VMware's performance engineers observed, under certain conditions, that ESXi IO (in versions 6.x – 7.0) with some NFS servers experienced unexpectedly low read throughput in the presence of extremely low packet loss, due to an undesirable TCP interaction between the ESXi host and the NFS server.
In this paper, Kinson Ho explains how this TCP interaction leads to poor ESXi NFS read performance, describes ways to determine whether this interaction is occurring in an environment, and presents a workaround for ESXi 7.0 that could improve performance significantly when this interaction is detected.
