Methods, systems and apparatus to improve cluster efficiency

A method to improve cluster efficiency, the method comprising:

• generating a candidate cluster specification in response to determining that cluster resources are unavailable;
• identifying cluster resource details to execute a workload, the cluster resource details based on the candidate cluster specification;
• parsing the workload to identify services to be executed by candidate cluster resources corresponding to the cluster resource details, respective ones of the identified services having one or more default values for one or more parameters of the candidate cluster resources;
• identifying service optimization formulas that, when evaluated, calculate values of respective ones of the one or more parameters to execute the respective ones of the identified services on the candidate cluster resources, the service optimization formulas obtained from a database of pre-defined service optimization formulas associated with a plurality of possible services capable of being executed by the candidate cluster resources, the possible services including the identified services;
• in response to a user change to the one or more default values corresponding to the identified services, evaluating one or more corresponding default safety equations with the user change to prevent one or more performance problems when the identified services are executed by the candidate cluster resources;
• and
• evaluating the identified service optimization formulas to calculate the values of the respective ones of the one or more parameters for the candidate cluster resources, the values of the respective ones of the one or more parameters to be applied to the candidate cluster resources to configure the candidate cluster resources prior to runtime execution of the workload.

A tangible computer-readable storage medium comprising instructions which, when executed, cause a processor to at least:

• generate a candidate cluster specification in response to determining that cluster resources are unavailable;
• identify cluster resource details to execute a workload, the cluster resource details based on the candidate cluster specification;
• parse the workload to identify services to be executed by candidate cluster resources corresponding to the cluster resource details, respective ones of the identified services having one or more default values for one or more parameters of the candidate cluster resources;
• identify service optimization formulas that, when evaluated, calculate values of respective ones of the one or more parameters to execute the respective ones of the identified services on the candidate cluster resources, the service optimization formulas obtained from a database of pre-defined service optimization formulas associated with a plurality of possible services capable of being executed with the candidate cluster resources, the possible services including the identified services;
• in response to a user change to the one or more default values corresponding to the identified services, evaluate one or more corresponding default safety equations with the user change to prevent one or more performance problems when the identified services are executed by the candidate cluster resources;
• and
• evaluate the identified service optimization formulas to calculate the values of the respective ones of the one or more parameters for the candidate cluster resources, the values of the respective ones of the one or more parameters to be applied to the candidate cluster resources to configure the candidate cluster resources prior to runtime execution of the workload.

An apparatus comprising:

• at least one tangible computer-readable storage medium;
• instructions;
• a processor to execute the instructions to:
• generate a candidate cluster specification in response to determining that cluster resources are unavailable;
• identify cluster resource details to execute a workload, the cluster resource details based on the candidate cluster specification;
• parse the workload to identify services to be executed by candidate cluster resources corresponding to the cluster resource details, respective ones of the identified services having one or more default values for one or more parameters of the candidate cluster resources;
• identify service optimization formulas that, when evaluated, calculate values of respective ones of the one or more parameters to execute the respective ones of the identified services on the candidate cluster resources, the service optimization formulas obtained from a database of pre-defined service optimization formulas associated with a plurality of possible services capable of being executed with the candidate cluster resources, the possible services including the identified services;
• in response to a user change to the one or more default values corresponding to the identified services, evaluate one or more corresponding default safety equations with the user change to prevent one or more performance problems when the identified services are executed by the candidate cluster resources;
• and
• evaluate the identified service optimization formulas to calculate the values of the respective ones of the one or more parameters for the candidate cluster resources, the values of the respective ones of the one or more parameters to be applied to the candidate cluster resources to configure the candidate cluster resources prior to runtime execution of the workload.