[{"id":"dynamodbadvanced-001","courseId":"dynamodb-advanced","domain":"data-modeling","topic":"LSI item collection size limit","prompt":"A team adds a local secondary index to its orders table, partitioned by CustomerId. One customer's items and their index entries keep growing. What is the largest total that customer's item collection can reach?","hint":"Consider where a local index's copies of items are stored compared with the base table.","explain":"A local index sits in the same partition as its base table rows, so one partition key's data cannot be spread across machines. Consider what that implies for the total size of everything sharing that key.","source":"https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/LSI.html","interaction":{"type":"single"},"options":[{"id":"a","text":"10 GB, counting base table items and projected index entries together","why":"Items sharing a partition key, including their projected LSI entries, cannot exceed 10 GB in total."},{"id":"b","text":"10 GB, counting only base table items, since index entries are stored apart","whyNot":"The projected LSI views count toward the 10 GB too, so index entries are not free."},{"id":"c","text":"400 KB, the same limit that applies to one item and its index entries","whyNot":"400 KB is the per-item limit for an item plus its corresponding index entries, not a cap on the whole collection."},{"id":"d","text":"No fixed cap, since a growing collection is split across several partitions","whyNot":"With a local index the collection stays in one partition and cannot be distributed across partitions."}],"answer":"a","explanation":"For a table with one or more LSIs, an item collection (items sharing a partition key, including their projected LSI entries) cannot exceed 10 GB and cannot be split across partitions.","concept":"For a table with one or more LSIs, an item collection (items sharing a partition key, including their projected LSI entries) cannot exceed 10 GB and cannot be split across partitions.","rev":1,"updatedAt":"2026-10-26","kind":"scenario","takeaways":["An LSI shares its base table's partition, so every index entry counts toward one 10 GB collection limit."]},{"id":"dynamodbadvanced-002","courseId":"dynamodb-advanced","domain":"data-modeling","topic":"Materialized graph pattern","prompt":"A social app stores people, friendships and skills in one table using the materialized graph pattern. Which TWO statements about this design are correct?","hint":"Think about what a partition holds, and which key a second index needs to find who points at a person.","explain":"A graph can live in one table when each node and each edge gets its own item under a shared key. Think about how sort keys and secondary indexes let you find relationships from either end.","source":"https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/bp-adjacency-graphs.html","interaction":{"type":"multiple","tellCount":true},"options":[{"id":"a","text":"Each person's partition holds one node item and multiple edge items.","why":"The partition key is the entity ID, and each partition contains one node item plus many edge items."},{"id":"b","text":"A second GSI keyed on TypeTarget finds everyone who lists a given person as a friend.","why":"The second GSI uses TypeTarget as its partition key, so querying Friend-Person-2 returns people who list Person-2 as a friend."},{"id":"c","text":"Node items use the edge type as their sort key, so friendships sort ahead of profile data.","whyNot":"For node items the sort key is the entity's own ID; the edge type and target form the sort key only for edge items."},{"id":"d","text":"Edge items go in a separate table, so the primary table holds only people.","whyNot":"The pattern stores nodes and edges together in a single DynamoDB table."},{"id":"e","text":"A popular skill should stay in one GSI partition so that queries touch only one place.","whyNot":"The page recommends sharding large aggregations across several logical partitions to avoid hot read and write problems."},{"id":"f","text":"Graph summaries should always come from a scheduled EMR job, even when results must be real-time.","whyNot":"A scheduled EMR process suits use cases that are not sensitive to real-time consistency; it is an option, not a requirement."}],"answer":"ab","explanation":"The materialized graph pattern stores nodes and edges in one table: each partition holds a node item and its edge items, and a GSI keyed on TypeTarget supports reverse lookups.","concept":"The materialized graph pattern stores nodes and edges in one table: each partition holds a node item and its edge items, and a GSI keyed on TypeTarget supports reverse lookups.","rev":1,"updatedAt":"2026-10-26","kind":"scenario","takeaways":["One table can hold a whole graph; a GSI on a Type-Target composite key gives reverse lookups."]},{"id":"dynamodbadvanced-003","courseId":"dynamodb-advanced","domain":"data-modeling","topic":"Overloaded GSI attributes","prompt":"With the default quota of 20 global secondary indexes per table, a table can index no more than about 20 different data fields.","hint":"Think about what the quota counts, and what one attribute can hold when items differ.","explain":"An index is defined by a partition key and a sort attribute. If one attribute holds different kinds of content in different items, a single index can answer several questions. What does a quota on indexes bound?","source":"https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/bp-gsi-overloading.html","interaction":{"type":"true-false"},"options":[{"id":"a","text":"True","whyNot":"The quota limits indexes, not the fields they can serve; one overloaded index can cover many kinds of data."