you want to make the priority of the heat level greater than the priority of the number of hops of the transmission link, you can make it greater than ⁇ . . If it is desired that the weight of the priority of the heat level is less than the weight of the priority of the hop of the transmission link, it may be) ⁇ is greater than x%.
- the first area network node After determining the priority of each target data, the first area network node ranks the target data according to the priority, and sequentially determines whether the target data needs to be stored in the cache device of the first area network node.
- the first area network node determines whether there is enough remaining space in the cache device to store the first target data. If the cache device has enough free space to store the first target data, the first target data is stored in the cache device of the first regional network node.
- the first regional network node is further configured to determine a priority of each data in the cache device, according to a priority of each target data, The priority of each data in the cache device and the remaining space of the cache device update the data in the cache device. Specifically, the first regional network node first determines The lowest priority data in the cache device (for the sake of description, the following cartridge is referred to as the first cache data). If the priority of the first cached data is higher than the priority of the first target data, the first target data is not stored in the cache device.
- the priority of the first cached data is lower than the priority of the first target data, determining whether the remaining space in the cache device is greater than the storage occupied by the first target data after deleting the priority of the first cached data space. If the remaining space after deleting the first cached data is greater than the space occupied by the first target data, the first cached data is deleted and the first target data is stored in the cache device. Similarly, the target data of the first regional network node may be continuously compared with the data saved by the cache device of the first regional network node, so that the data with high priority is saved to the cache of the first regional network node. In the device. In this way, the data stored in the cache device of the first regional network node can be made to be high priority data.
- the data held in the cache device of the first regional network node is the data that is often requested by the user equipment.
- the user equipment can quickly obtain the required data from the first regional network node without having to acquire data from the data center node. In this way, the speed at which the user device can acquire data can be accelerated, and the energy consumption required for acquiring data can be saved.
- the first regional network node determines three target data, which are D1, D2, and D3 in descending order of priority.
- the data of the first regional network node stores three data of dl, d2 and d3. If the remaining space of the cache device of the first regional network node is capable of storing D1, D1 is saved to the cache device of the first regional network node. At this time, the data held in the cache device of the first area network node is dl, d2, d3, and D1. If there is not enough storage space D1 in the cache device of the first regional network node, the priority of the three data stored in the cache device of the first regional network node is determined.
- the priority of d3 and D1 is determined. 4 If the priority of d3 is greater than the priority of D1, the data in the cache device of the network node of the first area is kept unchanged. Assuming that the priority of d3 is less than the priority of D1, it is determined whether the remaining space in the cache device of the first regional network node after the deletion of d3 can save D1. If it is determined that the remaining space after deleting d3 can hold D1, then d3 is deleted and D1 is saved to the cache device of the first regional network node.
- the data held in the cache device of the first regional network node is dl, d2, and D1.
- the first regional network node compares the other target data with the data stored in the cache device to update the data in the cache device, so that the data stored in the cache device is the data with higher priority.
- FIG. 1 is a schematic diagram of a data center system provided in accordance with an embodiment of the present invention.
- the number shown in Figure 1 According to the central system, there are three data center nodes and four regional network nodes.
- FIG. 1 is only a specific example of a data center system. This example is only intended to assist those skilled in the art to better understand the embodiments of the present invention and not to limit the scope of the embodiments of the present invention.
- the data center node shown in Figure 2 is a structural block diagram of a data center node according to an embodiment of the present invention.
- the data center node shown in Figure 2 may be any one of the N data center nodes in the data center system provided by the embodiment of the present invention.
- the data center node shown in Figure 2 can be a network device located on the core network, such as a server.
- the data center node 200 includes a storage unit 201, a control unit 202, and a communication unit 203.
- the storage unit 201 is configured to save data.
- the control unit 202 is configured to use a click rate of each data saved by the storage unit 201 based on the first time granularity statistics.
- the control unit 202 is further configured to determine a heat level of each data according to the click rate of each data.
- the control unit 202 is further configured to determine a mirror node corresponding to each heat level based on the energy saving optimization strategy.
- the communication unit 203 is configured to copy the data saved by the storage unit 201 to a mirror node corresponding to the heat level of the saved data.
- the data center node 200 shown in FIG. 2 can determine the click rate of each data saved by itself according to the first time granularity, determine the heat level of each data according to the click rate of each data, and adopt the optimized minimum average.
- the algorithm of hop count and minimum average transmission delay determines the mirror node corresponding to each heat level, and copies the saved data to other data center nodes corresponding to the heat level of the data.
- the user equipment can retrieve data from any of the plurality of data center nodes.
- these mirror nodes are determined according to the minimum average hop count and the minimum average transmission delay, the efficiency of the user equipment to acquire data on these mirror nodes can be improved and the power consumption when acquiring data can be reduced.
- the communication unit 203 is further configured to send the hop count of the first user equipment to the data center node 200, the transmission delay of the first user equipment to the data center node 200, and the load information of the data center node 200.
- the first user equipment is provided to the user equipment, wherein the first user equipment is a user equipment that requests data held by the data center node 200.
- the user equipment sends data to the regional network node in the data center system. begging. If the data requested by the data request is not in the cache device of the regional network node that receives the data request, the regional network node forwards the data request sent by the user equipment to all data center nodes in the data center node list. A data center node, wherein the data center node list is stored in the regional network node. For the sake of description, if the data requested by a user equipment is not in the cache device of the regional network node, the user equipment is referred to as the first user equipment.
- the control unit 202 may be configured to determine the hop count of the transmission link of the first user equipment to the data center node 200, the first user equipment to the data. The transmission delay of the central node 200 and the load of the data center node 200. Then, the communication unit 203 may be configured to send the hop count of the transmission link, the transmission delay, and the load information of the data center node 200 to the first user equipment, so that the first user equipment according to the transmission The number of hops of the link, the transmission delay, and the load information of the data center node 200 determine whether data is obtained from the data center node 200.
- the control unit 202 may be configured to count the click rate of the data requested by the user equipment based on the first time granularity, and then divide the data stored by the storage unit 201 into different heat levels according to the click rate, and reflect the heat of the data by using the heat level.
- the control unit 202 is further configured to determine, according to the energy saving optimization strategy, a mirror node corresponding to each heat level. Different heat levels can correspond to different mirror nodes.
- the control unit 202 when determining the mirroring node, may determine different numbers of mirror nodes for different heat levels based on the energy saving optimization strategy. For example, there are more mirror nodes corresponding to the high heat level than mirror nodes corresponding to the low heat level. In this way, the user equipment can have more ways to obtain data of high heat level.
- the security of the data of the high heat level is also improved due to the increase in the backup of the data of the high heat level.
- control unit 202 is further configured to determine a hop count and a transmission delay of a transmission link of the other data center node to the data center node 200, where the other data center node is a data center system in which the data center node 200 is located. Other data center nodes other than the central node 200.
- the control unit 202 is specifically configured to determine, according to the hop count and the transmission delay of the transmission link, the mirror node corresponding to each heat level.
- control unit 202 is specifically configured to determine, according to the hop count and the transmission delay of the transmission link, an algorithm for optimizing a minimum average hop count and a minimum average transmission delay, and determining a mirror node corresponding to each heat level. .
- control unit 202 is specifically configured to determine a mirror corresponding to each heat level by using a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the target equation of the linear programming algorithm is aimed at energy saving, so that the user equipment can efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during the data transmission process are optimized, thereby achieving the purpose of energy saving and high efficiency.
- the limiting equation of the linear programming algorithm mainly includes the limitation of the flow transfer equation and the number of mirror nodes.
- the minimum average hop count and the minimum average transmission delay during data transmission are average values based on the hop counts of the transmission links of the data center nodes other than the data center node 200 to the data center node 200 among the N data center nodes. The average of the transmission delays.
- control unit 202 is specifically configured to determine a mirror node corresponding to each heat level by using a genetic algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the modest calculation function of the genetic algorithm is determined by the minimum average hop count and the minimum average transmission delay of the data transmission.
- the genetic algorithm aims to save energy, enabling the user equipment to efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during data transmission are average values based on the hop counts of the transmission links of the data center nodes other than the data center node 200 to the data center node 200 among the N data center nodes. The average of the transmission delays.
- FIG. 3 is a structural block diagram of a regional network node according to an embodiment of the present invention.
- the regional network node shown in FIG. 3 may be any one of the M regional network nodes in the data center system provided by the embodiment of the present invention.
- the regional network node shown in Figure 3 may be a network device located in a metropolitan area network, such as a server.
- the regional network node 300 includes a storage unit 301 and a control unit 302.
- the storage unit 301 is configured to save data.
- the control unit 302 is configured to determine a priority of each target data, where the target data is data that is saved by the user equipment and stored in the at least one data center node and is not saved in the storage unit 301.
- the control unit 302 is further configured to update the storage unit based on the priority of each target data.
- the area network node 300 shown in FIG. 3 can update the data stored in its own cache device in real time according to the data requested by the user equipment, so that the data held by the storage unit 301 is data with high user equipment click rate.
- the user equipment can quickly obtain the required data from the regional network node without acquiring data from the data center node. In this way, the speed at which the user device can acquire data can be accelerated, and the user experience is improved.
- the control unit 302 is specifically configured to determine a heat level of each target data and a hop count of a transmission link of each target data, according to a heat level of each target data and each target data The number of hops of the transmission link, and the priority of each of the target data is determined.
- control unit 302 is configured to determine the hop count of the transmission link of each data center node storing the target data to the regional network node 300. Then, the control unit 302 determines the priority of each target data based on the heat level of each target data and the hop count of the transmission link of each target data. For example, the control unit 302 can determine the priority of the target data by using the following formula:
- P (i) pop (i) ⁇ xo + hop(i) ⁇ yo , Equation 1.3
- P(i) represents the priority of the i-th target data
- pop(i) represents the heat level of the i-th target data
- Hop(i) represents the hop count of the transmission link of the i-th target data
- y% represents the hop count of the transmission link of the target data as the target data.
- the weight of the priority Those skilled in the art will appreciate that _1% and ⁇ can be designed as needed. For example, if you want the priority of the heat level to be greater than the priority of the number of hops on the transmission link, you can make it greater than y%. If it is desired that the weight of the priority of the heat level is less than the weight of the priority of the hop of the transmission link, it may be) ⁇ is greater than x%.
- control unit 302 is further configured to determine the priority of the data held by the storage unit 301.
- the control unit 302 is specifically configured to update the data held by the storage unit 301 based on the priority of each of the target data, the remaining space of the storage unit 301, and the priority of the data held by the storage unit 301.
- control unit 302 after determining the priority of each target data, arranges the target data according to the priority, and sequentially determines whether the target data needs to be stored in the storage unit 301.
- the control unit 302 determines whether there is enough remaining space in the storage unit 301 to store the first target data. If the storage unit 301 has enough free space to store the first target data, the control unit 302 stores the first target data in the storage unit 301.
- the control unit 302 is further configured to determine a priority of each data in the storage unit 301, according to the priority of each target data, the storage unit 301.
- the data in the storage unit 301 is updated by the priority of each of the data and the remaining space of the storage unit 301.
- the control unit 302 first determines the lowest priority data in the storage unit 301 (for the sake of description, the following cartridge is referred to as the first cache number. According to). If the priority of the first cache data is higher than the priority of the first target data, the control unit 302 does not store the first target data into the storage unit 301.
- the control unit 302 determines whether the remaining space in the storage unit 301 after the priority of the first cached data is deleted is greater than the first target data. Occupied storage space. If the remaining space after deleting the first cache data is larger than the space occupied by the first target data, the control unit 302 deletes the first cache data and stores the first target data into the storage unit 301.
- the target data is stored in at least one data center node based on an algorithm for a user minimum average hop count and a minimum average transmission delay.
- FIG. 4 is a structural block diagram of a network device according to an embodiment of the present invention.
- the network device shown in Figure 4 may be any one of the N data center nodes in the data center system provided by the embodiment of the present invention.
- the network device shown in Figure 4 can be a network device located on the core network, such as a server.
- network device 400 includes: memory 401, processor 402, and transceiver 403.
- the memory 401 is configured to save data.
- the processor 402 is configured to calculate a click rate of each data stored by the memory 401 based on the first time granularity statistics.
- the processor 402 is further configured to determine a heat level of each data according to the click rate of each data.
- the processor 402 is further configured to determine, according to the energy saving optimization strategy, a mirror node corresponding to each heat level.
- the transceiver 403 is configured to copy the data saved by the memory 401 to a mirror node corresponding to the heat level of the saved data.
- the network device 400 shown in FIG. 4 can determine the click rate of each data saved by itself according to the first time granularity, determine the heat level of each data according to the click rate of each data, and adopt the optimized minimum average jump.
- the algorithm for the number and minimum average transmission delay determines the mirror node corresponding to each heat level and copies the saved data to other data center nodes corresponding to the heat level of the data.
- the user equipment can retrieve data from any of the plurality of data center nodes.
- these mirror nodes are determined according to the minimum average hop count and the minimum average transmission delay, the efficiency of the user equipment to acquire data on these mirror nodes can be improved and the power consumption when acquiring data can be reduced.
- the transceiver 403 is further configured to send, to the first, the hop count of the first user equipment to the network device 400, the transmission delay of the first user equipment to the network device 400, and the load information of the network device 400.
- the user equipment sends a data request to the regional network node in the data center system. If the data requested by the data request is not in the cache device of the regional network node that receives the data request, the regional network node forwards the data request sent by the user equipment to all data center nodes in the data center node list.
- the processor 402 may be configured to determine the hop count of the transmission link of the first user equipment to the network device 400, and the first user equipment to the network device 400. The transmission delay and the load of the network device 400.
- the transceiver 403 may be configured to send the hop count of the transmission link, the transmission delay, and load information of the network device 400 to the first user equipment, so that the first user equipment is configured according to the transmission chain.
- the hop count of the road, the transmission delay, and the load information of the network device 400 determine whether data is acquired from the network device 400.
- the processor 402 is configured to calculate a click rate of the data requested by the user equipment based on the first time granularity, and then divide the data stored in the memory 401 into different heat levels according to the click rate, and reflect the heat of the data by using the heat level.
- the processor 402 is further configured to determine, according to the energy saving optimization strategy, a mirror node corresponding to each heat level. Different heat levels can correspond to different mirror nodes. Specifically, the processor 402 can determine a different number of mirror nodes for different heat levels when determining the mirror node. For example, there are more mirror nodes corresponding to the high heat level than mirror nodes corresponding to the low heat level. In this way, the user equipment can have more ways to obtain data of high heat level. At the same time, the security of the data of the high heat level is also improved due to the increase in the backup of the data of the high heat level.
