A multi-tenant oriented performance monitoring, detecting and scheduling architecture based on SLA

Abstract

Software as a Service (SaaS) is thriving as a new mode of service delivery and operation with the development of network technology and the maturity of application software. SaaS application providers offer services for multiple tenants through the “single-instance multi-tenancy” model, which can effectively reduce service costs due to scale effect. Meanwhile, the providers allocate resources according to the SLA signed with tenants to meet the different needs of service quality. However, the service quality of some tenants will be affected by some abnormal consumption of system resources since both hardware and software resources are shared by tenants. In order to deal with this issue, we proposed a multi-tenant oriented monitoring, detecting and scheduling architecture based on SLA for performance isolation. It would monitor service quality of per tenant, discover abnormal status and dynamically adjust the use of resources based on quantization of SLA parameters to ensure the full realization of SLA tasks.