},{"id":"b","text":"False","why":"The quota caps indexes, not fields; one index can serve many kinds of data, so far more than 20 fields can be queried."}],"answer":"b","explanation":"The default quota is 20 GSIs per table, but overloading one attribute across item types lets a single index serve many kinds of data, so far more than 20 fields can be indexed.","concept":"The default quota is 20 GSIs per table, but overloading one attribute across item types lets a single index serve many kinds of data, so far more than 20 fields can be indexed.","rev":1,"updatedAt":"2026-10-26","kind":"scenario","takeaways":["A GSI quota limits index count, not how many data fields you can query through overloaded indexes."]},{"id":"dynamodbadvanced-004","courseId":"dynamodb-advanced","domain":"data-modeling","topic":"Multi-attribute GSI keys","prompt":"A developer follows a guide to build a global secondary index with multi-attribute keys on a tournament table. Put these four steps in the order the guide performs them.","hint":"Some steps need the table and its data to exist before they can run.","explain":"Building a multi-attribute index means defining it, loading records, and reading them back. Consider what each kind of work needs in place before it can succeed.","source":"https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/GSI.DesignPattern.MultiAttributeKeys.html","interaction":{"type":"sort","items":[{"text":"Create a table with GSIs that use multi-attribute keys","why":"The table and its indexes are defined first, since nothing can be written to or queried from them before they exist."},{"text":"Insert the items using their native attributes, with no synthetic keys","why":"Items are written next, and the GSI indexes their existing attributes without any concatenated key."},{"text":"Query the GSI with equality on all partition key attributes","why":"Queries begin by pinning every partition key attribute with an equality condition, because the partition must be known."},{"text":"Query the GSI's sort key attributes left to right, without skipping any","why":"Sort key conditions narrow the results from the first attribute onward, and a gap in the middle is rejected."}]},"options":[],"answer":"o:abcd","explanation":"A multi-attribute GSI is built in this order: create the table and its indexes, write items with their natural attributes, query with equality on every partition key attribute, then query sort key attributes left to right.","concept":"A multi-attribute GSI is built in this order: create the table and its indexes, write items with their natural attributes, query with equality on every partition key attribute, then query sort key attributes left to right.","rev":1,"updatedAt":"2026-10-26","kind":"scenario","takeaways":["Query a multi-attribute GSI with equality on every partition key attribute, then sort key attributes left to right."]},{"id":"dynamodbadvanced-005","courseId":"dynamodb-advanced","domain":"data-modeling","topic":"Empty set values","prompt":"A team is designing item attributes for a table, and some attributes are not keys. Which TWO of these values can a non-key attribute hold in DynamoDB?","hint":"Look at which container types are exempt from the rule for empty values, and where key attributes stand.","explain":"Attribute values follow emptiness rules that depend on the value's type and on the role the attribute plays. Think about which container types may hold nothing, and which key roles reject an empty value.","source":"https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/Constraints.html","interaction":{"type":"multiple","tellCount":true},"options":[{"id":"a","text":"An empty String value","why":"Empty String values are allowed in an attribute that is not used as a key for a table or index."},{"id":"b","text":"An empty List value","why":"Empty Lists and Maps are allowed; the rule against emptiness covers Sets of every kind, not these."},{"id":"c","text":"An empty String Set","whyNot":"An attribute value cannot be an empty Set, and that includes String Sets."},{"id":"d","text":"An empty Binary Set","whyNot":"Empty Sets of any kind are rejected as attribute values, Binary Sets included."},{"id":"e","text":"An empty String as the partition key of a GSI","whyNot":"Empty String values are not allowed when the attribute is used as a key attribute for a table or index."},{"id":"f","text":"An empty Binary value as the sort key of a table","whyNot":"Key attributes cannot hold empty Binary or String values, so this is rejected."}],"answer":"ab","explanation":"Empty String and Binary values are allowed in non-key attributes, and empty Lists and Maps are allowed, but an attribute value cannot be an empty Set of any type.","concept":"Empty String and Binary values are allowed in non-key attributes, and empty Lists and Maps are allowed, but an attribute value cannot be an empty Set of any type.","rev":1,"updatedAt":"2026-10-26","kind":"scenario","takeaways":["Empty Sets are never valid attribute values; empty Strings and Binaries cannot be key values."]}]