- the processor 402 is further configured to determine a hop count and a transmission delay of a transmission link of another data center node to the network device 400, where the other data center node is a network center device in which the network device 400 is located. Other data center nodes than .
- the processor 402 is specifically configured to determine, according to the hop count and the transmission delay of the transmission link, the mirror node corresponding to each heat level.
- the processor 402 is specifically configured to determine, according to the hop count and the transmission delay of the transmission link, an algorithm for optimizing a minimum average hop count and a minimum average transmission delay, and determining a mirror node corresponding to each heat level. .
- the processor 402 is specifically configured to determine a mirror node corresponding to each heat level by using a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the objective equation of the linear programming algorithm is aimed at energy saving, enabling the user equipment to efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during data transmission are optimized, so as to achieve energy saving and high efficiency. That is to say, the limiting equation of the linear programming algorithm mainly includes the flow transfer equation and the limitation of the number of mirror nodes.
- the minimum average hop count and the minimum average transmission delay during data transmission are average values and transmission delays of the number of hops of the transmission link from the data center node other than the network device 400 to the network device 400 among the N data center nodes. The average value of the time.
- the processor 402 is specifically configured to determine a mirror node corresponding to each heat level by using a genetic algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the modest calculation function of the genetic algorithm is determined by the minimum average hop count and the minimum average transmission delay of the data transmission.
- the genetic algorithm aims to save energy, enabling the user equipment to efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during data transmission are average values and transmission delays of the number of hops of the transmission link from the data center node other than the network device 400 to the network device 400 among the N data center nodes. The average value of the time.
- FIG. 5 is a structural block diagram of a network device according to an embodiment of the present invention.
- the network device shown in FIG. 5 may be any one of the M regional network nodes in the data center system provided by the embodiment of the present invention.
- the network device shown in Figure 5 may be a network device located in a metropolitan area network, such as a server.
- the network device 500 includes: a memory 501 and a processor 502.
- the memory 501 is configured to save data.
- the processor 502 is configured to determine a priority of each target data, where the target data is data that is saved by the user equipment and stored in the at least one data center node and not stored in the memory 501.
- the processor 502 is further configured to update the data held by the memory 501 based on the priority of each of the target data.
- the network device 500 shown in FIG. 5 can update the data stored in its own cache device in real time according to the data requested by the user equipment, so that the data held by the memory 501 is the user equipment point. High hit rate data.
- the user equipment can quickly obtain the required data from the cache device of the regional network node without acquiring data from the data center node. In this way, the speed at which the user device can acquire data can be accelerated, and the user experience is improved.
- the processor 502 is specifically configured to determine a heat level of each target data and a hop count of a transmission link of each target data, according to a heat level of each target data and each target data The number of hops of the transmission link, and the priority of each of the target data is determined.
- the processor 502 is configured to determine a hop count of a transmission link of each data center node storing the target data to the network device 500. Then, the processor 502 determines the priority of each target data according to the heat level of each target data and the hop count of the transmission link of each target data. For example, the processor 502 can determine the priority of the target data using the following formula:
- the processor 502 is further configured to determine a priority of data held by the memory 501.
- the processor 502 is specifically configured to update the data held by the memory 501 based on the priority of each of the target data, the remaining space of the memory 501, and the priority of the data held by the memory 501.
- the processor 502 sorts the target data according to the priority, and sequentially determines whether the target data needs to be stored in the memory 501.
- the processor 502 determines whether there is enough remaining space in the memory 501 to store the first target data. If the memory 501 has enough free space to store the first target data, the processor 502 stores the first target data in the memory 501.
- the processor 502 is further configured to determine a priority of each data in the memory 501, according to the priority of each target data, each in the memory 501. The priority of the data and the remaining space of the memory 501, The data in the new memory 501. Specifically, the processor 502 first determines the lowest priority data in the memory 501 (the following cartridge is referred to as the first cache data for the sake of description). If the priority of the first cached data is higher than the priority of the first target data, the processor 502 does not store the first target data in the memory 501.
- the processor 502 determines whether the remaining space in the memory 501 after the priority of the first cached data is deleted is greater than the occupied by the first target data. Storage space. If the remaining space after deleting the first cache data is greater than the space occupied by the first target data, the processor 502 deletes the first cache data and stores the first target data into the memory 501.
- the target data is stored in at least one data center node based on an algorithm for a user minimum average hop count and a minimum average transmission delay.
- FIG. 6 is a schematic flowchart of a method for providing a data service according to an embodiment of the present invention.
- the method shown in Figure 6 can be performed by a data center node.
- the data center node that performs the method shown in Figure 6 can be the data center node shown in Figure 3.
- Count according to the first time granularity, a click rate of each data saved by the data center node. 602. Determine, according to the click rate of each data, a heat level of each data.
- the data center node can determine the click rate of each data saved by itself according to the first time granularity, determine the heat level of each data according to the click rate of each data, and adopt the optimization for use.
- the algorithm of minimum average hop count and minimum average transmission delay determines the mirror node corresponding to each heat level, and copies the saved data into other data center nodes corresponding to the heat level of the data.
- the user equipment can retrieve data from any of the plurality of data center nodes.
- the mirror nodes are determined according to the minimum average hop count and the minimum average transmission delay, the efficiency of the user equipment to acquire data on the mirror nodes can be improved and the energy consumption of the user equipment to obtain data can be reduced.
- the method shown in FIG. 6 further includes: a hop count of a transmission link of the first user equipment to the data center node, a transmission delay of the first user equipment to the data center node, and a load of the data center node
- the information is sent to the first user equipment, where the first user equipment is a user equipment requesting data held by the data center node.
- the user equipment sends a data request to the regional network node in the data center system. If the data requested by the data request is not in the cache device of the regional network node that receives the data request, the regional network node forwards the data request sent by the user equipment to all data center nodes in the data center node list.
- a data center node wherein the data center node list is stored in the regional network node.
- the user equipment is referred to as the first user equipment.
- the data center node receives the data request, the data center node determines the hop count of the transmission link of the first user equipment to the data center node, and the transmission delay of the first user equipment to the data center node. And the load of the data center node. Then, the data center node sends the hop count of the transmission link, the transmission delay, and the load information of the data center node to the first user equipment, so that the first user equipment hops according to the transmission link.
- the number, the transmission delay, and the load information of the data center node determine whether data is obtained from the data center node.
- the data center node may count the click rate of the data requested by the user equipment based on the first time granularity. The data center node then divides the stored data into different heat levels according to the click rate, and uses the heat level to reflect the heat of the data. The data center node determines a mirror node corresponding to each heat level based on the energy saving optimization strategy. Different heat levels can correspond to different mirror nodes. Specifically, the data center node may determine a different number of mirror nodes for different heat levels when determining the mirror node. For example, there are more mirror nodes corresponding to the high heat level than mirror nodes corresponding to the low heat level. In this way, the user equipment can have more ways to obtain data of high heat level. At the same time, the security of the data of the high heat level is also improved due to the increase in the backup of the data of the high heat level.
- the method shown in FIG. 6 further includes: determining a hop count and a transmission delay of a transmission link of another data center node to the data center node, wherein the other data center node is in a data center system where the data center node is located A data center node other than the data center node.
- the determining the mirroring node corresponding to each heat level based on the energy saving optimization strategy includes: determining, according to the hop count and the transmission delay of the transmission link, the mirror node corresponding to each heat level.
- determining the mirror node corresponding to each heat level according to the hop count and the transmission delay of the transmission link including: optimizing the minimum average hop count according to the hop count and the transmission delay of the transmission link.
- the algorithm of the minimum average transmission delay determining the mirror node corresponding to each heat level.
- the algorithm for optimizing the minimum average hop count and the minimum average transmission delay may be a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the first data center node may determine a mirror node corresponding to each heat level by using a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the target equation of the linear programming algorithm is aimed at energy saving, enabling the user equipment to efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during the data transmission process are optimized, thereby achieving the purpose of energy saving and high efficiency.
- the limiting equation of the linear programming algorithm mainly includes the limitation of the flow transfer equation and the number of mirror nodes.
- the minimum average hop count and the minimum average transmission delay during data transmission are based on the hop count of the transmission link of the data center node other than the first data center node to the first data center node among the N data center nodes The average of the average and the transmission delay.
- the algorithm for optimizing the minimum average hop count and the minimum average transmission delay may be a genetic algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the first data center node may use a genetic algorithm for optimizing the minimum average hop count and the minimum average transmission delay to determine the mirror node corresponding to each heat level.
- the modest calculation function in the genetic algorithm is determined by the minimum average hop count and the minimum average transmission delay of the data transmission.
- the genetic algorithm aims to save energy, enabling the user equipment to efficiently obtain the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during data transmission are based on the hop count of the transmission link of the data center node other than the first data center node to the first data center node among the N data center nodes The average of the average and the transmission delay.
- FIG. 7 is a schematic flowchart of a method for providing a data service according to an embodiment of the present invention.
- the method illustrated in Figure 7 can be performed by a regional network node that includes a cache device for storing data.
- the regional network node performing the method shown in Fig. 7 may be the regional network shown in Fig. 6 to the node.
- Target data is data that is requested by the user equipment and is saved in at least one data center node and is not saved in the network node of the area.
- the method shown in FIG. 7 can update the data saved in the cache device in real time according to the data requested by the user equipment, so that the data saved by the regional network node is data with high user equipment click rate.
- the user equipment can quickly obtain the required data from the regional network node without acquiring data from the data center node. In this way, the speed at which the user device can obtain data can be accelerated, and the user can be improved. Experience.
- the target data is stored in at least one data center node based on an algorithm for a user minimum average hop count and a minimum average transmission delay. (this same as the previous question)
- the method shown in FIG. 7 may further include: determining a heat level of each of the target data and a hop count of the transmission link of the each target data.
- the determining the priority of each target data includes: determining a priority of each of the target data according to a heat level of each of the target data and a hop count of a transmission link of the each target data.
- the regional network node can determine the hop count of the transmission link of each data center node storing the target data to the first regional network node. Then, the regional network node can determine the priority of each target data based on the heat level of each target data and the hop count of the transmission link of each target data. For example, the regional network node can determine the priority of the target data using the following formula:
- the priority can be made larger than ⁇ ⁇ If the weight of the priority of the heat level is desired to be less than the weight of the hop of the transmission link, it may be) ⁇ is greater than x%.
- the method shown in FIG. 7 may further include: determining a priority of data held by the network node of the area.
- updating the data saved by the network node of the area based on the priority of each target data includes: a priority based on the target data, a remaining space of the network node of the area, and a network node save in the area The priority of the data, update the data saved by the network node in the area.
- the regional network node after determining the priority of each target data, the regional network node arranges the target data according to the priority, and sequentially determines whether the target data needs to be stored in the cache device of the regional network node.
- the regional network node determines whether there is enough remaining space in the cache device to store the first target data. If the cache device has enough free space to store the first target number According to the data, the regional network node stores the first target data in the cache device of the regional network node.
- the regional network node determines a priority of each data in the cache device, according to the priority of each target data, in the cache device The priority of each data and the remaining space of the cache device are updated to update the data in the cache device. Specifically, the regional network node first determines the lowest priority data in the cache device (for the sake of description, the following cartridge is referred to as the first cache data). If the priority of the first cached data is higher than the priority of the first target data, the regional network node does not store the first target data in the cache device.
- the regional network node determines whether the remaining space in the cache device is greater than the first target data after deleting the priority of the first cached data The storage space used. If the remaining space after deleting the first cached data is greater than the space occupied by the first target data, the first cached data is deleted and the first target data is stored in the cache device.
- Figure 8 will be described in terms of a data center system comprising data center nodes and regional network nodes.
- both the data center node and the regional network node in the data center system can independently execute their respective process steps when implementing the inventive scheme.
- FIG. 8 is a schematic flowchart of a method for providing a data service according to an embodiment of the present invention.
- the first data center node and the first regional network node in the method shown in Fig. 8 belong to the same data center system.
- the data center system may include N data center nodes and M regional network nodes, wherein each of the M regional network nodes has a cache device, and N and M are positive integers.
- the first data center node is any one of the N data center nodes.
- the first regional network node is any one of the M regional network nodes.
- the data center system may further include one or more regional networks without cache devices to the nodes.
- the first data center node calculates, according to the first time granularity, a click rate of each data saved by the first data center node, and determines a heat level of each data according to the click rate of each data, based on the energy saving optimization strategy. Determining, by using an algorithm for optimizing a minimum average hop count and a minimum average transmission delay, a mirror node corresponding to each heat level, and copying data saved by the first data center node to a mirror node corresponding to a heat level of the saved data,
- the mirror node is any one of the N data center nodes except the first data center node. 802.
- the first area network node determines a priority of each target data, and based on a priority of the each target data, updating data stored in a cache device of the first area network node, where the target data is requested by the user equipment. There is no saved data stored in at least one data center node and in the cache device of the first regional network node.
- the user equipment can quickly acquire highly hot data (i.e., data with a high click rate) from the regional network node.
- the user equipment can also obtain the required data from multiple data center nodes and obtain more efficient data. Therefore, the method provided by the embodiment of the present invention can save the energy consumption required when the user equipment acquires data.
- the user equipment may send a data request to the regional network node connected to the user equipment, where the data request is used to request to acquire data saved by the data center network.
- the regional network node will The data request sent by the device is forwarded to all data center nodes of all data center nodes in the data center node list, wherein the data center node list is saved in the regional network node.
- each data center node can send data corresponding to the data request to the user equipment according to the data request of the user equipment.
- the first data center node may calculate the click rate of the data requested by the user equipment based on the first time granularity, and then divide the data into different heat levels according to the click rate, and use the heat level to reflect the heat of the data.
- the first data center node may also determine a mirror node corresponding to each heat level based on the energy saving optimization policy. Different heat levels can correspond to different numbers of mirror nodes. The mirror nodes corresponding to the high heat level are more than the mirror nodes corresponding to the low heat level. In this way, the user equipment can have more ways to obtain data of high heat level.
- the security of the data of the high heat level is also improved due to the increase in the backup of the data of the high heat level.
- the first data center node can also receive and save data sent by other data center nodes in the data center system.
