Scanning tables in DynamoDB

A Scan operation in Amazon DynamoDB reads every item in a table or a secondary index. By default, a Scan operation returns all of the data attributes for every item in the table or index. You can use the ProjectionExpression parameter so that Scan only returns some of the attributes, rather than all of them.

Scan always returns a result set. If no matching items are found, the result set is empty.

A single Scan request can retrieve a maximum of 1 MB of data. Optionally, DynamoDB can apply a filter expression to this data, narrowing the results before they are returned to the user.

Filter expressions for scan

If you need to further refine the Scan results, you can optionally provide a filter expression. A filter expression determines which items within the Scan results should be returned to you. All of the other results are discarded.

A filter expression is applied after a Scan finishes but before the results are returned. Therefore, a Scan consumes the same amount of read capacity, regardless of whether a filter expression is present.

A Scan operation can retrieve a maximum of 1 MB of data. This limit applies before the filter expression is evaluated.

With Scan, you can specify any attributes in a filter expression—including partition key and sort key attributes.

The syntax for a filter expression is identical to that of a condition expression. Filter expressions can use the same comparators, functions, and logical operators as a condition expression. See Condition and filter expressions, operators, and functions in DynamoDB for more information about logical operators.

aws dynamodb scan \
     --table-name Thread \
     --filter-expression "LastPostedBy = :name" \
     --expression-attribute-values '{":name":{"S":"User A"}}'

Limiting the number of items in the result set

The Scan operation enables you to limit the number of items that it returns in the result. To do this, set the Limit parameter to the maximum number of items that you want the Scan operation to return, prior to filter expression evaluation.

For example, suppose that you Scan a table with a Limit value of 6 and without a filter expression. The Scan result contains the first six items from the table.

Now suppose that you add a filter expression to the Scan. In this case, DynamoDB applies the filter expression to the six items that were returned, discarding those that do not match. The final Scan result contains six items or fewer, depending on the number of items that were filtered.

Paginating the results

DynamoDB paginates the results from Scan operations. With pagination, the Scan results are divided into "pages" of data that are 1 MB in size (or less). An application can process the first page of results, then the second page, and so on.

A single Scan only returns a result set that fits within the 1 MB size limit.

To determine whether there are more results and to retrieve them one page at a time, applications should do the following:

In other words, the LastEvaluatedKey from a Scan response should be used as the ExclusiveStartKey for the next Scan request. If there is not a LastEvaluatedKey element in a Scan response, you have retrieved the final page of results. (The absence of LastEvaluatedKey is the only way to know that you have reached the end of the result set.)

You can use the Amazon CLI to view this behavior. The Amazon CLI sends low-level Scan requests to DynamoDB, repeatedly, until LastEvaluatedKey is no longer present in the results. Consider the following Amazon CLI example that scans the entire Movies table but returns only the movies from a particular genre.

aws dynamodb scan \
    --table-name Movies \
    --projection-expression "title" \
    --filter-expression 'contains(info.genres,:gen)' \
    --expression-attribute-values '{":gen":{"S":"Sci-Fi"}}' \
    --page-size 100  \
    --debug

Ordinarily, the Amazon CLI handles pagination automatically. However, in this example, the Amazon CLI --page-size parameter limits the number of items per page. The --debug parameter prints low-level information about requests and responses.

If you run the example, the first response from DynamoDB looks similar to the following.

2017-07-07 12:19:14,389 - MainThread - botocore.parsers - DEBUG - Response body:
b'{"Count":7,"Items":[{"title":{"S":"Monster on the Campus"}},{"title":{"S":"+1"}},
{"title":{"S":"100 Degrees Below Zero"}},{"title":{"S":"About Time"}},{"title":{"S":"After Earth"}},
{"title":{"S":"Age of Dinosaurs"}},{"title":{"S":"Cloudy with a Chance of Meatballs 2"}}],
"LastEvaluatedKey":{"year":{"N":"2013"},"title":{"S":"Curse of Chucky"}},"ScannedCount":100}'

The LastEvaluatedKey in the response indicates that not all of the items have been retrieved. The Amazon CLI then issues another Scan request to DynamoDB. This request and response pattern continues, until the final response.

