Asynchronous update systems for updating replicated database might result in inconsistent databases. Explain why.

What will be an ideal response?

The resulting schedule might not be serializable because, for example, a transaction, T1 might read two data items, x and y, of a committed transaction T2, but it might read the value of x that T2 updated while it was executing and the value of y, which was a replicated value, before it was updated by the transaction that started after T2 committed.