| CPC G06F 1/3296 (2013.01) | 14 Claims |

|
1. A control method for a power consumption upper limit of a server, comprising:
determining, according to a server policy, a power consumption upper limit of a management engine (ME) to be a first expected upper limit;
in response to an intelligent platform management interface (IPMI) command, controlling the ME to perform a power consumption limitation on the server according to a first desired upper limit as a current power consumption upper limit of the ME;
acquiring a current actual power consumption of the server;
comparing the current actual power consumption with the first expected upper limit and a second expected upper limit; and
in response to the current actual power consumption being greater than the first expected upper limit and not exceeding the second expected upper limit, lowering the current power consumption upper limit of the ME until the current actual power consumption is lowered by the ME such that the current actual power consumption of the server does not exceed the first expected upper limit;
wherein, the lowering the current power consumption upper limit of the ME until the current actual power consumption is lowered by the ME such that the current actual power consumption of the server does not exceed the first expected upper limit comprises:
updating the current power consumption upper limit of the ME according to a lowering formula, the lowering formula comprising:
an+1=an−en+1,en+1=s(xn−an)+(1−s)en,n=1,2, . . . ;
wherein xn refers to the current actual power consumption of the server at an nth lowering; an refers to the current power consumption upper limit of the ME at the nth lowering; a1 is set at a value equal to the first expected upper limit; en refers to a current allowable error interval value at the nth lowering; e1 is set at a difference value between the second expected upper limit and the first expected upper limit; and s refers to a correction coefficient.
|