App-V 5: Revisiting Registry Staging

Last year, I wrote a blog post on registry staging in App-V and how it can affect initial publishing times – particularly on VDI (https://blogs.technet.com/b/gladiatormsft/archive/2013/09/25/app-v-on-registry-staging-and-how-it-can-affect-vdi-environments.aspx) and many people wrote me to tell me it was very helpful in reducing publishing times. This was especially important especially for those non-persistent VDI and RDS/XenApp environments.

A colleague at Microsoft recently reminded me of an alleged “controversy” regarding our registry staging configuration recommendations within current recommended practices that has grown out there in the webosphere. As App-V 5 has changed a lot with service releases and hotfixes over the last two years, have my recommendations changed? Well . . . it depends.

After Service Pack 2 and definitely after Hotfix 4 for Service Pack 2, App-V 5 has improved in terms of publishing time and has been adjusted to work better with user environment management solutions and non-persistent environments.

When questioned about whether the previous blog article is valid or not still post App-V 5 SP2 HF4, customers refer to the TechEd 2014 presentation by Login VSI (https://channel9.msdn.com/Events/TechEd/NorthAmerica/2014/WIN-B362#fbid=) on App-V 5 performance in VDI environments. [This was an excellent presentation by Login VSI and prior to the presentation, Ment van der Plas and I discussed the findings prior to the presentation.] Customers are basically asking me, “Are they right or are you right?”

We are both right – within the right context. My article discussed how disabling background registry staging would reduce the time to complete publishing. The net effect of this means that registry staging would be shifted to on-demand which means the staging begins upon first launch. In essence, we are punting – to use an American football analogy. However, there are no free lunches in nature – and that includes computer science. So to decrease the publishing time, we are moving the burden to the application’s first launch. Could this cause a performance hit on launch? Yes – especially for applications with large registries. Turning back on background registry staging will allow the registry staging to occur prior to first launch which reduces overhead on the application initial launch. This is why Login VSI is right to say this value has a negative effect on first launch.

Am I wrong in the sense that I should have specifically pointed this out more clearly in my article? I will concede that. ;)