- the first data center node may adopt, according to the hop count and transmission delay of the transmission link of the data center node other than the first data center node to the first data center node of the N data center nodes.
- the burst for optimizing the minimum average hop count and the minimum average transmission delay determines the mirror node corresponding to each heat level.
- the algorithm for optimizing the minimum average hop count and the minimum average transmission delay may be a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay. Law.
- the first data center node may determine a mirror node corresponding to each heat level by using a linear programming algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the target equation of the linear programming algorithm is aimed at energy saving, enabling the user equipment to efficiently acquire the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during the data transmission process are optimized, thereby achieving the purpose of energy saving and high efficiency. That is to say, the limiting equation of the linear programming algorithm mainly includes the limitation of the flow transfer equation and the number of mirror nodes.
- the minimum average hop count and the minimum average transmission delay during data transmission are based on the hop count of the transmission link of the data center node other than the first data center node to the first data center node among the N data center nodes The average of the average and the transmission delay.
- the algorithm for optimizing the minimum average hop count and the minimum average transmission delay may be a genetic algorithm for optimizing a minimum average hop count and a minimum average transmission delay.
- the first data center node may use a genetic algorithm for optimizing the minimum average hop count and the minimum average transmission delay to determine the mirror node corresponding to each heat level.
- the modest calculation function in the genetic algorithm is determined by the minimum average hop count and the minimum average transmission delay of the data transmission.
- the genetic algorithm aims to save energy, enabling the user equipment to efficiently obtain the required data from the mirror node.
- the minimum average hop count and the minimum average transmission delay during data transmission are based on the hop count of the transmission link of the data center node other than the first data center node to the first data center node among the N data center nodes The average of the average and the transmission delay.
- the first data center node may further include a hop count of the first user equipment to the first data center, a transmission delay of the first user equipment to the first data center, and a first data center node
- the load information is sent to the first user equipment, where the first user equipment is a user equipment that requests data saved by the first data center node.
- the user equipment sends a data request to the regional network node in the data center system. If the data requested by the data request is not in the cache device of the regional network node that receives the data request, the regional network node forwards the data request sent by the user equipment to all data center nodes in the data center node list.
- the user equipment is referred to as the first user equipment.
- the first data center node determines the hop count of the first user equipment to the first data center node transmission link, and the first user equipment to the first Data center node transmission delay and the first data center node Load. Then, the first data center sends the hop count of the transmission link, the transmission delay, and the load information of the first data center to the first user equipment.
- the other data center nodes in the data center system may also receive the data request sent by the first user equipment and send the hop count and transmission of the transmission link to the first user equipment to the first user equipment. Delay and their respective load information.
- the first user equipment may receive the hop count, transmission delay, and load information of the transmission link sent by the multiple data center nodes.
- the first user equipment may select an appropriate data center node for data transmission according to the hop count, transmission delay, and load information of the transmission link. For example, you can use the following formula to determine the level of a data center node:
- L(i) hop (i) * a% + delay (i) * b9o + workload (i) * c% , Equation 1.6
- L(i) represents the rank of the i-th data center node
- hop(i) Indicates the hop count of the transmission link of the i-th data center node to the first user equipment
- delay (i) indicates the transmission delay of the i-th data center node to the first user equipment
- workload(i) indicates the i-th
- the load of the data center nodes, a%, b%, and c%, respectively, indicate the hop count of the transmission link, the transmission delay, and the weight of the load occupying the level of the data center node.
- a%, /?%, and ⁇ "% can be designed as needed. For example, if the number of hops of the transmission link is expected to have the greatest impact on the level of the data center node, the value of a% can be made. Greater than % and c% If you want the transmission delay to have the greatest impact on the level of the data center node, you can make the value of % greater than ⁇ % and c%.
- the first regional network node may determine a heat level of each target data and a hop count of a transmission link of each target data, and determine according to the heat level of each target data and the hop count of the transmission link of each target data. The priority of each target data.
- the first regional network node may determine the number of hops of the transmission link of each of the data center nodes storing the target data to the first regional network node. Then, the first area network node determines the priority of each target data based on the heat level of each target data and the number of hops of the transmission link of each target data. For example, the following formula can be used to prioritize target data:
- P(i) pop(iYx% + hop ⁇ i) * y% , Equation 1.7
- P(i) represents the priority of the i-th target data
- pop(i) represents the heat level of the i-th target data.
- Hop(i) indicates the hop count of the transmission link of the i-th target data
- y% indicates that the hop count of the transmission link of the target data occupies the priority of the target data.
- the weight of the level Those skilled in the art will appreciate that_1% and ⁇ can be designed as needed.
- the priority of the heat level may be greater than y%. If you want to prioritize the heat level
- the weight of the transmission link is smaller than the priority of the hop of the transmission link, which can be ⁇ . Greater than X%.
- the first area network node After determining the priority of each target data, the first area network node ranks the target data according to the priority, and sequentially determines whether the target data needs to be stored in the cache device of the first area network node.
- the first area network node determines whether there is enough remaining space in the cache device to store the first target data. If the cache device has enough free space to store the first target data, the first target data is stored in the cache device of the first regional network node.
