… As we illustrate in the next section, platforms often end up with spare resources around its
clusters that are not allocated to any customers. Fixedsize Spot VMs are the current approach …

… resources of a large cloud. • We propose Harvest VMs to harvest unallocated resources. • …
availability of unallocated resources and provide a new SLO for these resources. • We build …

… Our results show that SmartHarvest can harvest a significant amount of CPU resources
without increasing the 99th-percentile tail latency of latency-critical primary workloads by more …

… spare resources, which cloud providers … harvested resources improves efficiency and
reduces cost. Under the same cost budget, running serverless platforms on harvested resources …

… harvesting for … cloud users with low harvesting overhead. To develop BlockFlex, we first
conduct a characterization study of storage resources that could be harvested in a cloud platform. …

… At the top layer resides the cloud with its abundant resources that can also be … cloud
architectures, enabling them to interact with this new edge-centric ecosystem of devices/resources, …

… on-demand allocation, and to ensure fairness, a dynamic quote price mechanism for
edge-cloud resource is designed based on the gametheoretic approach. Moreover, the optimal …

… cloud resources. … resource harvesting in serverless computing, we compare the function
response latency achieved by the default resource manager (Fixed RM) and greedy resource …

… cloud computing platform infrastructure by harvesting underused generic computing resources
… Most of the current cloud computing platforms are mainly based on using a dedicating …

… resource harvesting framework for multi-node serverless clusters. Libra makes precise
harvest… Infrastructure-as-a-Service (IaaS) cloud platforms package harvested resources as low-…