Exam AWS Certified Developer Associate topic 1 question 140 discussion

The main point in the question is that they want to query table by using EITHER unique identifier OR email address. A and B is not enough if we only has one primary key in a table Assump that we configure base table has partition key is "unique identifier" so that we can query base on this key For B), the LSI index must have the same partition key with base table. So in case the assumption above, you cannot query only email address For D) the GSI can have different partition key and sort key from base table. So we can create partition key is "email address" and sort key is "unique identifier" for GSI. Then we can use only "email address" for query

The partition key must be also decided by the query pattern. The query can happen with the unique if or the email address. So its best to keep the unique id as the partition key of the primary table. However this will not solve the requirement of querying with the email address. So create a GSI with Email as the partition key. This cannot be LSI because the LSI can not have a different partition key, just a different sort key.

To design the indexes we have to study the table. The table is composed by an unique identifier and by an email. We can notice that both of the fields are unique and are optimal to partition the table. So, in both cases we don't need a sort key because will be unuseful (doesn't make sense to order only one row). So we can create the primary key using the userID as partition key. Regarding the second query, doesn't make sense to create a LSI because the partion key is shared with the primary key. Instead, we have to create a GSI by using the email address (also in this case we don't need the sort key because the email is unique). So, D is the correct answer

D - Query for LSI will require to specify the partition key with email since its an index local to that partition and we are required to query by email id which means we are required to retreive records which exist in multiple partitions .

What's the difference between implementing A and D? I thought setting up A you would always get an index, A. For the primary key of the table, specify the unique identifier as the partition key and specify the email address as the sort key.) D. For the primary key of the table, specify the unique identifier as the partition key. Create a global secondary index (GSI) based on the email address.

The correct answer is D. I initially thought it is B but no. LSI is alternate sort key , we want to search using either of the keys, so GSI GSI --> provides you option to define alternate primary key

In general, you should use global secondary indexes rather than local secondary indexes. The exception is when you need strong consistency in your query results, which a local secondary index can provide but a global secondary index cannot (global secondary index queries only support eventual consistency).

Global Secondary index It is D

The Query operation in Amazon DynamoDB finds items based on primary key values.

It is D. At first I thought it was B, but the requirement is to be able to search by one or the other, the unique id or the email. With a local index, you need the primary key as well. Only a global index would allow you to search using only the email (once the primary key is set to the unique id)

Option D - https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/bp-indexes-general.html

We need different primary key in second index so we need global secondary index

question sucks its trying to convey this: GSI can be a single attribute key which is called a partition key or it can be a composite key with two attributes,

I think its D