No abstract

No abstract

Summary Serverless computing has rapidly grown following the launch of Amazon's Lambda platform. Function‐as‐a‐Service (FaaS) a key enabler of serverless computing allows an application to be decomposed into simple, standalone functions that are executed on a FaaS platform. The FaaS platform is responsible for deploying and facilitating resources to the functions. Several of today's cloud applications spread over heterogeneous connected computing resources and are highly dynamic in their structure and resource requirements. However, FaaS platforms are limited to homogeneous clusters and homogeneous functions and do not account for the data access behavior of functions before scheduling. We introduce an extension of FaaS to heterogeneous clusters and to support heterogeneous functions through a network of distributed heterogeneous target platforms called Function Delivery Network (FDN). A target platform is a combination of a cluster of homogeneous nodes and a FaaS platform on top of it. FDN provides Function‐Delivery‐as‐a‐Service (FDaaS), delivering the function to the right target platform. We showcase the opportunities such as varied target platform's characteristics, possibility of collaborative execution between multiple target platforms, and localization of data that the FDN offers in fulfilling two objectives: Service Level Objective (SLO) requirements and energy efficiency when scheduling functions by evaluating over five distributed target platforms using the FDNInspector, a tool developed by us for benchmarking distributed target platforms. Scheduling functions on an edge target platform in our evaluation reduced the overall energy consumption by 17× without violating the SLO requirements in comparison to scheduling on a high‐end target platform.

With changing workloads, cloud service providers can leverage vertical container scaling (adding/removing resources) so that Service Level Objective (SLO) violations are minimized and spare resources are maximized. In this paper, we investigate a solution to the self-adaptive problem of vertical elasticity for co-located containerized applications. First, the system learns performance models that relate SLOs to workload, resource limits and service level indicators. Second, it derives limits that meet SLOs and minimize resource consumption via a combination of optimization and restricted brute-force search. Third, it vertically scales containers based on the derived limits. We evaluated our technique on a Kubernetes private cloud of 8 nodes with three deployed applications. The results registered two SLO violations out of 16 validation tests; acceptably low derivation times facilitate realistic deployment. Violations are primarily attributed to application specifics, such as garbage collection, which require further research to be circumvented.

No abstract