> From: Fam Zheng > Sent: Wednesday, December 21, 2016 12:32 AM > > This series adds a new protocol driver that is intended to achieve about 20% > better performance for latency bound workloads (i.e. synchronous I/O) than > linux-aio when guest is exclusively accessing a NVMe device, by talking to the > device directly instead of through kernel file system layers and its NVMe > driver. >
Curious... if the NVMe device is exclusively owned by the guest, why not directly passing through to the guest? is it a tradeoff between performance (better than linux-aio) and composability (snapshot and live migration which not supported by direct passthrough)? Thanks Kevin