![]() Software-emulated devices appear to the VM guest as an I/O device, but all functionality is handled in the VM implementation. VM hypervisors scale resources through device virtualization. Dynamically scaling up or down based on current workload requirements, and being able to partitioning available physical resources, leads to more efficient utilization in the cluster. In this respect, efficient and dynamic resource sharing and configuration is important as it is desirable to be able to scale up and allocate more resources on demand, or scale down and release them when the resources are no longer needed. Workloads come and go, and even vary themselves with number of users and amount of data to process. The demand for processing power and I/O resources in a cluster may, to a large degree, vary over time. With our added VM and multi-device support, Device Lending offers highly customizable configurations of remote devices that can be dynamically reassigned and shared to optimize resource utilization, thus enabling a flexible composable I/O infrastructure for VMs as well as bare-metal machines. I/O intensive workloads run seamlessly using both local and remote resources. Additionally, we have also implemented multi-device support, enabling shortest-path peer-to-peer transfers between remote devices residing in different hosts.Our experimental results prove that multiple remote devices can be used, achieving bandwidth and latency close to native PCIe, and without requiring any additional support in device drivers. Physical remote devices can be “passed through” to VM guests, enabling direct access to physical resources while still retaining the flexibility of virtualization. In this work, we have extended our solution with support for a virtual machine (VM) hypervisor. ![]() Any PCIe device, such as network cards disks, and GPUs, can easily be shared among the connected hosts. The method has extremely low computing overhead, and does not require any application- or device-specific distribution mechanisms. For PCI Express (PCIe) clusters, we have previously proposed Device Lending as a solution for enabling direct low latency access to remote devices. ![]() This limits the availability of resources to specific devices and drivers that are supported by the virtualization technology being used, as well as what the interconnection technology supports. For computers interconnected in a cluser, access to remote hardware resources often requires facilitation both in hardware and specialized drivers with virtualization support. Modern workloads often exceed the processing and I/O capabilities provided by resource virtualization, requiring direct access to the physical hardware in order to reduce latency and computing overhead.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |