Enabling a cloud to effectively assign workloads to servers

摘要

In accordance with some embodiments, a public infrastructure as a service (IaaS) user can provide a file, to a cloud service provider, with information about the specific instructions and opcodes that may be used in an application run on the cloud service provider's system. This information may be developed at compile time by the user before the user deploys the workload onto the public IaaS cloud. Thus the user has complete control over the information that is provided.

主权项

1. A method comprising:
receiving an application at a server of a plurality of servers, including a first server that is adapted to perform encryption or decryption and a second server that is not adapted to perform encryption or decryption; receiving, by the server, a profile catalog that includes an indication of at least one hardware-accelerated instruction set architecture that is to be present in a server that is to execute the application; reading, at said catalog, a second indication that the application includes instructions or opcodes that require encryption or decryption; reading, by the server, a third indication that is indicative of a time sensitivity of the application; and assigning, by the server, said application to at least one of the plurality of servers as a function of the indication of the hardware-accelerated instruction set architecture included in the profile catalog and the indication of the time sensitivity of the application.