1Y1-200 · Question #76
If an administrator does NOT create a highly available SQL environment for a XenDesktop deployment, what will happen when the database goes down?
The correct answer is A. An new connections CANNOT be brokered, but existing connections are maintained. See the full explanation below for the reasoning.
Question
Options
- AAn new connections CANNOT be brokered, but existing connections are maintained
- BNew connections can be brokered, but existing desktop connections are dropped
- CExisting connections are maintained, and new connections are still brokered to end users
- DExisting connections to desktops are dropped, and new connections CANNOT be brokered
How the community answered
(60 responses)- A78% (47)
- B13% (8)
- C5% (3)
- D3% (2)
Community Discussion
13
Mei-Ling H.May 3, 2026The answer is A. The key thing to understand here is that the SQL database is in the brokering path, not the session maintenance path. When the database goes offline, the Delivery Controller Broker Service loses the ability to look up machine catalogs, delivery groups, and user assignments, so it cannot make new connection decisions. Existing sessions, though, are a direct communication between the Citrix Receiver on the client side and the VDA on the desktop or server, so the database being down does not interrupt those established connections. Options B and D are both wrong because they say existing connections get dropped, which would only happen if the VDA or the ICA connection itself failed, not the SQL database. Option C is the trap answer if you misread the qualifier "still brokered," because without the database the Broker Service is essentially blind to any new requests.
Giselle K.May 4, 2026Local host cache kicks in on the Delivery Controller before the Broker Service goes fully blind, so in current CVAD builds the "no new connections" window is narrower than it used to be unless LHC is explicitly disabled.
Nina C.Mar 25, 2026The answer is A, new connections cannot be brokered but existing connections stay up. The database is what the Delivery Controller needs to broker new sessions, so once it goes down that process stops, but sessions already established do not depend on the database actively and keep running.
Yusuf A.Mar 26, 2026Nina is right that existing sessions survive, though it is worth adding that if the database stays down long enough the Delivery Controllers will eventually lose the ability to perform load evaluations and zone preference lookups too, so the impact grows the longer the outage runs.
Giselle K.Apr 14, 2026A is right, but do you know WHY existing sessions survive the DB going down?
Jovan I.Mar 13, 2026Went with D first because losing the data store sounded catastrophic across the board, and I was thinking about all the session metadata the Delivery Controller is constantly writing back. What snapped me out of it was remembering that once an ICA session is brokered and the VDA handoff is done, the ongoing traffic flows directly between the client and the VDA, so the broker database is not in the data path for an established session. The Controller needs SQL to make a new brokering decision, look up entitlements, pick a machine, all of that, so new connections die when the database does, but your already-running sessions keep going until something else kills them.
Viktor S.Mar 16, 2026You nailed the core mechanism, though worth adding that session persistence also depends on whether the VDA can still reach its registration controller, because if controller communication drops long enough the VDA deregisters and the session dies even without a new brokering attempt.
Viktor S.Apr 1, 2026A is right. D tempts you but the broker losing DB contact kills new sessions only, existing ICA connections stay alive.
Nina C.Apr 4, 2026Yes, and the thing that helped it click for me is that ICA is a direct client-to-VDA connection once it is up, so the broker dropping its DB link only breaks the front door, not the sessions that already got through it.
Brenda K.Apr 21, 2026Quick win on this one, do not overthink it and do not let D steal your time. D sounds dramatic and dramatic usually sells on test day, but Citrix keeps existing ICA sessions alive through what the docs call "local host cache" or connection leasing depending on your version, so the desktop sessions themselves stay up even when the data store is gone. The broker is the piece that breaks, meaning no new session tickets get issued, which is exactly what A says. B is the trap for people who have the logic exactly backwards, and C is the fantasy where nothing breaks at all, which you should eliminate in about two seconds. Mark this one, bubble A, move on, and spend your saved seconds on a question that actually deserves them.
Isolde G.Apr 24, 2026Thought D at first, but the broker handles reconnects in memory, so A is right.
Orla P.May 10, 2026A is the one, and the reason makes sense once you think about what the database actually does during an active session. The broker uses the database to make decisions about where to send a new connection, so without it, new requests just stall out at the Delivery Controller. What the database does NOT do is carry your live ICA session traffic, so a user who is already on their desktop stays on their desktop because that connection runs directly between the client and the VDA. D is the trap here because it sounds dramatic and complete, but dropping existing sessions would require something to actively terminate them, and a database outage does not do that.
Yusuf A.Apr 7, 2026I almost picked D, but the broker handles new sessions so existing ones stay alive.