I have a client who is seeing a recompose operation (5 desktops, View 5.2) running about 15 minutes. They are on aggregated 2GbE links to management, vMotion, and VM Network vSwitches (each separate) with 2 ESXi hosts dedicated to VDI. The storage is FC and over 4Gbps fabric. We see no extreme high utilization or latency on the storage (peak bandwidth of 75MBps and peak latency of 17ms) during the recompose nor do we see it on the network switch ports. We are wondering if 10GbE connections would speed up the recompose significantly or if its a moot point. We cannot find the bottleneck. For the life of me I have searched for days now all over the net for a detailed traffic flow diagram of how the traffic flows during different View operations. A clone is taking about 5 minutes on this same infrastructure and they feel that is way too slow as well so we would really like to understand what links are being used for what traffic during these operations. Thank you for any help or direction you can suggest to find this information.
↧