Method and device for scheduling resources

1. A method for scheduling resources, the method comprising:

traversing a user instance distribution of multiple user instances of a user on multiple hosts in a cluster;

detecting that a trigger condition for user scheduling of adjustment of the user instance distribution is satisfied;

determining one or more first hosts from which one or more user instances are to be migrated out at least based on equilibrium information among the multiple user instances, the equilibrium information including a deviation value of a respective user instance from an equilibrium expectation value of host resources of the multiple hosts in the cluster based on a maximum host resources consumption of the multiple hosts in the cluster and current resource usage of the multiple hosts in the cluster, each of the one or more first hosts comprising at least one user instance to be migrated out; and

migrating the one or more user instances from the one or more first hosts to one or more second hosts.

2. The method of claim 1 , wherein the trigger condition for user scheduling comprises that the user instance distribution includes at least one user instance that is higher than a corresponding threshold information of user instance.

3. The method of claim 1 , wherein the trigger condition for user scheduling comprises that the user instance distribution includes at least one user instance of which deviation value is greater than a corresponding warning threshold information of user instance.

4. The method of claim 1 , wherein the trigger condition for user scheduling comprises that the user instance distribution includes at least one user instance that is greater than an upper limit of user instance equalization.

5. The method of claim 1 , wherein the migrating the one or more user instances from the one or more first hosts to the one or more second hosts comprises:

migrating, based on a priority descending order, the one or more user instances from the one or more first hosts to the one or more second hosts.

6. The method of claim 5 , further comprising:

determining the priority descending order based on at least one of:

quantity information of the one or more user instances to be migrated out; and

resource consumption for the multiple user instances.

7. The method of claim 1 , wherein the one or more second hosts comprise an idle host or a host that has capacity for a respective user instance to be migrated in and corresponds to the respective user instance.

8. The method of claim 1 , wherein the migrating the one or more user instances from the one or more first hosts to the one or more second hosts comprises:

migrating the one or more user instances from the one or more first hosts to the one or more second hosts based on a condition that a host resource consumption is not greater than a corresponding equilibrium expected threshold of resource consumption after the one or more second hosts accept the one or more user instance.

9. The method of claim 1 , further comprising:

detecting whether a deviation value of host resource is greater than a warning information of resource balance in response to determining that the trigger condition for user scheduling is not satisfied; and

determining the one or more first hosts in response to determining that the deviation value of the host resource is greater than the warning information of resource balance.

10. The method of claim 9 , wherein the migrating the one or more user instances from the one or more first hosts to the one or more second hosts comprises:

migrating the one or more user instances from the one or more first hosts to the one or more second hosts based on an order of resource consumption for user instances from high to low.

11. The method of claim 9 , wherein the migrating the one or more user instances from the one or more first hosts to the one or more second hosts n comprises:

migrating the one or more user instances from the one or more first hosts to the one or more second hosts when the user instances satisfy an equalization expectation of the user instances after the one or more user instances are migrated to the one or more second hosts.

12. An apparatus comprising:

one or more processors; and

one or more memories storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform acts comprising

traversing a user instance distribution of user instances of a user on one or more hosts in a cluster;

detecting that a trigger condition for user scheduling of adjustment of the user instance distribution is satisfied;

determining one or more first hosts from which one or more user instances are to be migrated out at least based on equilibrium information among the multiple user instances, the equilibrium information including a deviation value of a respective user instance from an equilibrium expectation value of host resources of the multiple hosts in the cluster, the equilibrium expectation value of the host resources of the multiple hosts in the cluster representing a ratio between a total volume of the resources consumed by the cluster and a maximum host resource consumption of the cluster, each of the one or more first hosts comprising at least one user instance to be migrated out; and

migrating the one or more user instances from the one or more first hosts to one or more second hosts.

13. The apparatus of claim 12 , wherein the trigger condition for user scheduling comprises at least one of:

the user instance distribution includes at least one user instance that is higher than a corresponding threshold information of user instance;

the user instance distribution includes at least one user instance of which deviation value is greater than a corresponding warning threshold information of user instance; and

the user instance distribution includes at least one user instance that is greater than an upper limit of user instance equalization.

14. The apparatus of claim 12 , wherein the migrating the one or more user instances from the one or more first hosts to the one or more second hosts comprises:

migrating, based on a priority descending order, the one or more user instances from the one or more first hosts to the one or more second hosts.

15. The apparatus of claim 14 , wherein the acts further comprise:

determining a priority of the one or more user instance based on at least one of:

quantity information of user instances to be migrated out;

and

resource consumption for the multiple user instances.

16. The apparatus of claim 12 , wherein the one or more second hosts comprises an idle host or a host that is available for a respective user instance to be migrated in and corresponds to the respective user instance.

17. The apparatus of claim 12 , wherein the migrating the one or more user instances from the one or more first hosts to the one or more second hosts comprises:

migrating the one or more user instances from the one or more first hosts to the one or more second hosts when a host resource consumption is not greater than a corresponding equilibrium expected threshold of resource consumption after the one or more second hosts accept the one or more user instances.

18. The apparatus of claim 12 , wherein the acts further comprise:

detecting whether a deviation value of host resource is greater than warning information of resource balance in response to determining that the trigger condition for user scheduling is not satisfied; and

determining the one or more first hosts in response to determining that the deviation value of the host resource is greater than the warning information of resource balance.

19. The apparatus of claim 18 , wherein the migrating the one or more user instances from the one or more first hosts to the one or more second hosts comprises:

migrating the one or more user instances from the one or more first hosts to the one or more second hosts based on an order of resource consumption for user instances from high to low.

20. One or more memories storing computer-executable instructions that, when executed by one or more processors, cause the one or more processors to perform acts comprising:

traversing a user instance distribution of multiple user instances of a user on one or more hosts in a cluster;

detecting that a trigger condition for user scheduling of adjustment of the user instance distribution is satisfied;

determining one or more first hosts from which one or more user instances are to be migrated out at least based on equilibrium information among the multiple user instances, the equilibrium information including a deviation value of a respective user instance from an equilibrium expectation value of host resources of the multiple hosts in the cluster based on a maximum host resources consumption of the multiple hosts in the cluster and current resource usage of the multiple hosts in the cluster, each of the one or more first hosts comprising at least one user instance to be migrated out; and

migrating the one or more user instances from the one or more first hosts to one or more second hosts.