Cassandra Practice Test

Which of the following structures are accessed directly from disk?

1. MemTable

2. SStable

3. Commit log

4. Partition index

The correct answer is: SStable

The correct answer relates to the SStable, which stands for Sorted String Table. SStables are the primary data structure used by Apache Cassandra to store data on disk. When data is written to Cassandra, it is initially written to the MemTable (in memory) and, upon reaching a certain threshold, the MemTable is flushed to disk in the form of an SStable. This means that SStables represent a persistent version of the data that is stored directly on disk. SStables are designed for efficient read operations, as they are immutable once created. They are organized in a way that allows for fast retrieval of data, using the sorted order of keys. When a read request is made, Cassandra will access the SStable directly from disk to fetch the requested data, which makes it a fundamental component in the disk storage architecture of Cassandra. In contrast, the MemTable is a memory-based structure and not directly accessed from disk, as it exists in memory for speedier write operations. The Commit log, while crucial for ensuring durability of writes, also resides on disk as a separate entity and is not accessed in the manner SStables are when querying data. Lastly, the Partition index is a structure that helps locate data within an SStable