- the first regional network node may further determine a priority of each data in the cache device, according to the priority of each target data, The priority of each data in the cache device and the remaining space of the cache device are updated to update the data in the cache device. Specifically, the first area network node first determines the lowest priority data in the cache device (for the sake of description, the following cartridge is referred to as the first cache data). If the priority of the first cached data is higher than the priority of the first target data, the first target data is not stored in the cache device.
- the priority of the first cached data is lower than the priority of the first target data, determining whether the remaining space in the cache device is greater than the storage occupied by the first target data after deleting the priority of the first cached data space. If the remaining space after deleting the first cache data is larger than the space occupied by the first target data, the first cache data is deleted and the first target data is stored in the cache device.
- the disclosed systems, devices, and methods may be implemented in other ways.
- the device embodiments described above are merely illustrative.
- the division of the unit is only a logical function division.
- there may be another division manner for example, multiple units or components may be combined or Can be integrated into another The system, or some features can be ignored, or not executed.
- the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
- the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solution of the embodiment.
- each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
- the functions, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium.
- the technical solution of the present invention which is essential to the prior art or part of the technical solution, may be embodied in the form of a software product stored in a storage medium, including
- the instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to perform all or part of the steps of the methods described in various embodiments of the present invention.
- the foregoing storage medium includes: a U disk, a removable hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and the like, which can store program code. .
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Transfer Between Computers (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明实施例公开了提供数据服务的方法和网络设备,包括:基于第一时间粒度统计该数据中心节点保存的每个数据的点击率;根据该每个数据的点击率,确定该每个数据的热度等级;基于节能优化策略,确定每个热度等级对应的镜像节点;基于节能优化策略,将该数据中心节点保存的数据复制到与该保存的数据的热度等级对应的镜像节点。根据本发明实施例,用户设备可以从多个数据中心节点中的任一数据中心节点来获取数据。此外,由于这些镜像节点是根据最小平均跳数和最小平均传输延时确定的,因此可以提高用户设备获取这些镜像节点上的数据的效率并降低获取数据时的能耗。
Description
提供数据服务的方法和网络设备 技术领域
本发明实施例涉及信息技术领域, 并且更具体地, 涉及提供数据服务的 方法和网络设备。 背景技术
互联网中多个为用户提供数据服务的装置通常会组成一个数据中心系 统。 这些为用户提供数据服务的装置通常称为数据中心节点。 数据中心节点 会根据所存储的数据的点击率, 将数据划分为不同热度等级, 并将不同热度 等级的数据复制到其他数据中心节点上。 这样, 用户设备可以根据功耗和传 输延时选取数据的获取源。 但是, 用户设备获取数据是需要历经的网络层数 并不会有效的减少。 因此, 这样的数据中心系统的节能效果并不理想。 发明内容
本发明实施例提供数据服务的方法和网络设备, 能够节约获取数据时需 要的能耗。
第一方面,本发明实施例提供一种数据中心节点,该数据中心节点包括: 存储单元, 用于保存数据; 控制单元, 用于基于第一时间粒度统计的该存储 单元保存的每个数据的点击率, 根据该每个数据的点击率, 确定该每个数据 的热度等级, 并且基于节能优化策略, 确定每个热度等级对应的镜像节点; 通信单元, 用于将该存储单元保存的数据复制到与该保存的数据的热度等级 对应的镜像节点。
结合第一方面, 在第一种可能的实现方式中, 该控制单元, 还用于确定 其他数据中心节点到该数据中心节点的传输链路的跳数和传输延时, 其中该 其他数据中心节点是该数据中心节点所在的数据中心系统中除该数据中心 节点以外的数据中心节点; 该控制单元, 具体用于基于节能优化策略, 根据 该传输链路的跳数和传输延时, 确定该每个热度等级对应的镜像节点。
结合第二种可能的实现方式,在第三种可能的实现方式中,该控制单元, 具体用于根据该传输链路的跳数和传输延时,采用用于优化最小平均跳数和 最小平均传输延时的算法, 确定该每个热度等级对应的镜像节点。
结合第一方面或上述任一种可能的实现方式,在第四种可能的实现方式 中, 该通信单元, 还用于将第一用户设备到该数据中心节点的传输链路的跳 数、该第一用户设备到该数据中心节点的传输延时和该数据中心节点的负载 信息发送给该第一用户设备, 其中该第一用户设备为请求该数据中心节点保 存的数据的用户设备。
第二方面,本发明实施例提供一种区域网络节点,该区域网络节点包括: 存储单元, 用于保存数据; 控制单元, 用于确定每个目标数据的优先级, 并 且基于该每个目标数据的优先级, 更新该存储单元保存的数据, 其中, 该目 标数据为用户设备请求的保存在至少一个数据中心节点且没有保存在该存 储单元的数据。
结合第二方面, 在第一种可能的实现方式中, 该控制单元, 具体用于确 定该每个目标数据的热度等级和该每个目标数据的传输链路的跳数,根据该 每个目标数据的热度等级和该每个目标数据的传输链路的跳数,确定该每个 目标数据的优先级。
结合第二方面或第一种可能的实现方式, 在第二种可能的实现方式中, 该每个目标数据基于用于优化最小平均跳数和最小平均传输延时的算法保 存在至少一个数据中心节点。
结合第二方面或上述任一种可能的实现方式,在第三种可能的实现方式 中,该控制单元,还用于确定该存储单元保存的数据的优先级; 该控制单元, 具体用于基于该每个目标数据的优先级、该存储单元的剩余空间以及该存储 单元保存的数据的优先级, 更新该存储单元保存的数据。
第三方面, 本发明实施例提供一种提供数据服务的方法, 该方法由数据 中心节点执行, 该方法包括: 基于第一时间粒度统计该数据中心节点保存的 每个数据的点击率;根据该每个数据的点击率,确定该每个数据的热度等级; 基于节能优化策略, 确定每个热度等级对应的镜像节点; 将该数据中心节点 保存的数据复制到与该保存的数据的热度等级对应的镜像节点。
结合第三方面, 在第一种可能的实现方式中, 该方法还包括: 确定其他 数据中心节点到该数据中心节点的传输链路的跳数和传输延时,其中该其他 数据中心节点是该数据中心节点所在的数据中心系统中除该数据中心节点 以外的数据中心节点, 其中, 该确定每个热度等级对应的镜像节点, 包括: 根据该传输链路的跳数和传输延时, 确定该每个热度等级对应的镜像节点。
结合第一种可能的实现方式, 在第二种可能的实现方式中, 该基于节能 优化策略, 根据该传输链路的跳数和传输延时, 确定该每个热度等级对应的 镜像节点, 包括: 根据该传输链路的跳数和传输延时, 采用用于优化最小平 均跳数和最小平均传输延时的算法, 确定该每个热度等级对应的镜像节点。
结合第三方面或上述任一种可能的实现方式,在第三种可能的实现方式 中, 该方法还包括: 将第一用户设备到该数据中心节点的传输链路的跳数、 该第一用户设备到该数据中心节点的传输延时和该数据中心节点的负载信 息发送给该第一用户设备, 其中该第一用户设备为请求该数据中心节点保存 的数据的用户设备。
第四方面, 本发明实施例提供一种提供数据服务的方法, 该方法由区域 网络节点执行, 该方法包括: 确定每个目标数据的优先级, 其中该目标数据 为用户设备请求的保存在至少一个数据中心节点且没有保存在该存储单元 的数据; 基于该每个目标数据的优先级, 更新该区域网络节点保存的数据。
结合第四方面, 在第一种可能的实现方式中, 该方法还包括: 确定该每 个目标数据的热度等级和该每个目标数据的传输链路的跳数, 其中, 该确定 每个目标数据的优先级, 包括: 根据该每个目标数据的热度等级和该每个目 标数据的传输链路的跳数, 确定该每个目标数据的优先级。
结合第四方面或第一种可能的实现方式, 在第二种可能的实现方式中, 该每个目标数据基于用于优化最小平均跳数和最小平均传输延时的算法保 存在至少一个数据中心节点。
结合第四方面或上述任一种可能的实现方式,在第三种可能的实现方式 中, 该方法还包括: 确定该区域网络节点保存的数据的优先级, 其中, 该基 于该每个目标数据的优先级, 更新该区域网络节点保存的数据, 包括: 基于 该每个目标数据的优先级、该区域网络节点的剩余空间以及该区域网络节点 保存的数据的优先级, 更新该区域网络节点保存的数据。
根据本发明所提供的实施例, 由于多个数据中心节点保存相同的数据, 用户设备可以从多个数据中心节点中的任一数据中心节点来获取数据。 此 夕卜, 由于这些镜像节点是根据最小平均跳数和最小平均传输延时确定的, 因 此可以提高用户设备获取这些镜像节点上的数据的效率并降低获取数据时 的能耗。 同时区域网络节点可以根据用户设备所请求的数据, 实时更新自身 的緩存装置中所保存的数据, 以便所保存的数据是用户设备点击率高的数
据。 用户设备可以快速的从区域网络节点获取需要的数据, 而不需要从数据 中心节点获取数据。 这样, 可以加快用户设备获取数据的速度, 提升用户的 体验。 附图说明
为了更清楚地说明本发明实施例的技术方案, 下面将对本发明实施例中 所需要使用的附图作筒单地介绍, 显而易见地, 下面所描述的附图仅仅是本 发明的一些实施例, 对于本领域普通技术人员来讲, 在不付出创造性劳动的 前提下, 还可以根据这些附图获得其他的附图。
图 1是根据本发明实施例提供的数据中心系统的示意图。
图 2是根据本发明实施例提供的数据中心节点的结构框图。
图 3是根据本发明实施例提供的区域网络节点的结构框图。
图 4是根据本发明实施例提供的网络设备的结构框图。
图 5是根据本发明实施例提供的网络设备的结构框图。
图 6是根据本发明实施例提供的提供数据服务的方法的示意性流程图。 图 7是根据本发明实施例提供的提供数据服务的方法的示意性流程图。 图 8是根据本发明实施例提供的提供数据服务的方法的示意性流程图。 具体实施方式
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行 清楚、 完整地描述, 显然, 所述的实施例是本发明的一部分实施例, 而不是 全部实施例。 基于本发明中的实施例, 本领域普通技术人员在没有做出创造 性劳动的前提下所获得的所有其他实施例, 都应属于本发明保护的范围。
本发明实施例提供一种数据中心系统,该数据中心系统包括 N个数据中 心节点和 M个区域网络节点, 其中该区域网络节点具有緩存装置, N、 M为 正整数。 本领域技术人员可以理解, 在实际应用中, 该数据中心节点和区域 网络节点可以是不同功能的服务器,其中该数据中心节点可以是位于核心网 中的服务器, 该区域网络节点可以是位于城域网中的服务器。
第一数据中心节点, 用于基于第一时间粒度统计该第一数据中心节点保 存的每个数据的点击率, 根据该每个数据的点击率, 确定该每个数据的热度 等级, 基于节能优化测量, 确定每个热度等级对应的镜像节点, 将该第一数
据中心节点保存的数据复制到该保存的数据的热度等级所对应数量的镜像 节点,其中该第一数据中心节点为该 N个数据中心节点中任一个数据中心节 点,该镜像节点的选择可为该 N个数据中心节点中除该第一数据中心节点以 外的数据中心节点。
换句话说,该 N个数据中心节点中的每一个数据中心节点都可以基于第 一时间粒度统计自身所保存的每个数据的点击率, 根据每个数据的点击率, 确定每个数据的热度等级, 并基于节能优化策略, 采用用于优化最小平均跳 数和最小平均传输延时的算法确定每个热度等级对应的镜像节点, 并将所保 存的数据复制到与该数据的热度等级对应数量的其他数据中心节点中。 这 样, 由于多个数据中心节点保存相同的数据, 用户设备可以从多个数据中心 节点中的任一数据中心节点来获取数据。 此外, 由于这些镜像节点是根据最 小平均跳数和最小平均传输延时确定的, 因此可以提高用户设备获取这些镜 像节点上的数据的效率。
第一区域网络节点, 用于确定每个目标数据的优先级, 基于该每个目标 数据的优先级, 更新该第一区域网络节点的緩存装置中保存的数据, 其中该 目标数据为用户设备请求的保存在至少一个数据中心节点并且该第一区域 网络节点的緩存装置中没有保存的数据, 该第一区域网络节点为该 M个区 域网络节点中的任一区域网络节点。
进一步, 该每个目标数据基于用于优化最小平均跳数和最小平均传输延 时的算法保存在至少一个数据中心节点。 换句话说, 该每个目标数据保存在 至少一个数据中心节点中, 该至少一个数据中心节点是基于用于用户最小平 均跳数和最小平均传输延时的算法确定的。
换句话说, 该 M个区域网络节点中的每一个区域网络节点都可以根据 用户设备所请求的数据, 实时更新自身的緩存装置中所保存的数据, 以便区 域网络节点的緩存装置所保存的数据是用户设备点击率高的数据。用户设备 可以快速的从区域网络节点获取需要的数据, 而不需要从数据中心节点获取 数据。 这样, 可以加快用户设备获取数据的速度, 提升用户的体验。
本发明实施例提供的数据中心系统, 能够使得用户设备快速地从区域网 络节点获取到热度高的数据 (即点击率高的数据)。 此外, 用户设备还可以 从多个数据中心节点中获取到需要的数据, 并且获取到需要数据的效率较 高。 因此, 本发明实施例所提供的数据中心系统能够节约用户设备获取数据
