Abstract:
Systems and methods are disclosed for query processing in a big data analytics platform by enumerating plans for a current query using a processor; building a dominance graph for the current query; for each plan, determining a regret value and a score for the plan based on the regret value and cost; and selecting query plans in an online fashion for query processing in big data analytics platforms where intermediate results are materialized and can be reused later.
Abstract:
A method for database consolidation includes generating a model for expected penalty estimation; determining a tenant's value as a function of query arrival rate and SLA penalty; placing a tenant to minimize a total expected cost in the order of the tenant value; and progressively using additional servers to prevent any server from being saturated to guarantee a tenant placement that costs no more than four times the cost of any other placement
Abstract:
Systems and methods are disclosed for query processing in a big data analytics platform by enumerating plans for a current query using a processor; building a dominance graph for the current query; for each plan, determining a regret value and a score for the plan based on the regret value and cost; and selecting query plans in an online fashion for query processing in big data analytics platforms where intermediate results are materialized and can be reused later.
Abstract:
A method for tenant placement in a multi-tenant system is shown that includes creating a weighted graph of tenants and sharing relationships between tenants, where a weight for each tenant and each sharing relationship represents an associated degree of resource consumption; and adding one or more tenants to a database using a processor based on said weighted graph and a database capacity, such that the combined weight of the added tenants and the sharing relationships belonging to the added tenants are within the database capacity. If a tenant cannot be added to the database without exceeding the database capacity, a new database is created and the one or more tenants are added to the new database, subject to a new database capacity. The adding and creating steps are repeated until all tenants have been added to a database.
Abstract:
A method for tenant placement in a multi-tenant system is shown that includes creating a weighted graph of tenants and sharing relationships between tenants, where a weight for each tenant and each sharing relationship represents an associated degree of resource consumption; and adding one or more tenants to a database using a processor based on said weighted graph and a database capacity, such that the combined weight of the added tenants and the sharing relationships belonging to the added tenants are within the database capacity. If a tenant cannot be added to the database without exceeding the database capacity, a new database is created and the one or more tenants are added to the new database, subject to a new database capacity. The adding and creating steps are repeated until all tenants have been added to a database.
Abstract:
A system for fair costing of dynamic data sharing in a cloud market is disclosed. The system uses an online method for sharing plan selection, as well as a set of fair costing criteria and a method that maximizes fairness.
Abstract:
Methods for generating a data fetching plan in a multi-tenant system include placing tenants in consecutively allocated databases according to a weighted graph of tenants and sharing relationships between tenants, where at least one sharing relationship includes multiple accessing tenants accessing a given set of data from a provider tenant. For each sharing relationship, if a current database has one or more accessing tenants and does not have the provider tenant, data is fetched from the latest-allocated database prior to the current database that has accessing tenants, if such a database exists; if a current database has the provider tenant, data is provided to the earliest-allocated database after the current database that has accessing tenants if such a database exists. The fetching and providing steps are repeated for each allocated database.
Abstract:
Systems and methods are disclosed for placing tenants in a cloud based database server, by estimating with a processor a cost of placing a set of tenants on a server using a simulator; estimating a relative importance of different tenant's queries; and selecting an optimal server for each tenant based with a cost-based schedular to maximize profit using the simulator.
Abstract:
Methods for generating a data fetching plan in a multi-tenant system include placing tenants in consecutively allocated databases according to a weighted graph of tenants and sharing relationships between tenants, where at least one sharing relationship includes multiple accessing tenants accessing a given set of data from a provider tenant. For each sharing relationship, if a current database has one or more accessing tenants and does not have the provider tenant, data is fetched from the latest-allocated database prior to the current database that has accessing tenants, if such a database exists; if a current database has the provider tenant, data is provided to the earliest-allocated database after the current database that has accessing tenants if such a database exists. The fetching and providing steps are repeated for each allocated database.