Querying DynamoDB by date

Question

This question concerns AWS' Dynamo DB:

I have a table with a hash key and a range. Also a bunch of items in it.

I am specifically trying  to get all the items that were created after a specific date and sort them  by date. Which is pretty straightforward that is for a relational Database.

This was the problem however, to perform a query i need a hash key.

 by any means is my table schema wrong, now shouldn't the hash key be unique?

Answer 1

DynamoDB allows for specification of secondary indexes to aid in this sort of query. Secondary indexes can either be global, meaning that the index spans the whole table across hash keys, or local meaning that the index would exist within each hash key partition, thus requiring the hash key to also be specified when making the query.

For the use case in this question, you would want to use a global secondary index on the "CreatedAt" field.

This sort of queries are dealt in differently in DynamoDB due to the specification of Secondary indexes.These can be both global and local.

Global the span of index is across whole table. Where local means the hash key will have to be specified when making the query. Yourr problem will be benefited by Global . This is for the sake of reference : https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/SecondaryIndexes.html

Answer 2

There are different query you can run on Dynamo DB. You can find the different queries here:

https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/EMRforDynamoDB.Querying.html?tag=duckduckgo-d-20

Answer 3

Read the blog related to #dynamoDB. This blog covers all the different topics for the different query that can be implemented over Dynamo DB.
https://www.dynamodbguide.com/querying/

Answer 4

You can use the "day" portion in the timestamp because of the hash function and use the full timestamp as the range.

Answer 5

See the different method of querying in DynamoDb here

https://docs.aws.amazon.com/amazondynamodb/latest/developerguide/Query.html

Answer 6

You could use the "day" portion of the timestamp as the hash and use the full timestamp as the range.

Answer 7

You could store the "year-month" portion of the timestamp as the hash key. That would significantly reduce the number of queries required when someone opens their app for the first time in, say, one year — compared to using the "day" portion.
At that point, it also seems that the range key becomes irrelevant, since querying by hash key would bring in all of the items that have been updated since lastRefreshedAt.

Answer 8

You could make the Hash key something along the lines of a 'unique' id, then the range key as a combination of a timestamp with a unique id appended on the end. That way you know the hash key and can still query the date with greater than.

Answer 9

As per my understanding of these is that it should now allow you to perform the desired queries without having a full scan. The downside is that these indexes have to be specified at table creation, and also (I believe) cannot be blank when creating a unique item. So it's not a perfect solution, but a viable alternative, for some.