时所需的能耗。
具体来说, 用户设备在需要获取该数据中心网络中所保存的数据时, 可 以向与该用户设备连接的区域网络节点发送数据请求, 该数据请求用于请求 获取该数据中心网络所保存的数据。在该区域网络节点不具有保存数据的緩 存装置的情况下, 或者在该区域网络节点的保存数据的緩存装置中没有该数 据请求消息所请求的数据的情况下,该区域网络节点会将该用户设备发送的 数据请求转发给数据中心节点列表内的所有数据中心节点所有数据中心节 点, 其中该数据中心节点列表保存在该区域网络节点中。 这样, 各个数据中 心节点可以根据用户设备的数据请求,将该数据请求对应的数据发送给该用 户设备。 在此情况下, 第一数据中心节点可以基于第一时间粒度, 统计用户 设备所请求的数据的点击率, 然后根据点击率将数据划分为不同的热度等 级, 利用热度等级反映数据的热度。 第一数据中心节点还可以基于节能优化 策略, 确定每个热度等级对应的镜像节点。 不同的热度等级可以对应于不同 数量的镜像节点。对应于高热度等级的镜像节点多于对应于低热度等级的镜 像节点。 这样, 用户设备可以有更多的途径获取高热度等级的数据。 同时, 由于高热度等级的数据的备份增加, 高热度等级的数据的安全性也同时得到 提升。 并且, 同随机确定的镜像节点相比, 用户设备从这些基于节能优化策 略确定的镜像节点获取数据所需的能耗可能较低并且有较高的效率。 相应 的, 该第一数据中心节点还可以用于接收并保存该数据中心系统中的其他数 据中心节点发送的数据。
例如, 第一数据中心节点保存有 Al、 A2、 A3、 A4 四个数据。 第一数 据中心节点可以统计 2个小时内这四个数据的点击率。例如,在 2个小时内, 第一数据中心收到了 220个用于请求 A1 的数据请求, 250个用于请求 A2 的请求, 1000个用于请求 A3的请求以及 2000个用于请求 A4的请求。换句 话说, Al、 A2、 A3、 A4这四个数据在 2个小时内的点击率分别为 200、 250、 500和 2000。 如果 2个小时内的点击率为 0至 199的数据的热度等级为 1 , 点击率 201至 500的数据的热度等级为 2,点击率为 501至 1000的数据的热 度等级为 3, 点击率 1001以上的数据的热度等级为 4, 则 A1和 A2的热度 等级为 2, A3的热度等级为 3, A4的热度等级为 4。 这样, 就可以通过热度 等级反映出第一数据中心存储的数据的热度。 进一步, 第一数据中心根据热 度等级, 确定出不同的镜像节点, 并将对应的数据复制到这些镜像节点上。
例如, 第一数据中心可以确定该数据中心系统中所有的数据中心节点都是对 应于热度等级为 4的镜像节点, 该数据中心系统中二分之一的数据中心节点 是对应于热度等级为 3的镜像节点,该数据中心系统中三分之一的数据中心 节点是对应于热度等级为 2的镜像节点,而热度等级为 1则不设置镜像节点。 也就是说, 第一数据中心节点可以将 Α4复制到该数据中心系统中所有的数 据中心节点上, 将 A3复制到该数据中心系统中的二分之一的数据中心节点 上, 将 A1和 Α2复制到该数据中心系统中三分之一的数据中心节点上。 如 果第一数据中心节点还保存数据 Α5并且 Α5两个小时内的点击率为 50, 则 第一数据中心节点不将 Α5复制到任何其他的数据中心节点上。
本领域技术人员可以理解, 上述例子仅是为了说明第一时间粒度以及点 击率和热度等级的关系。 第一时间粒度、 点击率和热度等级的关系完全可以 根据需要进行设定。 例如, 第一时间粒度可以更长或更短。 点击率和热度等 级的范围也可以改变。
第一数据中心节点,具体用于根据该 Ν个数据中心节点中除该第一数据 中心节点以外的数据中心节点到该第一数据中心节点的传输链路的跳(英 文: hop )数和传输延时, 采用用于优化最小平均跳数和最小平均传输延时 的算法确定该每个热度等级对应的镜像节点。
可选的, 作为一个实施例, 该用于优化最小平均跳数和最小平均传输延 时的算法可以是用于优化最小平均跳数和最小平均传输延时的线性规划算 法。 在此情况下, 该第一数据中心节点, 具体用于采用用于优化最小平均跳 数和最小平均传输延时的线性规划算法, 确定每个热度等级对应的镜像节 点。 该线性规划算法的目标方程以节能为目的, 使得用户设备能够高效地从 镜像节点上获取需要的数据。在数据传输过程中的最小平均跳数和最小平均 传输延时为优化对象, 从而达到节能高效的目的。 也就是说, 该线性规划算 法的限制方程主要包括流量转移方程和镜像节点数目的限制。数据传输过程 中的最小平均跳数和最小平均传输延时是根据该 N 个数据中心节点中除该 第一数据中心节点以外的数据中心节点到该第一数据中心节点的传输链路 的跳数的平均值和传输延时的平均值。
可选的, 作为另一个实施例, 该用于优化最小平均跳数和最小平均传输 延时的算法可以是用于优化最小平均跳数和最小平均传输延时的基因算法。 在此情况下, 该第一数据中心节点, 具体用于采用用于优化最小平均跳数和
最小平均传输延时的基因算法, 确定每个热度等级对应的镜像节点。 该基因 算法中适度计算函数是由数据传输的最小平均跳数和最小平均传输延时来 共同决定的。 该基因算法以节能为目的, 使得用户设备能够高效地从镜像节 点上获取需要的数据。数据传输过程中的最小平均跳数和最小平均传输延时 是根据该 N 个数据中心节点中除该第一数据中心节点以外的数据中心节点 到该第一数据中心节点的传输链路的跳数的平均值和传输延时的平均值。
该第一数据中心节点,还用于将第一用户设备到该第一数据中心的传输 链路的跳数、该第一用户设备到该第一数据中心的传输延时和该第一数据中 心节点的负载信息发送给该第一用户设备, 其中该第一用户设备为请求该第 一数据中心节点保存的数据的用户设备。
具体来说, 用户设备向该数据中心系统中的区域网络节点发送数据请 求。如果接收到该数据请求的区域网络节点的緩存装置中没有该数据请求所 请求的数据, 则该区域网络节点会将该用户设备发送的数据请求转发给数据 中心节点列表内的所有数据中心节点所有数据中心节点,其中该数据中心节 点列表保存在该区域网络节点中。 为了筒便描述, 如果一个用户设备发送的 数据请求所请求的数据不在区域网络节点的緩存装置中,则称这个用户设备 为第一用户设备。 如果该第一数据中心节点接收到该数据请求, 则该第一数 据中心节点会确定该第一用户设备到该第一数据中心节点传输链路的跳数、 该第一用户设备到该第一数据中心节点的传输延时和该第一数据中心节点 的负载。 然后, 该第一数据中心会将该传输链路的跳数、 该传输延时以及该 第一数据中心的负载信息发送给该第一用户设备。 类似的, 该数据中心系统 中的其他数据中心节点也可以接收到该第一用户设备发送的数据请求并向 该第一用户设备发送各自到该第一用户设备的传输链路的跳数、传输延时以 及各自的负载信息。 第一用户设备可以接收到多个数据中心节点发送的传输 链路的跳数、 传输延时以及负载信息。 第一用户设备可以根据传输链路的跳 数、 传输延时以及负载信息, 选择合适的数据中心节点来进行数据传输。 例 如, 可以采用以下公式确定数据中心节点的等级:
L(i) = hop (i) * a% + delay (i) * b% + workload (i) * c% , 公式 1.1 其中, L(i)表示第 i个数据中心节点的等级, hop(i)表示第 i个数据中心 节点到该第一用户设备的传输链路的跳数, delay(i)表示第 i个数据中心节点 到该第一用户设备的传输延时, workload(i)表示第 i个数据中心节点的负载,
a%、 b%和 c%分别表示传输链路的跳数、 传输延时以及负载占数据中心节 点的等级的权重。 本领域技术人员可以理解, a%、 /?%和^"%可以根据需要 进行设计。例如,如果希望传输链路的跳数对数据中心节点的等级影响最大, 则可以使 a%的取值大于 %和 c%。 如果希望传输延时对数据中心节点的等 级影响最大, 则可以使 %的取值大于 %和^>%。
第一区域网络节点, 具体用于确定每个目标数据的热度等级和每个目标 数据的传输链路的跳数,根据该每个目标数据的热度等级和每个目标数据的 传输链路的跳数, 确定每个目标数据的优先级。
具体来说, 第一区域网络节点确定每个存储目标数据的数据中心节点到 该第一区域网络节点的传输链路的跳数。 然后, 根据每个目标数据的热度等 级和每个目标数据的传输链路的跳数, 确定每个目标数据的优先级。 例如, 可以采用以下公式确定目标数据的优先级:
P(i) = pop (i) * x% + hop(i) * y% , 公式 1.2 其中, P(i)表示第 i个目标数据的优先级, pop(i)表示第 i个目标数据的 热度等级, hop(i)表示第 i个目标数据的传输链路的跳数, 表示目标数据 的热度等级占目标数据的优先级的权重, y%表示目标数据的传输链路的跳 数占目标数据的优先级的权重。本领域技术人员可以理解 _1%和}^可以根据 需要进行设计。 例如, 如果希望使热度等级占优先级的权重大于传输链路的 跳数占优先级的权重,则可以使 大于) ^。。如果希望使热度等级占优先级 的权重小于传输链路的跳数占优先级的权重, 则可以是) ^大于 x%。
第一区域网络节点在确定了每个目标数据的优先级后,对目标数据按照 优先级进行排列,依次判断是否需要将该目标数据存储到该第一区域网络节 点的緩存装置中。
以目标数据中优先级最高的目标数据 (为了筒便描述, 以下称为第一目 标数据)为例, 第一区域网络节点会判断緩存装置中是否具有足够的剩余空 间存储该第一目标数据。如果该緩存装置有足够的剩余空间存储该第一目标 数据, 则将该第一目标数据存储到该第一区域网络节点的緩存装置中。
如果该緩存装置中没有足够的剩余空间存储该第一目标数据, 则该第一 区域网络节点还用于确定该緩存装置中的每个数据的优先级,根据该每个目 标数据的优先级、该緩存装置中的每个数据的优先级以及该緩存装置的剩余 空间, 更新该緩存装置中的数据。 具体来说, 该第一区域网络节点首先确定
出该緩存装置中优先级最低的数据 (为了筒便描述, 以下筒称为第一緩存数 据)。 如果第一緩存数据的优先级高于该第一目标数据的优先级, 则不将该 第一目标数据存储到该緩存装置中。如果该第一緩存数据的优先级低于该第 一目标数据的优先级, 则确定删除该第一緩存数据的优先级后该緩存装置中 的剩余空间是否大于该第一目标数据所占用的存储空间。如果删除该第一緩 存数据后的剩余空间大于该第一目标数据所占用的空间, 则将该第一緩存数 据删除并将该第一目标数据存储到该緩存装置中。 类似的, 可以继续将该第 一区域网络节点统计的目标数据与该第一区域网络节点的緩存装置所保存 的数据进行比较, 以便将优先级高的数据保存到该第一区域网络节点的緩存 装置中。 这样, 可以使得第一区域网络节点的緩存装置中保存的数据都是优 先级高的数据。 换句话说, 第一区域网络节点的緩存装置中所保存的数据是 经常被用户设备所请求的数据。用户设备可以快速的从第一区域网络节点获 取所需的数据, 而不必从数据中心节点获取数据。 这样, 可以加快用户设备 获取数据的速度, 并节约获取数据时需要的能耗。
例如, 第一区域网络节点确定 3个目标数据, 这 3个目标数据按照优先 级从大到小排列依次是 Dl、 D2和 D3。 同时, 该第一区域网络节点的緩存 装置中保存有 dl、 d2和 d3三个数据。 如果该第一区域网络节点的緩存装置 的剩余空间能够存储 D1 , 则将 D1 保存到该第一区域网络节点的緩存装置 中。 此时, 第一区域网络节点的緩存装置中保存的数据为 dl、 d2、 d3和 Dl。 如果第一区域网络节点的緩存装置中没有足够的存储空间存储 D1 , 则确定 第一区域网络节点的緩存装置中保存的三个数据的优先级。假设第一区域网 络节点的緩存装置中保存的三个数据的优先级从大到小排列依次为 dl、 d2 和 d3, 则判断 d3与 D1的优先级。 4 设 d3的优先级大于 D1的优先级, 则 保持第一区域网络节点的緩存装置中的数据不变。假设 d3的优先级小于 D1 的优先级, 则确定删除 d3后的第一区域网络节点的緩存装置中的剩余空间 是否能够保存 Dl。 如果确定删除 d3后的剩余空间能够保存 D1 , 则将 d3删 除并将 D1保存到该第一区域网络节点的緩存装置中。 此时, 该第一区域网 络节点的緩存装置中保存的数据为 dl、 d2和 Dl。 类似的, 该第一区域网络 节点将其他的目标数据和緩存装置中保存的数据也进行比较, 以便对緩存装 置中的数据进行更新, 从而使得緩存装置中保存的数据为优先级高的数据。
图 1是根据本发明实施例提供的数据中心系统的示意图。 图 1所示的数
据中心系统中包括 3个数据中心节点和 4个区域网络节点。
需要说明的是, 图 1仅是一个数据中心系统的具体例子。 该例子只是为 了帮助本领域技术人员更好地理解本发明实施例, 而非限制本发明实施例的 范围
图 2是根据本发明实施例提供的数据中心节点的结构框图。 图 2所示的 数据中心节点可以是本发明实施例提供的数据中心系统中 N 个数据中心节 点中的任一个数据中心节点。 图 2所示的数据中心节点可以是位于核心网的 网络设备,例如服务器。如图 2所示,数据中心节点 200包括:存储单元 201、 控制单元 202和通信单元 203。
存储单元 201 , 用于保存数据。
控制单元 202, 用于基于第一时间粒度统计的存储单元 201保存的每个 数据的点击率。
控制单元 202, 还用于根据该每个数据的点击率, 确定该每个数据的热 度等级。
控制单元 202, 还用于基于节能优化策略, 确定每个热度等级对应的镜 像节点。
通信单元 203 , 用于将存储单元 201保存的数据复制到与该保存的数据 的热度等级对应的镜像节点。
图 2所示的数据中心节点 200可以基于第一时间粒度统计自身所保存的 每个数据的点击率, 根据每个数据的点击率, 确定每个数据的热度等级, 并 采用用于优化最小平均跳数和最小平均传输延时的算法确定每个热度等级 对应的镜像节点, 并将所保存的数据复制到与该数据的热度等级对应的其他 数据中心节点中。 这样, 由于多个数据中心节点保存相同的数据, 用户设备 可以从多个数据中心节点中的任一数据中心节点来获取数据。 此外, 由于这 些镜像节点是根据最小平均跳数和最小平均传输延时确定的, 因此可以提高 用户设备获取这些镜像节点上的数据的效率并降低获取数据时的能耗。
进一步, 通信单元 203 , 还用于将第一用户设备到数据中心节点 200的 传输链路的跳数、该第一用户设备到数据中心节点 200的传输延时和数据中 心节点 200的负载信息发送给该第一用户设备, 其中该第一用户设备为请求 数据中心节点 200保存的数据的用户设备。
具体来说, 用户设备向该数据中心系统中的区域网络节点发送数据请
求。如果接收到该数据请求的区域网络节点的緩存装置中没有该数据请求所 请求的数据, 则该区域网络节点会将该用户设备发送的数据请求转发给数据 中心节点列表内的所有数据中心节点所有数据中心节点, 其中该数据中心节 点列表保存在该区域网络节点中。 为了筒便描述, 如果一个用户设备发送的 数据请求所请求的数据不在区域网络节点的緩存装置中, 则称这个用户设备 为第一用户设备。如果数据中心节点 200的通信单元 203接收到该数据请求, 则控制单元 202, 可以用于会确定该第一用户设备到数据中心节点 200的传 输链路的跳数、该第一用户设备到数据中心节点 200的传输延时和数据中心 节点 200的负载。 然后, 通信单元 203 , 可以用于会将该传输链路的跳数、 该传输延时以及该数据中心节点 200的负载信息发送给该第一用户设备, 以 便于该第一用户设备根据该传输链路的跳数、该传输延时以及数据中心节点 200的负载信息确定是否从数据中心节点 200中获取数据。
控制单元 202, 可以用于基于第一时间粒度, 统计用户设备所请求的数 据的点击率, 然后根据点击率将存储单元 201存储的数据划分为不同的热度 等级, 利用热度等级反映数据的热度。 控制单元 202, 还可以用于基于节能 优化策略, 确定每个热度等级对应的镜像节点。 不同的热度等级可以对应于 不同的镜像节点。 具体来说, 控制单元 202, 在确定镜像节点时, 可以基于 节能优化策略, 为不同的热度等级确定不同数量的镜像节点。 例如, 对应于 高热度等级的镜像节点多于对应于低热度等级的镜像节点。 这样, 用户设备 可以有更多的途径获取高热度等级的数据。 同时, 由于高热度等级的数据的 备份增加, 高热度等级的数据的安全性也同时得到提升。
进一步, 控制单元 202, 还用于确定其他数据中心节点到数据中心节点 200的传输链路的跳数和传输延时, 其中该其他数据中心节点是数据中心节 点 200所在的数据中心系统中除数据中心节点 200 以外的其他数据中心节 点。 控制单元 202, 具体用于根据该传输链路的跳数和传输延时, 确定该每 个热度等级对应的镜像节点。
具体地, 控制单元 202, 具体用于根据该传输链路的跳数和传输延时, 采用用于优化最小平均跳数和最小平均传输延时的算法,确定该每个热度等 级对应的镜像节点。
可选的, 作为一个实施例, 控制单元 202, 具体用于采用用于优化最小 平均跳数和最小平均传输延时的线性规划算法,确定每个热度等级对应的镜