2017-07-07 12:19:17,830 - MainThread - botocore.parsers - DEBUG - Response body:
b'{"Count":1,"Items":[{"title":{"S":"WarGames"}}],"ScannedCount":6}'

The absence of LastEvaluatedKey indicates that there are no more items to retrieve.

Counting the items in the results

In addition to the items that match your criteria, the Scan response contains the following elements:

If the size of the Scan result set is larger than 1 MB, ScannedCount and Count represent only a partial count of the total items. You need to perform multiple Scan operations to retrieve all the results (see Paginating the results).

Each Scan response contains the ScannedCount and Count for the items that were processed by that particular Scan request. To get grand totals for all of the Scan requests, you could keep a running tally of both ScannedCount and Count.

Capacity units consumed by scan

You can Scan any table or secondary index. Scan operations consume read capacity units, as follows.

By default, a Scan operation does not return any data on how much read capacity it consumes. However, you can specify the ReturnConsumedCapacity parameter in a Scan request to obtain this information. The following are the valid settings for ReturnConsumedCapacity:

DynamoDB calculates the number of read capacity units consumed based on the number of items and the size of those items, not on the amount of data that is returned to an application. For this reason, the number of capacity units consumed is the same whether you request all of the attributes (the default behavior) or just some of them (using a projection expression). The number is also the same whether or not you use a filter expression. Scan consumes a minimum read capacity unit to perform one strongly consistent read per second, or two eventually consistent reads per second for an item up to 4 KB. If you need to read an item that is larger than 4 KB, DynamoDB needs additional read request units. Empty tables and very large tables which have a sparse amount of partition keys might see some additional RCUs charged beyond the amount of data scanned. This covers the cost of serving the Scan request, even if no data exists.

Read consistency for scan

A Scan operation performs eventually consistent reads, by default. This means that the Scan results might not reflect changes due to recently completed PutItem or UpdateItem operations. For more information, see DynamoDB read consistency.

If you require strongly consistent reads, as of the time that the Scan begins, set the ConsistentRead parameter to true in the Scan request. This ensures that all of the write operations that completed before the Scan began are included in the Scan response.

Setting ConsistentRead to true can be useful in table backup or replication scenarios, in conjunction with DynamoDB Streams. You first use Scan with ConsistentRead set to true to obtain a consistent copy of the data in the table. During the Scan, DynamoDB Streams records any additional write activity that occurs on the table. After the Scan is complete, you can apply the write activity from the stream to the table.

Parallel scan

By default, the Scan operation processes data sequentially. Amazon DynamoDB returns data to the application in 1 MB increments, and an application performs additional Scan operations to retrieve the next 1 MB of data.

The larger the table or index being scanned, the more time the Scan takes to complete. In addition, a sequential Scan might not always be able to fully use the provisioned read throughput capacity: Even though DynamoDB distributes a large table's data across multiple physical partitions, a Scan operation can only read one partition at a time. For this reason, the throughput of a Scan is constrained by the maximum throughput of a single partition.

To address these issues, the Scan operation can logically divide a table or secondary index into multiple segments, with multiple application workers scanning the segments in parallel. Each worker can be a thread (in programming languages that support multithreading) or an operating system process. To perform a parallel scan, each worker issues its own Scan request with the following parameters:

The following diagram shows how a multithreaded application performs a parallel Scan with three degrees of parallelism.

In this diagram, the application spawns three threads and assigns each thread a number. (Segments are zero-based, so the first number is always 0.) Each thread issues a Scan request, setting Segment to its designated number and setting TotalSegments to 3. Each thread scans its designated segment, retrieving data 1 MB at a time, and returns the data to the application's main thread.

The values for Segment and TotalSegments apply to individual Scan requests, and you can use different values at any time. You might need to experiment with these values, and the number of workers you use, until your application achieves its best performance.