on Feb 11, 2019We want to use the VMware version in test / Dev but have production on Gseries nodes.
But from a Test / Dev environment perspective if they tested a HCP code upgrade and it works in VM then it will work on G series?
Any idea what the performance difference might be?
There is no way to give an answer to this question. In a VM environment, we provide guidelines, but there is nothing stopping the end user from utilizing the VM infrastructure in a different manner. For example, if they decide to run all 4 nodes of HCP on the same ESXi Server, the CPU and memory resources will be in high demand. Or if they decide to not put each HCP Node storage on a different data store, all HCP nodes will be competing for the same backend disk I/O.
From an upgrade point of view, this is going down to another level. The HCP install/upgrade software has be be very aware of environment that it is executing on top of. Performing an upgrade in a VM environment is no guarantee that upgrade will succeed on a G-Node environment. Too many variables could come into play.
Are you referring to customer code upgrade and test against the dev HCP instance ? or are you talking about the actual HCP software upgrade ?
From an application point of view, it is the same code. Just don't do any performance characterization and expect the results to be the same on G-Nodes.