像节点。 该线性规划算法的目标方程以节能为目的, 使得用户设备能够高效 地从镜像节点上获取需要的数据。在数据传输过程中的最小平均跳数和最小 平均传输延时为优化对象, 从而达到节能高效的目的。 也就是说, 该线性规 划算法的限制方程主要包括流量转移方程和镜像节点数目的限制。数据传输 过程中的最小平均跳数和最小平均传输延时是根据该 N 个数据中心节点中 除数据中心节点 200以外的数据中心节点到数据中心节点 200的传输链路的 跳数的平均值和传输延时的平均值。
可选的, 作为另一个实施例, 控制单元 202, 具体用于采用用于优化最 小平均跳数和最小平均传输延时的基因算法,确定每个热度等级对应的镜像 节点。该基因算法中适度计算函数是由数据传输的最小平均跳数和最小平均 传输延时来共同决定的。 该基因算法以节能为目的, 使得用户设备能够高效 地从镜像节点上获取需要的数据。数据传输过程中的最小平均跳数和最小平 均传输延时是根据该 N个数据中心节点中除数据中心节点 200以外的数据中 心节点到数据中心节点 200 的传输链路的跳数的平均值和传输延时的平均 值。
图 3是根据本发明实施例提供的区域网络节点的结构框图。 图 3所示的 区域网络节点可以是本发明实施例提供的数据中心系统中 M个区域网络节 点中的任一个区域网络节点。 图 3所示的区域网络节点可以是位于城域网的 网络设备,例如服务器。如图 3所示, 区域网络节点 300包括:存储单元 301 和控制单元 302。
存储单元 301 , 用于保存数据。
控制单元 302, 用于确定每个目标数据的优先级, 其中所述目标数据为 用户设备请求的保存在至少一个数据中心节点并且没有保存在存储单元 301 的数据。
控制单元 302, 还用于基于所述每个目标数据的优先级, 更新存储单元
301保存的数据。
图 3所示的区域网络节点 300可以根据用户设备所请求的数据, 实时更 新自身的緩存装置中所保存的数据, 以便存储单元 301所保存的数据是用户 设备点击率高的数据。 用户设备可以快速的从区域网络节点获取需要的数 据, 而不需要从数据中心节点获取数据。 这样, 可以加快用户设备获取数据 的速度, 提升用户的体验。
控制单元 302, 具体用于确定所述每个目标数据的热度等级和所述每个 目标数据的传输链路的跳数,根据所述每个目标数据的热度等级和所述每个 目标数据的传输链路的跳数, 确定所述每个目标数据的优先级。
具体来说, 控制单元 302, 用于确定每个存储目标数据的数据中心节点 到区域网络节点 300的传输链路的跳数。 然后, 控制单元 302, 根据每个目 标数据的热度等级和每个目标数据的传输链路的跳数,确定每个目标数据的 优先级。 例如, 控制单元 302, 可以采用以下公式确定目标数据的优先级:
P (i) = pop (i) ^ x o + hop(i) ^ y o , 公式 1.3 其中, P(i)表示第 i个目标数据的优先级, pop(i)表示第 i个目标数据的 热度等级, hop(i)表示第 i个目标数据的传输链路的跳数, 表示目标数据 的热度等级占目标数据的优先级的权重, y%表示目标数据的传输链路的跳 数占目标数据的优先级的权重。本领域技术人员可以理解 _1%和}^可以根据 需要进行设计。 例如, 如果希望使热度等级占优先级的权重大于传输链路的 跳数占优先级的权重,则可以使 大于 y%。如果希望使热度等级占优先级 的权重小于传输链路的跳数占优先级的权重, 则可以是) ^大于 x%。
进一步,控制单元 302,还用于确定存储单元 301保存的数据的优先级。 控制单元 302, 具体用于基于所述每个目标数据的优先级、 存储单元 301的 剩余空间以及存储单元 301保存的数据的优先级, 更新存储单元 301保存的 数据。
具体来说, 控制单元 302, 在确定了每个目标数据的优先级后, 对目标 数据按照优先级进行排列,依次判断是否需要将该目标数据存储到存储单元 301中。
以目标数据中优先级最高的目标数据 (为了筒便描述, 以下称为第一目 标数据)为例, 控制单元 302会判断存储单元 301中是否具有足够的剩余空 间存储该第一目标数据。如果存储单元 301有足够的剩余空间存储该第一目 标数据, 则控制单元 302将该第一目标数据存储到存储单元 301中。
如果存储单元 301中没有足够的剩余空间存储该第一目标数据, 则控制 单元 302还用于确定存储单元 301中的每个数据的优先级,根据该每个目标 数据的优先级、存储单元 301中的每个数据的优先级以及存储单元 301的剩 余空间, 更新存储单元 301中的数据。 具体来说, 控制单元 302首先确定出 存储单元 301中优先级最低的数据 (为了筒便描述, 以下筒称为第一緩存数
据)。 如果第一緩存数据的优先级高于该第一目标数据的优先级, 则控制单 元 302不将该第一目标数据存储到存储单元 301中。如果该第一緩存数据的 优先级低于该第一目标数据的优先级, 则控制单元 302确定删除该第一緩存 数据的优先级后存储单元 301中的剩余空间是否大于该第一目标数据所占用 的存储空间。如果删除该第一緩存数据后的剩余空间大于该第一目标数据所 占用的空间, 则控制单元 302将该第一緩存数据删除并将该第一目标数据存 储到存储单元 301中。
进一步, 该目标数据基于用于用户最小平均跳数和最小平均传输延时的 算法保存在至少一个数据中心节点。
图 4是根据本发明实施例提供的网络设备的结构框图。 图 4所示的网络 设备可以是本发明实施例提供的数据中心系统中 N 个数据中心节点中的任 一个数据中心节点。 图 4所示的网络设备可以是位于核心网的网络设备, 例 如服务器。 如图 4所示, 网络设备 400包括: 存储器 401、 处理器 402和收 发器 403。
存储器 401 , 用于保存数据。
处理器 402, 用于基于第一时间粒度统计的存储器 401保存的每个数据 的点击率。
处理器 402, 还用于根据该每个数据的点击率, 确定该每个数据的热度 等级。
处理器 402, 还用于基于节能优化策略, 确定每个热度等级对应的镜像 节点。
收发器 403 , 用于将存储器 401保存的数据复制到与该保存的数据的热 度等级对应的镜像节点。
图 4所示的网络设备 400可以基于第一时间粒度统计自身所保存的每个 数据的点击率, 根据每个数据的点击率, 确定每个数据的热度等级, 并采用 用于优化最小平均跳数和最小平均传输延时的算法确定每个热度等级对应 的镜像节点, 并将所保存的数据复制到与该数据的热度等级对应的其他数据 中心节点中。 这样, 由于多个数据中心节点保存相同的数据, 用户设备可以 从多个数据中心节点中的任一数据中心节点来获取数据。 此外, 由于这些镜 像节点是根据最小平均跳数和最小平均传输延时确定的, 因此可以提高用户 设备获取这些镜像节点上的数据的效率并降低获取数据时的能耗。
进一步, 收发器 403 , 还用于将第一用户设备到网络设备 400的传输链 路的跳数、该第一用户设备到网络设备 400的传输延时和网络设备 400的负 载信息发送给该第一用户设备, 其中该第一用户设备为请求网络设备 400保 存的数据的用户设备。
具体来说, 用户设备向该数据中心系统中的区域网络节点发送数据请 求。如果接收到该数据请求的区域网络节点的緩存装置中没有该数据请求所 请求的数据, 则该区域网络节点会将该用户设备发送的数据请求转发给数据 中心节点列表内的所有数据中心节点所有数据中心节点, 其中该数据中心节 点列表保存在该区域网络节点中。 为了筒便描述, 如果一个用户设备发送的 数据请求所请求的数据不在区域网络节点的緩存装置中, 则称这个用户设备 为第一用户设备。 如果网络设备 400的收发器 403接收到该数据请求, 则处 理器 402, 可以用于会确定该第一用户设备到网络设备 400的传输链路的跳 数、 该第一用户设备到网络设备 400的传输延时和网络设备 400的负载。 然 后, 收发器 403 , 可以用于会将该传输链路的跳数、 该传输延时以及该网络 设备 400的负载信息发送给该第一用户设备, 以便于该第一用户设备根据该 传输链路的跳数、该传输延时以及网络设备 400的负载信息确定是否从网络 设备 400中获取数据。
处理器 402, 可以用于基于第一时间粒度, 统计用户设备所请求的数据 的点击率,然后根据点击率将存储器 401存储的数据划分为不同的热度等级, 利用热度等级反映数据的热度。处理器 402,还可以用于基于节能优化策略, 确定每个热度等级对应的镜像节点。 不同的热度等级可以对应于不同的镜像 节点。 具体来说, 处理器 402, 在确定镜像节点时, 可以为不同的热度等级 确定不同数量的镜像节点。 例如, 对应于高热度等级的镜像节点多于对应于 低热度等级的镜像节点。 这样, 用户设备可以有更多的途径获取高热度等级 的数据。 同时, 由于高热度等级的数据的备份增加, 高热度等级的数据的安 全性也同时得到提升。
进一步, 处理器 402, 还用于确定其他数据中心节点到网络设备 400的 传输链路的跳数和传输延时, 其中该其他数据中心节点是网络设备 400所在 的数据中心系统中除网络设备 400以外的其他数据中心节点。 处理器 402, 具体用于根据该传输链路的跳数和传输延时,确定该每个热度等级对应的镜 像节点。
具体地, 处理器 402, 具体用于根据该传输链路的跳数和传输延时, 采 用用于优化最小平均跳数和最小平均传输延时的算法,确定该每个热度等级 对应的镜像节点。
可选的, 作为一个实施例, 处理器 402, 具体用于采用用于优化最小平 均跳数和最小平均传输延时的线性规划算法,确定每个热度等级对应的镜像 节点。 该线性规划算法的目标方程以节能为目的, 使得用户设备能够高效地 从镜像节点上获取需要的数据。在数据传输过程中的最小平均跳数和最小平 均传输延时为优化对象, 从而达到节能高效的目的。 也就是说, 该线性规划 算法的限制方程主要包括流量转移方程和镜像节点数目的限制。数据传输过 程中的最小平均跳数和最小平均传输延时是根据该 N 个数据中心节点中除 网络设备 400以外的数据中心节点到网络设备 400的传输链路的跳数的平均 值和传输延时的平均值。
可选的, 作为另一个实施例, 处理器 402, 具体用于采用用于优化最小 平均跳数和最小平均传输延时的基因算法,确定每个热度等级对应的镜像节 点。该基因算法中适度计算函数是由数据传输的最小平均跳数和最小平均传 输延时来共同决定的。 该基因算法以节能为目的, 使得用户设备能够高效地 从镜像节点上获取需要的数据。数据传输过程中的最小平均跳数和最小平均 传输延时是根据该 N个数据中心节点中除网络设备 400以外的数据中心节点 到网络设备 400的传输链路的跳数的平均值和传输延时的平均值。
图 5是根据本发明实施例提供的网络设备的结构框图。 图 5所示的网络 设备可以是本发明实施例提供的数据中心系统中 M个区域网络节点中的任 一个区域网络节点。 图 5所示的网络设备可以是位于城域网的网络设备, 例 如服务器。 如图 5所示, 网络设备 500包括: 存储器 501和处理器 502。
存储器 501 , 用于保存数据。
处理器 502, 用于确定每个目标数据的优先级, 其中所述目标数据为用 户设备请求的保存在至少一个数据中心节点并且没有保存在存储器 501的数 据。
处理器 502, 还用于基于所述每个目标数据的优先级, 更新存储器 501 保存的数据。
图 5所示的网络设备 500可以根据用户设备所请求的数据, 实时更新自 身的緩存装置中所保存的数据, 以便存储器 501所保存的数据是用户设备点
击率高的数据。用户设备可以快速的从区域网络节点的緩存装置中获取需要 的数据, 而不需要从数据中心节点获取数据。 这样, 可以加快用户设备获取 数据的速度, 提升用户的体验。
处理器 502 , 具体用于确定所述每个目标数据的热度等级和所述每个目 标数据的传输链路的跳数,根据所述每个目标数据的热度等级和所述每个目 标数据的传输链路的跳数, 确定所述每个目标数据的优先级。
具体来说, 处理器 502, 用于确定每个存储目标数据的数据中心节点到 网络设备 500的传输链路的跳数。 然后, 处理器 502, 根据每个目标数据的 热度等级和每个目标数据的传输链路的跳数, 确定每个目标数据的优先级。 例如, 处理器 502, 可以采用以下公式确定目标数据的优先级:
Ρ(ή = ρορ (ή * χΨο + hop(i) * y9o , 公式 1.4 其中, P(i)表示第 i个目标数据的优先级, pop(i)表示第 i个目标数据的 热度等级, hop(i)表示第 i个目标数据的传输链路的跳数, 表示目标数据 的热度等级占目标数据的优先级的权重, y%表示目标数据的传输链路的跳 数占目标数据的优先级的权重。本领域技术人员可以理解 _1%和>^可以根据 需要进行设计。 例如, 如果希望使热度等级占优先级的权重大于传输链路的 跳数占优先级的权重,则可以使 大于 y%。如果希望使热度等级占优先级 的权重小于传输链路的跳数占优先级的权重, 则可以是}^。大于 x%。
进一步, 处理器 502, 还用于确定存储器 501保存的数据的优先级。 处 理器 502 , 具体用于基于所述每个目标数据的优先级、 存储器 501的剩余空 间以及存储器 501保存的数据的优先级, 更新存储器 501保存的数据。
具体来说, 处理器 502, 在确定了每个目标数据的优先级后, 对目标数 据按照优先级进行排列, 依次判断是否需要将该目标数据存储到存储器 501 中。
以目标数据中优先级最高的目标数据 (为了筒便描述, 以下称为第一目 标数据)为例, 处理器 502会判断存储器 501中是否具有足够的剩余空间存 储该第一目标数据。如果存储器 501有足够的剩余空间存储该第一目标数据, 则处理器 502将该第一目标数据存储到存储器 501中。
如果存储器 501中没有足够的剩余空间存储该第一目标数据, 则处理器 502还用于确定存储器 501中的每个数据的优先级, 根据该每个目标数据的 优先级、 存储器 501中的每个数据的优先级以及存储器 501的剩余空间, 更
新存储器 501中的数据。 具体来说, 处理器 502首先确定出存储器 501中优 先级最低的数据 (为了筒便描述, 以下筒称为第一緩存数据)。 如果第一緩 存数据的优先级高于该第一目标数据的优先级, 则处理器 502不将该第一目 标数据存储到存储器 501中。如果该第一緩存数据的优先级低于该第一目标 数据的优先级, 则处理器 502 确定删除该第一緩存数据的优先级后存储器 501中的剩余空间是否大于该第一目标数据所占用的存储空间。 如果删除该 第一緩存数据后的剩余空间大于该第一目标数据所占用的空间, 则处理器 502将该第一緩存数据删除并将该第一目标数据存储到存储器 501中。
进一步, 该目标数据基于用于用户最小平均跳数和最小平均传输延时的 算法保存在至少一个数据中心节点。
图 6是根据本发明实施例提供的提供数据服务的方法的示意性流程图。 图 6所示的方法可以由数据中心节点执行。执行图 6所示的方法的数据中心 节点可以是图 3所示的数据中心节点。
601 ,基于第一时间粒度统计该数据中心节点保存的每个数据的点击率。 602, 根据该每个数据的点击率, 确定该每个数据的热度等级。
603 , 基于节能优化策略, 确定该每个热度等级对应的镜像节点。
604 , 将该数据中心节点保存的数据复制到与该保存的数据的热度等级 对应的镜像节点。
根据图 6所示的方法,数据中心节点可以基于第一时间粒度统计自身所 保存的每个数据的点击率, 根据每个数据的点击率, 确定每个数据的热度等 级, 并采用用于优化最小平均跳数和最小平均传输延时的算法确定每个热度 等级对应的镜像节点, 并将所保存的数据复制到与该数据的热度等级对应的 其他数据中心节点中。 这样, 由于多个数据中心节点保存相同的数据, 用户 设备可以从多个数据中心节点中的任一数据中心节点来获取数据。 此外, 由 于这些镜像节点是根据最小平均跳数和最小平均传输延时确定的, 因此可以 提高用户设备获取这些镜像节点上的数据的效率并降低用户设备获取数据 的能耗。
进一步, 图 6所示的方法还包括: 将第一用户设备到该数据中心节点的 传输链路的跳数、该第一用户设备到该数据中心节点的传输延时和该数据中 心节点的负载信息发送给该第一用户设备, 其中该第一用户设备为请求该数 据中心节点保存的数据的用户设备。
具体来说, 用户设备向该数据中心系统中的区域网络节点发送数据请 求。如果接收到该数据请求的区域网络节点的緩存装置中没有该数据请求所 请求的数据, 则该区域网络节点会将该用户设备发送的数据请求转发给数据 中心节点列表内的所有数据中心节点所有数据中心节点, 其中该数据中心节 点列表保存在该区域网络节点中。 为了筒便描述, 如果一个用户设备发送的 数据请求所请求的数据不在区域网络节点的緩存装置中,则称这个用户设备 为第一用户设备。 如果该数据中心节点接收到该数据请求, 则该数据中心节 点会确定该第一用户设备到该数据中心节点的传输链路的跳数、该第一用户 设备到该数据中心节点的传输延时和该数据中心节点的负载。 然后, 该数据 中心节点会将该传输链路的跳数、该传输延时以及该数据中心节点的负载信 息发送给该第一用户设备, 以便于该第一用户设备根据该传输链路的跳数、 该传输延时以及该数据中心节点的负载信息确定是否从该数据中心节点中 获取数据。
