Citrix Certified Expert - Virtualization: Assessment and Design common mistakes
Citrix Certified Expert - Virtualization: Assessment and Design · Understand the traps and how to avoid them.
Calls a small office-only pilot proof for every workload.
Untested endpoint, graphics, peripheral, location or audio needs are hidden behind a single successful launch.
What to do instead
Segment the users and test representative required capabilities against agreed acceptance measures.
Draws StoreFront or a Controller as the renderer of every established session.
Authentication/enumeration/brokering are confused with the VDA workload and selected HDX transport path.
What to do instead
Draw and verify the control and session paths separately for the actual deployment.
Assumes removing an access-group membership instantly terminates every active session.
Authorization changes, token refresh and already established sessions are treated as one immediate action.
What to do instead
Specify revocation requirements and test each affected layer, including existing sessions and application access.
Equates random assignment with an automatic reset at every logon.
The allocation of machines to users is confused with disk persistence, restart and profile retention.
What to do instead
Document assignment, machine lifecycle and user-state persistence separately and verify each behaviour.
Assumes BDM proves that no DHCP role exists anywhere.
The bootstrap method is confused with how a target obtains its own network configuration.
What to do instead
Describe the bootstrap and target IP configuration independently for the supported implementation.
Claims Studio policy always wins against every domain GPO.
Different policy sources, filters and precedence are collapsed into a single unsupported universal rule.
What to do instead
Inspect applicable sources and resulting settings, then test the intended bounded scope.
Uses normal-load arithmetic as proof of full service after failure.
Surviving capacity and common dependencies are not included in the capacity model.
What to do instead
Calculate the specified failure case and validate compute, storage, control, identity and application dependencies.
Treats replication as sufficient protection against deletion or corruption.
A harmful change can propagate to another copy, leaving no suitable recoverable point.
What to do instead
Keep and verify protected recoverable versions and test both time and data-loss objectives.
Treats Local Host Cache and connection-lease continuity as identical.
Different mechanisms, prerequisites and outage restrictions are replaced with one broad availability promise.
What to do instead
Identify the mechanism and version, test its actual supported outage behaviour and document limits.