Method, electronic device, and computer program product for selecting backup destination
Implementations of the present disclosure provide a method, an electronic device, and a computer program product for selecting a backup destination. One method includes: receiving device information about storage devices in a storage device set, wherein a backup task is executed in the storage device set; receiving backup information about the backup task; acquiring a destination association relationship, wherein the destination association relationship describes an association relationship between a reference backup task in a reference storage device set and a reference backup destination of the reference backup task, the reference backup destination including a group of storage devices in a reference storage system; and selecting a backup destination for the backup task from the storage device set according to the destination association relationship and based on the device information and the backup information, the backup destination including a group of storage devices in the storage device set.
1. A method including:
receiving device information about storage devices in a storage device set, wherein a backup task is executed in the storage device set;
receiving backup information about the backup task;
training a machine learning system to determine at least one network model characterizing a destination association relationship, wherein the destination association relationship describes an association relationship between a reference backup task in a reference storage device set and a reference backup destination of the reference backup task, the reference backup destination including a group of storage devices in a reference storage system;
selecting a backup destination for the backup task from the storage device set according to the destination association relationship and based on the device information and the backup information, the backup destination including a group of storage devices in the storage device set; and
executing the backup task utilizing the selected backup destination.
2. The method according to claim 1 , wherein receiving the device information and the backup information further includes: receiving the device information and the backup information that are within a preset time period.
3. The method according to claim 1 , wherein the device information includes, for each of one or more of the storage devices in the storage device set, at least any one of the following:
a position of the storage device;
an available storage space of the storage device;
a network bandwidth of the storage device;
a CPU usage rate of the storage device;
a memory usage rate of the storage device; and
an exhaustion time of the storage device.
4. The method according to claim 1 , wherein the backup information includes at least any one of the following:
a number of backup copies specified by the backup task;
a size of source data specified by the backup task; and
a repetition rate of the source data.
5. The method according to claim 1 , wherein training the machine learning system to determine at least one network model characterizing the destination association relationship includes:
determining reference backup information about the reference backup task executed in the reference storage device set;
determining reference device information about each reference storage device in the reference storage device set; and
training the destination association relationship based on the reference backup information, the reference device information, and the reference backup destination of the reference backup task.
6. The method according to claim 5 , wherein the destination association relationship includes:
a first network model based on a convolutional neural network, wherein the first network model is used to map the reference backup information and the reference device information to an internal feature vector; and
a second network model based on a long short-term memory network, wherein the second network model is used to map the internal feature vector to the reference backup destination of the reference backup task.
7. The method according to claim 1 , wherein determining the backup destination includes:
mapping the backup information and the device information to an internal feature vector based on a first network model; and
mapping the internal feature vector to the backup destination based on a second network model.
8. The method according to claim 1 , wherein determining the backup destination further includes verifying the backup destination in response to the backup destination satisfying the following conditions:
a distance between any two storage devices in the group of storage devices that are in the storage device set and included in the backup destination is greater than a threshold distance;
an available resource amount of any storage device in the group of storage devices that are in the storage device set and included in the backup destination is greater than a threshold resource amount; and
a global balance degree associated with the backup destination is higher than a threshold balance degree, wherein the global balance degree indicates a usage balance degree of the storage device set in a situation where a storage device in the backup destination is used for the backup task.
9. The method according to claim 1 , wherein the storage device set and the reference storage device set satisfy at least any one of the following:
having same or similar numbers of storage devices; and
having same or similar device models.
10. The method according to claim 1 , wherein a number of copies specified by the backup task is not higher than a number of copies specified by the reference backup task.
11. An electronic device, including:
at least one processor; and
at least one memory storing computer program instructions, wherein the at least one memory and the computer program instructions are configured to cause, together with the at least one processor, the electronic device to perform actions, the actions including:
receiving device information about storage devices in a storage device set, wherein a backup task is executed in the storage device set;
receiving backup information about the backup task;
training a machine learning system to determine at least one network model characterizing a destination association relationship, wherein the destination association relationship describes an association relationship between a reference backup task in a reference storage device set and a reference backup destination of the reference backup task, the reference backup destination including a group of storage devices in a reference storage system;
selecting a backup destination for the backup task from the storage device set according to the destination association relationship and based on the device information and the backup information, the backup destination including a group of storage devices in the storage device set; and
executing the backup task utilizing the selected backup destination.
12. The device according to claim 11 , wherein receiving the device information and the backup information further includes: receiving the device information and the backup information that are within a preset time period.
13. The device according to claim 11 , wherein the device information includes, for each of one or more of the storage devices in the storage device set, at least any one of the following:
a position of the storage device;
an available storage space of the storage device;
a network bandwidth of the storage device;
a CPU usage rate of the storage device;
a memory usage rate of the storage device; and
an exhaustion time of the storage device.
14. The device according to claim 11 , wherein the backup information includes at least any one of the following:
a number of backup copies specified by the backup task;
a size of source data specified by the backup task; and
a repetition rate of the source data.
15. The device according to claim 11 , wherein training the machine learning system to determine at least one network model characterizing the destination association relationship includes:
determining reference backup information about the reference backup task executed in the reference storage device set;
determining reference device information about each reference storage device in the reference storage device set; and
training the destination association relationship based on the reference backup information, the reference device information, and the reference backup destination of the reference backup task.
16. The device according to claim 15 , wherein the destination association relationship includes:
a first network model based on a convolutional neural network, wherein the first network model is used to map the reference backup information and the reference device information to an internal feature vector; and
a second network model based on a long short-term memory network, wherein the second network model is used to map the internal feature vector to the reference backup destination of the reference backup task.
17. The device according to claim 11 , wherein determining the backup destination includes:
mapping the backup information and the device information to an internal feature vector based on a first network model; and
mapping the internal feature vector to the backup destination based on a second network model.
18. The device according to claim 11 , wherein determining the backup destination further includes verifying the backup destination in response to the backup destination satisfying the following conditions:
a distance between any two storage devices in the group of storage devices that are in the storage device set and included in the backup destination is greater than a threshold distance;
an available resource amount of any storage device in the group of storage devices that are in the storage device set and included in the backup destination is greater than a threshold resource amount; and
a global balance degree associated with the backup destination is higher than a threshold balance degree, wherein the global balance degree indicates a usage balance degree of the storage device set in a situation where a storage device in the backup destination is used for the backup task.
19. The device according to claim 11 , wherein the storage device set and the reference storage device set satisfy at least any one of the following:
having same or similar numbers of storage devices; and
having same or similar device models;
and further wherein a number of copies specified by the backup task is not higher than a number of copies specified by the reference backup task.
20. A computer program product tangibly stored on a non-volatile computer-readable medium and including machine-executable instructions, wherein the machine-executable instructions, when executed, cause a machine to perform steps of a method, the method including:
receiving device information about storage devices in a storage device set, wherein a backup task is executed in the storage device set;
receiving backup information about the backup task;
training a machine learning system to determine at least one network model characterizing a destination association relationship, wherein the destination association relationship describes an association relationship between a reference backup task in a reference storage device set and a reference backup destination of the reference backup task, the reference backup destination including a group of storage devices in a reference storage system;
selecting a backup destination for the backup task from the storage device set according to the destination association relationship and based on the device information and the backup information, the backup destination including a group of storage devices in the storage device set; and
executing the backup task utilizing the selected backup destination.