具体来说, 该数据中心节点在基于节能优化策略, 确定每个热度等级对 应的镜像节点时, 可以基于第一时间粒度, 统计用户设备所请求的数据的点 击率。 然后该数据中心节点根据点击率将存储的数据划分为不同的热度等 级, 利用热度等级反映数据的热度。 该数据中心节点基于节能优化策略, 确 定每个热度等级对应的镜像节点。 不同的热度等级可以对应于不同的镜像节 点。 具体来说, 该数据中心节点在确定镜像节点时, 可以为不同的热度等级 确定不同数量的镜像节点。 例如, 对应于高热度等级的镜像节点多于对应于 低热度等级的镜像节点。 这样, 用户设备可以有更多的途径获取高热度等级 的数据。 同时, 由于高热度等级的数据的备份增加, 高热度等级的数据的安 全性也同时得到提升。
进一步, 图 6所示的方法还包括: 确定其他数据中心节点到该数据中心 节点的传输链路的跳数和传输延时,其中该其他数据中心节点是该数据中心 节点所在的数据中心系统中除该数据中心节点以外的数据中心节点。在此情 况下, 该基于节能优化策略, 确定每个热度等级对应的镜像节点, 包括: 根 据该传输链路的跳数和传输延时, 确定该每个热度等级对应的镜像节点。
具体地, 该根据该传输链路的跳数和传输延时, 确定每个热度等级对应 的镜像节点, 包括: 根据该传输链路的跳数和传输延时, 采用用于优化最小 平均跳数和最小平均传输延时的算法, 确定每个热度等级对应的镜像节点。
可选的, 作为一个实施例, 该用于优化最小平均跳数和最小平均传输延 时的算法可以是用于优化最小平均跳数和最小平均传输延时的线性规划算 法。 在此情况下, 该第一数据中心节点可以采用用于优化最小平均跳数和最 小平均传输延时的线性规划算法, 确定每个热度等级对应的镜像节点。 该线 性规划算法的目标方程以节能为目的,使得用户设备能够高效地从镜像节点 上获取需要的数据。在数据传输过程中的最小平均跳数和最小平均传输延时 为优化对象, 从而达到节能高效的目的。 也就是说, 该线性规划算法的限制 方程主要包括流量转移方程和镜像节点数目的限制。数据传输过程中的最小 平均跳数和最小平均传输延时是根据该 N 个数据中心节点中除该第一数据 中心节点以外的数据中心节点到该第一数据中心节点的传输链路的跳数的 平均值和传输延时的平均值。
可选的, 作为另一个实施例, 该用于优化最小平均跳数和最小平均传输 延时的算法可以是用于优化最小平均跳数和最小平均传输延时的基因算法。 在此情况下, 该第一数据中心节点可以采用用于优化最小平均跳数和最小平 均传输延时的基因算法, 确定每个热度等级对应的镜像节点。 该基因算法中 适度计算函数是由数据传输的最小平均跳数和最小平均传输延时来共同决 定的。 该基因算法以节能为目的, 使得用户设备能够高效地从镜像节点上获 取需要的数据。数据传输过程中的最小平均跳数和最小平均传输延时是根据 该 N 个数据中心节点中除该第一数据中心节点以外的数据中心节点到该第 一数据中心节点的传输链路的跳数的平均值和传输延时的平均值。
图 7是根据本发明实施例提供的提供数据服务的方法的示意性流程图。 图 7所示的方法可以由区域网络节点执行, 其中该区域网络节点包括用于存 储数据的緩存装置。执行图 7所示的方法的区域网络节点可以是图 6所示的 区域网给节点。
701 , 确定每个目标数据的优先级, 其中该目标数据为用户设备请求的 保存在至少一个数据中心节点并且没有保存在该区域网络节点的数据。
702, 基于该每个目标数据的优先级, 更新该区域网络节点保存的数据。 图 7所示方法可以根据用户设备所请求的数据, 实时更新自身的緩存装 置中所保存的数据, 以便区域网络节点所保存的数据是用户设备点击率高的 数据。 用户设备可以快速的从区域网络节点获取需要的数据, 而不需要从数 据中心节点获取数据。 这样, 可以加快用户设备获取数据的速度, 提升用户
的体验。
进一步, 该目标数据基于用于用户最小平均跳数和最小平均传输延时的 算法保存在至少一个数据中心节点。 (这个同上一个问题 )
进一步, 图 7所示的方法还可以包括: 确定该每个目标数据的热度等级 和该每个目标数据的传输链路的跳数。 在此情况下, 该确定每个目标数据的 优先级, 包括: 根据该每个目标数据的热度等级和该每个目标数据的传输链 路的跳数, 确定该每个目标数据的优先级。
具体来说, 该区域网络节点可以确定每个存储目标数据的数据中心节点 到该第一区域网络节点的传输链路的跳数。 然后, 该区域网络节点可以根据 每个目标数据的热度等级和每个目标数据的传输链路的跳数,确定每个目标 数据的优先级。 例如, 该区域网络节点可以采用以下公式确定目标数据的优 先级:
Ρ(ή = ρορ (ή * χΨο + hop(i) * y9o , 公式 1.5 其中, P(i)表示第 i个目标数据的优先级, pop(i)表示第 i个目标数据的 热度等级, hop(i)表示第 i个目标数据的传输链路的跳数, 表示目标数据 的热度等级占目标数据的优先级的权重, y%表示目标数据的传输链路的跳 数占目标数据的优先级的权重。本领域技术人员可以理解 _1%和}^可以根据 需要进行设计。 例如, 如果希望使热度等级占优先级的权重大于传输链路的 跳数占优先级的权重,则可以使 大于) ^。。如果希望使热度等级占优先级 的权重小于传输链路的跳数占优先级的权重, 则可以是) ^大于 x%。
进一步, 图 7所示的方法还可以包括: 确定该区域网络节点保存的数据 的优先级。 在此情况下, 该基于该每个目标数据的优先级, 更新该区域网络 节点保存的数据, 包括: 基于该每个目标数据的优先级、 该区域网络节点的 剩余空间以及该区域网络节点保存的数据的优先级, 更新该区域网络节点保 存的数据。
具体来说, 该区域网络节点在确定了每个目标数据的优先级后, 对目标 数据按照优先级进行排列,依次判断是否需要将该目标数据存储到该区域网 络节点的緩存装置中。
以目标数据中优先级最高的目标数据 (为了筒便描述, 以下称为第一目 标数据)为例, 该区域网络节点会判断緩存装置中是否具有足够的剩余空间 存储该第一目标数据。如果该緩存装置有足够的剩余空间存储该第一目标数
据, 则该区域网络节点将该第一目标数据存储到该区域网络节点的緩存装置 中。
如果该緩存装置中没有足够的剩余空间存储该第一目标数据, 则该区域 网络节点还确定该緩存装置中的每个数据的优先级,根据该每个目标数据的 优先级、 该緩存装置中的每个数据的优先级以及该緩存装置的剩余空间, 更 新该緩存装置中的数据。 具体来说, 该区域网络节点首先确定出该緩存装置 中优先级最低的数据 (为了筒便描述, 以下筒称为第一緩存数据)。 如果第 一緩存数据的优先级高于该第一目标数据的优先级, 则该区域网络节点不将 该第一目标数据存储到该緩存装置中。如果该第一緩存数据的优先级低于该 第一目标数据的优先级, 则该区域网络节点确定删除该第一緩存数据的优先 级后该緩存装置中的剩余空间是否大于该第一目标数据所占用的存储空间。 如果删除该第一緩存数据后的剩余空间大于该第一目标数据所占用的空间, 则将该第一緩存数据删除并将该第一目标数据存储到该緩存装置中。
为了帮助更好的理解本发明, 图 8所示的实施例将以数据中心节点和区 域网络节点组成的数据中心系统的方式来对本发明进行介绍。但是该数据中 心系统中的数据中心节点和区域网络节点两种实体在实施本发明方案时可 以独立执行各自的流程步骤。
图 8是根据本发明实施例提供的提供数据服务的方法的示意性流程图。 图 8所示的方法中的第一数据中心节点和第一区域网络节点属于同一个数据 中心系统。 该数据中心系统可以包括 N个数据中心节点和 M个区域网络节 点,其中该 M个区域网络节点中的每一个区域网络节点都具有緩存装置, N、 M为正整数。 该第一数据中心节点是该 N个数据中心节点中的任一个数据 中心节点。 该第一区域网络节点是该 M个区域网络节点中的任一个区域网 络节点。 可选的, 该数据中心系统还可以包括一个或多个不具有緩存装置的 区域网给节点。
801 , 第一数据中心节点基于第一时间粒度统计第一数据中心节点保存 的每个数据的点击率, 根据该每个数据的点击率, 确定该每个数据的热度等 级, 基于节能优化策略, 采用用于优化最小平均跳数和最小平均传输延时的 算法确定每个热度等级对应的镜像节点,将该第一数据中心节点保存的数据 复制到该保存的数据的热度等级对应的镜像节点,其中该镜像节点是该 N个 数据中心节点中除该第一数据中心节点以外的任一个数据中心节点。
802, 第一区域网络节点确定每个目标数据的优先级, 基于该每个目标 数据的优先级, 更新该第一区域网络节点的緩存装置中保存的数据, 其中该 目标数据为用户设备请求的保存在至少一个数据中心节点并且该第一区域 网络节点的緩存装置中没有保存的数据。
根据图 8所示的方法, 用户设备可以快速地从区域网络节点获取到热度 高的数据 (即点击率高的数据)。 此外, 用户设备还可以从多个数据中心节 点中获取到需要的数据, 并且获取到需要数据的效率较高。 因此, 本发明实 施例所提供方法能够节约用户设备获取数据时所需的能耗。
具体来说, 用户设备在需要获取该数据中心网络中所保存的数据时, 可 以向与该用户设备连接的区域网络节点发送数据请求, 该数据请求用于请求 获取该数据中心网络所保存的数据。在该区域网络节点不具有保存数据的緩 存装置的情况下, 或者在该区域网络节点的保存数据的緩存装置中没有该数 据请求消息所请求的数据的情况下,该区域网络节点会将该用户设备发送的 数据请求转发给数据中心节点列表内的所有数据中心节点所有数据中心节 点, 其中该数据中心节点列表保存在该区域网络节点中。 这样, 各个数据中 心节点可以根据用户设备的数据请求,将该数据请求对应的数据发送给该用 户设备。 在此情况下, 第一数据中心节点可以基于第一时间粒度, 统计用户 设备所请求的数据的点击率, 然后根据点击率将数据划分为不同的热度等 级, 利用热度等级反映数据的热度。 第一数据中心节点还可以基于节能优化 策略, 确定每个热度等级对应的镜像节点。 不同的热度等级可以对应于不同 数量的镜像节点。对应于高热度等级的镜像节点多于对应于低热度等级的镜 像节点。 这样, 用户设备可以有更多的途径获取高热度等级的数据。 同时, 由于高热度等级的数据的备份增加, 高热度等级的数据的安全性也同时得到 提升。 相应的, 该第一数据中心节点还可以接收并保存该数据中心系统中的 其他数据中心节点发送的数据。
具体地,该第一数据中心节点可以根据该 N个数据中心节点中除该第一 数据中心节点以外的数据中心节点到该第一数据中心节点的传输链路的跳 数和传输延时,采用用于优化最小平均跳数和最小平均传输延时的窜发确定 该每个热度等级对应的镜像节点。
可选的, 作为一个实施例, 该用于优化最小平均跳数和最小平均传输延 时的算法可以是用于优化最小平均跳数和最小平均传输延时的线性规划算
法。 在此情况下, 该第一数据中心节点可以采用用于优化最小平均跳数和最 小平均传输延时的线性规划算法, 确定每个热度等级对应的镜像节点。 该线 性规划算法的目标方程以节能为目的,使得用户设备能够高效地从镜像节点 上获取需要的数据。在数据传输过程中的最小平均跳数和最小平均传输延时 为优化对象, 从而达到节能高效的目的。 也就是说, 该线性规划算法的限制 方程主要包括流量转移方程和镜像节点数目的限制。数据传输过程中的最小 平均跳数和最小平均传输延时是根据该 N 个数据中心节点中除该第一数据 中心节点以外的数据中心节点到该第一数据中心节点的传输链路的跳数的 平均值和传输延时的平均值。
可选的, 作为另一个实施例, 该用于优化最小平均跳数和最小平均传输 延时的算法可以是用于优化最小平均跳数和最小平均传输延时的基因算法。 在此情况下, 该第一数据中心节点可以采用用于优化最小平均跳数和最小平 均传输延时的基因算法, 确定每个热度等级对应的镜像节点。 该基因算法中 适度计算函数是由数据传输的最小平均跳数和最小平均传输延时来共同决 定的。 该基因算法以节能为目的, 使得用户设备能够高效地从镜像节点上获 取需要的数据。数据传输过程中的最小平均跳数和最小平均传输延时是根据 该 N 个数据中心节点中除该第一数据中心节点以外的数据中心节点到该第 一数据中心节点的传输链路的跳数的平均值和传输延时的平均值。
该第一数据中心节点还可以将第一用户设备到该第一数据中心的传输 链路的跳数、该第一用户设备到该第一数据中心的传输延时和该第一数据中 心节点的负载信息发送给该第一用户设备, 其中该第一用户设备为请求该第 一数据中心节点保存的数据的用户设备。
具体来说, 用户设备向该数据中心系统中的区域网络节点发送数据请 求。如果接收到该数据请求的区域网络节点的緩存装置中没有该数据请求所 请求的数据, 则该区域网络节点会将该用户设备发送的数据请求转发给数据 中心节点列表内的所有数据中心节点所有数据中心节点, 其中该数据中心节 点列表保存在该区域网络节点中。 为了筒便描述, 如果一个用户设备发送的 数据请求所请求的数据不在区域网络节点的緩存装置中,则称这个用户设备 为第一用户设备。 如果该第一数据中心节点接收到该数据请求, 则该第一数 据中心节点会确定该第一用户设备到该第一数据中心节点传输链路的跳数、 该第一用户设备到该第一数据中心节点的传输延时和该第一数据中心节点
的负载。 然后, 该第一数据中心会将该传输链路的跳数、 该传输延时以及该 第一数据中心的负载信息发送给该第一用户设备。 类似的, 该数据中心系统 中的其他数据中心节点也可以接收到该第一用户设备发送的数据请求并向 该第一用户设备发送各自到该第一用户设备的传输链路的跳数、传输延时以 及各自的负载信息。 第一用户设备可以接收到多个数据中心节点发送的传输 链路的跳数、 传输延时以及负载信息。 第一用户设备可以根据传输链路的跳 数、 传输延时以及负载信息, 选择合适的数据中心节点来进行数据传输。 例 如, 可以采用以下公式确定数据中心节点的等级:
L(i) = hop (i) * a% + delay (i) * b9o + workload (i) * c% , 公式 1.6 其中, L(i)表示第 i个数据中心节点的等级, hop(i)表示第 i个数据中心 节点到该第一用户设备的传输链路的跳数, delay (i)表示第 i个数据中心节点 到该第一用户设备的传输延时, workload(i)表示第 i个数据中心节点的负载, a%、 b%和 c%分别表示传输链路的跳数、 传输延时以及负载占数据中心节 点的等级的权重。 本领域技术人员可以理解, a%、 /?%和^"%可以根据需要 进行设计。例如,如果希望传输链路的跳数对数据中心节点的等级影响最大, 则可以使 a%的取值大于 %和 c%。 如果希望传输延时对数据中心节点的等 级影响最大, 则可以使 %的取值大于 α%和 c%。
第一区域网络节点可以确定每个目标数据的热度等级和每个目标数据 的传输链路的跳数,根据该每个目标数据的热度等级和每个目标数据的传输 链路的跳数, 确定每个目标数据的优先级。
具体来说, 第一区域网络节点可以确定每个存储目标数据的数据中心节 点到该第一区域网络节点的传输链路的跳数。 然后, 第一区域网络节点根据 每个目标数据的热度等级和每个目标数据的传输链路的跳数,确定每个目标 数据的优先级。 例如, 可以采用以下公式确定目标数据的优先级:
P(i) = pop(iYx% + hop{i) * y% , 公式 1.7 其中, P(i)表示第 i个目标数据的优先级, pop(i)表示第 i个目标数据的 热度等级, hop(i)表示第 i个目标数据的传输链路的跳数, 表示目标数据 的热度等级占目标数据的优先级的权重, y%表示目标数据的传输链路的跳 数占目标数据的优先级的权重。本领域技术人员可以理解 _1%和}^可以根据 需要进行设计。 例如, 如果希望使热度等级占优先级的权重大于传输链路的 跳数占优先级的权重,则可以使 大于 y%。如果希望使热度等级占优先级
的权重小于传输链路的跳数占优先级的权重, 则可以是}^。大于 X%。
第一区域网络节点在确定了每个目标数据的优先级后,对目标数据按照 优先级进行排列,依次判断是否需要将该目标数据存储到该第一区域网络节 点的緩存装置中。
以目标数据中优先级最高的目标数据 (为了筒便描述, 以下称为第一目 标数据)为例, 第一区域网络节点会判断緩存装置中是否具有足够的剩余空 间存储该第一目标数据。如果该緩存装置有足够的剩余空间存储该第一目标 数据, 则将该第一目标数据存储到该第一区域网络节点的緩存装置中。
如果该緩存装置中没有足够的剩余空间存储该第一目标数据, 则该第一 区域网络节点还可以确定该緩存装置中的每个数据的优先级,根据该每个目 标数据的优先级、该緩存装置中的每个数据的优先级以及该緩存装置的剩余 空间, 更新该緩存装置中的数据。 具体来说, 该第一区域网络节点首先确定 出该緩存装置中优先级最低的数据 (为了筒便描述, 以下筒称为第一緩存数 据)。 如果第一緩存数据的优先级高于该第一目标数据的优先级, 则不将该 第一目标数据存储到该緩存装置中。如果该第一緩存数据的优先级低于该第 一目标数据的优先级, 则确定删除该第一緩存数据的优先级后该緩存装置中 的剩余空间是否大于该第一目标数据所占用的存储空间。如果删除该第一緩 存数据后的剩余空间大于该第一目标数据所占用的空间, 则将该第一緩存数 据删除并将该第一目标数据存储到该緩存装置中。
本领域普通技术人员可以意识到, 结合本文中所公开的实施例描述的各 示例的单元及算法步骤, 能够以电子硬件、 或者计算机软件和电子硬件的结 合来实现。 这些功能究竟以硬件还是软件方式来执行, 取决于技术方案的特 定应用和设计约束条件。 专业技术人员可以对每个特定的应用来使用不同方 法来实现所描述的功能, 但是这种实现不应认为超出本发明的范围。
