Access control method, server device, and storage device

摘要

A method of controlling access from a virtual machine executed in a server device to a storage device, the method including acquiring a service level corresponding to a virtual machine identifier that identifies a virtual machine and is output from the virtual machine together with an input/output command by using the virtual machine identifier and by referring to a storage device that stores correspondence relationship between the virtual machine identifier and the service level and executing the input/output command from the virtual machine in accordance with the acquired service level in at least either the server device or the storage device.

主权项

1. A method for controlling access from a virtual machine executed in a server device to a storage device, the method comprising:
acquiring a service level corresponding to a virtual machine identifier, that identifies a virtual machine and is output from the virtual machine together with an input/output command, by using a correspondence relationship between the acquired virtual machine identifier and the service level; acquiring an allowable number corresponding to the acquired service level, the allowable number indicating an upper limit value of input/output commands executable at once in the virtual machine and being determined based on a maximum input/output number that is a maximum value of the number of input/output commands executable at once in an interface connecting between the server device and the storage device; executing the input/output command from the virtual machine within the acquired allowable number in at least either the server device or the storage device, in the storage device,
receiving the input/output command from the server device together with the virtual machine identifier;storing the received input/output command in an input/output queue for each service level of a service level corresponding to the virtual machine identifier and being placed in a waiting state;calculating the allowable number that is a number of input/output commands executable in each of virtual machines;reading out the input/output command stored in the input/output queue for each service level in an order of a queue in the allowable number, and executing an access to a storage device;including two operating modes for each service level including a normal mode in which a particular controller included in the storage device executes an input/output process and a load distribution mode in which a controller other than the particular controller included in the storage device executes the input/output command;monitoring an average processing time period of the input/output command for each virtual machine;when the average processing time period in any of the virtual machines of same service level in the normal mode exceeds a first threshold, switching the operating mode of the same service level to the load distribution mode; andwhen the average processing time period of all the virtual machines of same service level in the load distribution mode falls below a second threshold, switching the operating mode of the same service level to the normal mode.