Method, server system and computer program product for managing resources
Embodiments of the present disclosure provide a method, a server system and a computer program product of managing resources. The method may comprise receiving a request for a first amount of resources of the dedicated processing unit from an application with an assigned priority. The method may further comprise determining a total amount of resources of the dedicated processing unit to be occupied by the application based on the request. The method may also comprise in response to the total amount approximating or exceeding a predetermined quota associated with the priority, allocating the first amount of resources of the general-purpose processing unit to the application.
1. A method of managing resources, comprising:
receiving a request for a first amount of resources of a dedicated processing unit from an application with an assigned priority;
determining, based on the request, a total amount of resources of the dedicated processing unit to be occupied by the application;
in response to the total amount approximating or exceeding a predetermined quota associated with the priority, allocating the first amount of resources of a general-purpose processing unit to the application;
in response to the application releasing a portion of resources of the dedicated processing unit occupied previously, determining whether a total amount of resources of the dedicated processing unit occupied by the application after the releasing is below the predetermined quota by a threshold difference; and
in response to the total amount of resources after the releasing being below the predetermined quota by the threshold difference, migrating at least a portion of data occupying the allocated resources of the general-purpose processing unit to the resources of the dedicated processing unit.
2. The method of claim 1 , wherein the resources of the dedicated processing unit include storage resources of a Graphics Processing Unit (GPU), and the resources of the general-purpose processing unit include storage resources of a Central Processing Unit (CPU).
3. The method of claim 1 , wherein the resources of the dedicated processing unit and the resources of the general-purpose processing unit have unified addressing by means of a shared pointer, and allocating the first amount of resources of the general-purpose processing unit comprises:
setting the shared pointer to point to the first amount of resources of the general-purpose processing unit allocated to the application; and
returning information of the pointer to the application.
4. The method of claim 1 , further comprising:
detecting utilization of resources of the dedicated processing unit; and
in response to the utilization exceeding a predetermined threshold, sending a signal to the application to instruct the application to defer sending of a next request for the resources of the dedicated processing unit.
5. The method of claim 1 , further comprising:
receiving an indication of the predetermined quota from the application.
6. The method of claim 1 , wherein the total amount of resources of the dedicated processing unit to be occupied by the application comprises a sum of an amount of resources of the dedicated processing unit already occupied by the application plus the requested first amount of resources of the dedicated processing unit.
7. The method of claim 1 , wherein the application is allocated either the first amount of resources of the dedicated processing unit or the first amount of resources of the general-purpose processing unit based at least in part on whether or not the total amount approximates or exceeds the predetermined quota associated with the priority.
8. A server system, comprising:
a dedicated processing unit;
a general-purpose processing unit; and
a controller configured to:
receive a request for a first amount of resources of the dedicated processing unit from an application with an assigned priority;
determine, based on the request, a total amount of resources of the dedicated processing unit to be occupied by the application;
in response to the total amount approximating or exceeding a predetermined quota associated with the priority, allocate the first amount of resources of the general-purpose processing unit to the application;
in response to the application releasing a portion of resources of the dedicated processing unit occupied previously, determine whether a total amount of resources of the dedicated processing unit occupied by the application after the releasing is below the predetermined quota by a threshold difference; and
in response to the total amount of resources after the releasing being below the predetermined quota by the threshold difference, migrate at least a portion of data occupying the allocated resources of the general-purpose processing unit to the resources of the dedicated processing unit.
9. The server system of claim 8 , wherein the resources of the dedicated processing unit include storage resources of a Graphics Processing Unit (GPU), and the resources of the general-purpose processing unit include storage resources of a Central Processing Unit (CPU).
10. The server system of claim 8 , wherein the resources of the dedicated processing unit and the resources of the general-purpose processing unit have unified addressing by means of a shared pointer, and allocating the first amount of resources of the general-purpose processing unit comprises:
setting the shared pointer to point to the first amount of resources of the general-purpose processing unit allocated to the application; and
returning information of the pointer to the application.
11. The server system of claim 8 , wherein the controller is further configured to:
detect utilization of resources of the dedicated processing unit; and
in response to the utilization exceeding a predetermined threshold, send a signal to the application to instruct the application to defer sending of a next request for the resources of the dedicated processing unit.
12. The server system of claim 8 , wherein the controller is further configured to:
receive an indication of the predetermined quota from the application.
13. The server system of claim 8 , wherein the application is allocated either the first amount of resources of the dedicated processing unit or the first amount of resources of the general-purpose processing unit based at least in part on whether or not the total amount approximates or exceeds the predetermined quota associated with the priority.
14. A computer program product being tangibly stored on a non-volatile computer readable medium and including machine executable instructions which, when executed, cause a machine to perform a method of managing resources, the method comprising:
receiving a request for a first amount of resources of a dedicated processing unit from an application with an assigned priority;
determining, based on the request, a total amount of resources of the dedicated processing unit to be occupied by the application;
in response to the total amount approximating or exceeding a predetermined quota associated with the priority, allocating the first amount of resources of a general-purpose processing unit to the application;
in response to the application releasing a portion of resources of the dedicated processing unit occupied previously, determining whether a total amount of resources of the dedicated processing unit occupied by the application after the releasing is below the predetermined quota by a threshold difference; and
in response to the total amount of resources after the releasing being below the predetermined quota by the threshold difference, migrating at least a portion of data occupying the allocated resources of the general-purpose processing unit to the resources of the dedicated processing unit.
15. The computer program product of claim 14 , wherein the resources of the dedicated processing unit include storage resource of a Graphics Processing Unit (GPU), and the resources of the general-purpose processing unit include storage resources of a Central Processing Unit (CPU).
16. The computer program product of claim 14 , wherein the resources of the dedicated processing unit and the resources of the general-purpose processing unit have unified addressing by means of a shared pointer, and allocating the first amount of resources of the general-purpose processing unit comprises:
setting the shared pointer to point to the first amount of resources of the general-purpose processing unit allocated to the application; and
returning information of the pointer to the application.
17. The computer program product of claim 14 , wherein the method further comprises:
detecting utilization of resources of the dedicated processing unit; and
in response to the utilization exceeding a predetermined threshold, sending a signal to the application to instruct the application to defer sending of a next request for the resources of the dedicated processing unit.
18. The computer program product of claim 14 , wherein the method further comprises:
receiving an indication of the predetermined quota from the application.
19. The computer program product of claim 14 , wherein the total amount of resources of the dedicated processing unit to be occupied by the application comprises a sum of an amount of resources of the dedicated processing unit already occupied by the application plus the requested first amount of resources of the dedicated processing unit.
20. The computer program product of claim 14 , wherein the application is allocated either the first amount of resources of the dedicated processing unit or the first amount of resources of the general-purpose processing unit based at least in part on whether or not the total amount approximates or exceeds the predetermined quota associated with the priority.