所属领域的技术人员可以清楚地了解到, 为描述的方便和筒洁, 上述描 述的系统、 装置和单元的具体工作过程, 可以参考前述方法实施例中的对应 过程, 在此不再赘述。
在本申请所提供的几个实施例中, 应该理解到, 所揭露的系统、 装置和 方法, 可以通过其它的方式实现。 例如, 以上所描述的装置实施例仅仅是示 意性的, 例如, 所述单元的划分, 仅仅为一种逻辑功能划分, 实际实现时可 以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个
系统, 或一些特征可以忽略, 或不执行。 另一点, 所显示或讨论的相互之间 的耦合或直接耦合或通信连接可以是通过一些接口, 装置或单元的间接耦合 或通信连接, 可以是电性, 机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作 为单元显示的部件可以是或者也可以不是物理单元, 即可以位于一个地方, 或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或 者全部单元来实现本实施例方案的目的。
另外, 在本发明各个实施例中的各功能单元可以集成在一个处理单元 中, 也可以是各个单元单独物理存在, 也可以两个或两个以上单元集成在一 个单元中。
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使 用时, 可以存储在一个计算机可读取存储介质中。 基于这样的理解, 本发明 的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部 分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质 中, 包括若干指令用以使得一台计算机设备(可以是个人计算机, 服务器, 或者网络设备等)或处理器(processor )执行本发明各个实施例所述方法的 全部或部分步骤。 而前述的存储介质包括: U盘、 移动硬盘、 只读存储器 ( ROM , Read-Only Memory ), 随机存取存储器 (RAM , Random Access Memory )、 磁碟或者光盘等各种可以存储程序代码的介质。
以上所述, 仅为本发明的具体实施方式, 但本发明的保护范围并不局限 于此, 任何熟悉本技术领域的技术人员在本发明揭露的技术范围内, 可轻易 想到的变化或替换, 都应涵盖在本发明的保护范围之内, 因此本发明的保护 范围应以权利要求的保护范围为准。
Claims
1、 一种数据中心节点, 其特征在于, 所述数据中心节点包括: 存储单元, 用于保存数据;
控制单元, 用于基于第一时间粒度统计的所述存储单元保存的每个数据 的点击率, 根据所述每个数据的点击率, 确定所述每个数据的热度等级, 并 且基于节能优化策略, 确定每个热度等级对应的镜像节点;
通信单元, 用于将所述存储单元保存的数据复制到与所述保存的数据的 热度等级对应的镜像节点。
2、 如权利要求 1所述的数据中心节点, 其特征在于,
所述控制单元,还用于确定其他数据中心节点到所述数据中心节点的传 输链路的跳数和传输延时, 其中所述其他数据中心节点是所述数据中心节点 所在的数据中心系统中除所述数据中心节点以外的数据中心节点;
所述控制单元, 具体用于基于节能优化策略, 根据所述传输链路的跳数 和传输延时, 确定所述每个热度等级对应的镜像节点。
3、 如权利要求 2所述的数据中心节点, 其特征在于,
所述控制单元, 具体用于根据所述传输链路的跳数和传输延时, 采用用 于优化最小平均跳数和最小平均传输延时的算法,确定所述每个热度等级对 应的镜像节点。
4、 如权利要求 1至 3中任一项所述的数据中心节点, 其特征在于, 所述通信单元,还用于将第一用户设备到所述数据中心节点的传输链路 的跳数、所述第一用户设备到所述数据中心节点的传输延时和所述数据中心 节点的负载信息发送给所述第一用户设备, 其中所述第一用户设备为请求所 述数据中心节点保存的数据的用户设备。
5、 一种区域网络节点, 其特征在于, 所述区域网络节点包括: 存储单元, 用于保存数据;
控制单元, 用于确定每个目标数据的优先级, 并且基于所述每个目标数 据的优先级, 更新所述存储单元保存的数据, 其中, 所述目标数据为用户设 备请求的保存在至少一个数据中心节点且没有保存在所述存储单元的数据。
6、 如权利要求 5所述的区域网络节点, 其特征在于,
所述控制单元, 具体用于确定所述每个目标数据的热度等级和所述每个 目标数据的传输链路的跳数,根据所述每个目标数据的热度等级和所述每个
目标数据的传输链路的跳数, 确定所述每个目标数据的优先级。
7、 如权利要求 5或 6所述的区域网络节点, 其特征在于, 所述每个目 标数据基于用于优化最小平均跳数和最小平均传输延时的算法保存在至少 一个数据中心节点。
8、 如权利要求 5-7中任一项所述的区域网络节点, 其特征在于, 所述控制单元, 还用于确定所述存储单元保存的数据的优先级; 所述控制单元, 具体用于基于所述每个目标数据的优先级、 所述存储单 元的剩余空间以及所述存储单元保存的数据的优先级, 更新所述存储单元保 存的数据。
9、 一种提供数据服务的方法, 其特征在于, 所述方法由数据中心节点 执行, 所述方法包括:
基于第一时间粒度统计所述数据中心节点保存的每个数据的点击率; 根据所述每个数据的点击率, 确定所述每个数据的热度等级; 基于节能优化策略, 确定每个热度等级对应的镜像节点;
将所述数据中心节点保存的数据复制到与所述保存的数据的热度等级 对应的镜像节点。
10、 如权利要求 9所述的方法, 其特征在于, 所述方法还包括: 确定其 他数据中心节点到所述数据中心节点的传输链路的跳数和传输延时, 其中所 述其他数据中心节点是所述数据中心节点所在的数据中心系统中除所述数 据中心节点以外的数据中心节点,
其中, 所述确定每个热度等级对应的镜像节点, 包括:
根据所述传输链路的跳数和传输延时,确定所述每个热度等级对应的镜 像节点。
11、 如权利要求 10所述的方法, 其特征在于,
所述基于节能优化策略, 根据所述传输链路的跳数和传输延时, 确定所 述每个热度等级对应的镜像节点, 包括:
根据所述传输链路的跳数和传输延时,采用用于优化最小平均跳数和最 小平均传输延时的算法, 确定所述每个热度等级对应的镜像节点。
12、 如权利要求 9至 11 中任一项所述的方法, 其特征在于, 所述方法 还包括:
将第一用户设备到所述数据中心节点的传输链路的跳数、所述第一用户 设备到所述数据中心节点的传输延时和所述数据中心节点的负载信息发送 给所述第一用户设备,其中所述第一用户设备为请求所述数据中心节点保存 的数据的用户设备。
13、 一种提供数据服务的方法, 其特征在于, 所述方法由区域网络节点 执行, 所述方法包括:
确定每个目标数据的优先级, 其中所述目标数据为用户设备请求的保存 在至少一个数据中心节点且没有保存在所述存储单元的数据;
基于所述每个目标数据的优先级, 更新所述区域网络节点保存的数据。
14、 如权利要求 13所述的方法, 其特征在于, 所述方法还包括: 确定所述每个目标数据的热度等级和所述每个目标数据的传输链路的 跳数,
其中, 所述确定每个目标数据的优先级, 包括:
根据所述每个目标数据的热度等级和所述每个目标数据的传输链路的 跳数, 确定所述每个目标数据的优先级。
15、 如权利要求 13或 14所述的方法, 其特征在于, 所述每个目标数据 基于用于优化最小平均跳数和最小平均传输延时的算法保存在至少一个数 据中心节点。
16、 如权利要求 13至 15中任一项所述的方法, 其特征在于, 所述方法 还包括:
确定所述区域网络节点保存的数据的优先级,
其中, 所述基于所述每个目标数据的优先级, 更新所述区域网络节点保 存的数据, 包括:
基于所述每个目标数据的优先级、所述区域网络节点的剩余空间以及所 述区域网络节点保存的数据的优先级, 更新所述区域网络节点保存的数据。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201480000179.9A CN105191259B (zh) | 2014-03-03 | 2014-03-03 | 提供数据服务的方法和网络设备 |
PCT/CN2014/072796 WO2015131310A1 (zh) | 2014-03-03 | 2014-03-03 | 提供数据服务的方法和网络设备 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2014/072796 WO2015131310A1 (zh) | 2014-03-03 | 2014-03-03 | 提供数据服务的方法和网络设备 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2015131310A1 true WO2015131310A1 (zh) | 2015-09-11 |
Family
ID=54054325
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2014/072796 WO2015131310A1 (zh) | 2014-03-03 | 2014-03-03 | 提供数据服务的方法和网络设备 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN105191259B (zh) |
WO (1) | WO2015131310A1 (zh) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110995866B (zh) * | 2019-12-20 | 2022-11-08 | 深圳市网心科技有限公司 | 节点调度方法、装置、调度服务器及存储介质 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1604642A (zh) * | 2004-11-04 | 2005-04-06 | 复旦大学 | 一种广播视频节目系统中信息发布优先级排列的方法 |
CN101668046A (zh) * | 2009-10-13 | 2010-03-10 | 成都市华为赛门铁克科技有限公司 | 资源缓存方法、资源获取方法及其装置、系统 |
CN102508902A (zh) * | 2011-11-08 | 2012-06-20 | 西安电子科技大学 | 云存储系统中可变分块大小的块数据分块方法 |
WO2014022413A1 (en) * | 2012-07-30 | 2014-02-06 | Huawei Technologies Co., Ltd. | Method for peer to peer cache forwarding |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5699712B2 (ja) * | 2011-03-17 | 2015-04-15 | ソニー株式会社 | メモリ制御装置、メモリ装置、メモリ制御方法、及びプログラム |
CN102930056B (zh) * | 2012-11-21 | 2016-11-23 | 华为技术有限公司 | 云存储系统的检索方法及设备 |
-
2014
- 2014-03-03 WO PCT/CN2014/072796 patent/WO2015131310A1/zh active Application Filing
- 2014-03-03 CN CN201480000179.9A patent/CN105191259B/zh active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1604642A (zh) * | 2004-11-04 | 2005-04-06 | 复旦大学 | 一种广播视频节目系统中信息发布优先级排列的方法 |
CN101668046A (zh) * | 2009-10-13 | 2010-03-10 | 成都市华为赛门铁克科技有限公司 | 资源缓存方法、资源获取方法及其装置、系统 |
CN102508902A (zh) * | 2011-11-08 | 2012-06-20 | 西安电子科技大学 | 云存储系统中可变分块大小的块数据分块方法 |
WO2014022413A1 (en) * | 2012-07-30 | 2014-02-06 | Huawei Technologies Co., Ltd. | Method for peer to peer cache forwarding |
Also Published As
Publication number | Publication date |
---|---|
CN105191259A (zh) | 2015-12-23 |
CN105191259B (zh) | 2019-09-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11431791B2 (en) | Content delivery method, virtual server management method, cloud platform, and system | |
US9692849B2 (en) | Method and node entity for enhancing content delivery network | |
CN103150347B (zh) | 基于文件热度的动态副本管理方法 | |
WO2018000993A1 (zh) | 一种分布式存储的方法和系统 | |
WO2017097059A1 (zh) | 分布式数据库系统及其自适应方法 | |
CN105357246B (zh) | 基于信息中心网络的缓存方法和系统 | |
JP2016517080A (ja) | 複製書き込みのためのトークンベースの承認制御 | |
KR20120072907A (ko) | 오브젝트를 복수 개의 데이터 노드들의 위치에 기반하여 분산 저장하는 분산 저장 시스템 및 그 위치 기반 분산 저장 방법 및 컴퓨터에 의하여 독출 가능한 저장 매체 | |
WO2015196686A1 (zh) | 一种数据存储方法以及一种数据存储管理服务器 | |
JPWO2004063928A1 (ja) | データベース負荷軽減システムおよび負荷軽減プログラム | |
JP2021515295A (ja) | 分散ストレージ・ネットワークにおける動的認可バッチ処理 | |
JP2011215794A (ja) | 分散ストレージシステム及びプログラム | |
CN106648456A (zh) | 基于用户访问量以及预测机制的动态副本文件访问方法 | |
JP5845877B2 (ja) | 情報処理装置、データ制御方法およびデータ制御プログラム | |
KR20100073154A (ko) | 메타데이터 서버, 데이터 서버의 데이터 처리 방법 및 이를이용한 비대칭 클러스터 분산 파일 시스템 | |
JP2016048498A (ja) | キャッシュ制御装置およびキャッシュ制御方法 | |
JP2009122981A (ja) | キャッシュ配置方法 | |
JP2014229088A (ja) | データ処理システム、データ処理装置および記憶媒体 | |
WO2015131310A1 (zh) | 提供数据服务的方法和网络设备 | |
JP2019153189A (ja) | セッション情報セットをキャッシュするサーバ、及び、セッション情報セットのキャッシュ制御方法 | |
JP6035934B2 (ja) | データストア管理装置、データ提供システム及びデータ提供方法 | |
Jayalakshmi et al. | Dynamic data replication strategy in cloud environments | |
KR20200072128A (ko) | 라이브 서비스를 위한 분산 파일 시스템 및 파일 관리 방법 | |
JP7318899B2 (ja) | 二次ストレージにコンテンツアイテムを格納するためのシステムおよび方法 | |
Shoaib et al. | Fast Data Access through Nearest Location‐Based Replica Placement |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201480000179.9 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14884723 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14884723 Country of ref document: EP Kind code of ref document: A1 |