The $\texttt{dir}$ directory is already in memory.
From it, the system learns:
$\texttt{dir1}$ starts at cluster $48$.
So cluster $\boxed{48}$ must be read to obtain the contents of directory $\texttt{dir1}$.
From that directory, the system learns:
$\texttt{file1}$ starts at cluster $100$.
Now locate byte $5000$.
One cluster contains:
$4\text{ KB}=4096$ bytes.
Therefore:
- bytes $1$ through $4096$ are in the first cluster
- byte $5000$ is in the second cluster
The first cluster of $\texttt{file1}$ is $100$.
Normally, with ordinary linked allocation, block $100$ would have to be read to discover the next block.
But this system uses FAT, and the FAT is already in memory.
The system simply looks up:
$FAT[100]=106$
without performing a disk read.
Therefore it directly reads cluster $: \boxed{106}$
which contains byte $5000$.
So the only disk clusters actually accessed are: $\boxed{48,\ 106}$
Answer : B
