Chapter 3-4 UD

What kind of data is ideal for graph databases?

• Simple data
• Data consisting of complex relationships
• Numerical data
• Categorical data

What is the key feature of most NoSQL databases?

• Their use of aggregates
• Their use of relationships
• Their use of SQL queries
• Their use of transactions

Why might you need separate aggregates for orders and customers?

• To reduce the amount of data stored in the database
• To simplify the data model
• Because some applications want to process orders individually
• Because it's a requirement of all NoSQL databases

What is the difference between inter-aggregate relationships and intra-aggregate relationships in aggregate-oriented databases?

• Intra-aggregate relationships are more difficult to handle than inter-aggregate relationships.
• Inter-aggregate relationships are more difficult to handle than intra-aggregate relationships.
• There is no difference between inter-aggregate relationships and intra-aggregate relationships.
• Inter-aggregate relationships are stored in a different column family than intra-aggregate relationships.

How do aggregate-oriented databases handle updates across multiple aggregates?

• They provide ACID guarantees for all updates
• They only support atomicity within a single aggregate
• They automatically handle updates across multiple aggregates
• They treat all aggregates as a single unit of data-retrieval

What is the advantage of removing references to orders from the customer object?

• It allows for fast access to the data.
• It allows for real-time analytics.
• It allows for not updating the customer object when new orders are placed.
• All of the above.

What is the primary driver of interest in NoSQL databases?

• Their ability to scale up by buying bigger servers.
• Their ability to query by attributes inside the document.
• Their ability to run on a large cluster.
• Their ability to optimize read performance.

How can a slave be appointed a new master after a failure of the master?

• The slave can appoint itself as a new master.
• A new master can be appointed manually.
• A new master can be appointed automatically.
• The cluster will continue to function without a master.

What are materialized views?

• Views that are never updated.
• Views that are updated only when needed.
• Views that are updated constantly.
• Views that are only used for incremental map-reduce.

Which distribution option would the author recommend as the first and simplest option?

• Replication
• Sharding
• Peer-to-peer
• Single-server