Secondary indexes (GSI, LSI) in AWS - Step-by-Step Execution

Process Flow - Secondary indexes (GSI, LSI)

Create Main Table with Primary Key and LSI

↓

Add Global Secondary Index (GSI)

↓

Write Data to Table

↓

Data Automatically Indexed in LSI and GSI

↓

Query Using Primary Key or Secondary Index

This flow shows creating a main table with LSI, adding GSI, writing data, and querying using these indexes.

Execution Sample

AWS

Create DynamoDB table with PK userId and LSI on orderDate
Add GSI on attribute status
Write item {userId:1, orderDate:'2024-06-01', status:'shipped'}
Query by GSI status='shipped'

This example creates a table with primary key userId and LSI, adds GSI, writes an item, and queries using the GSI.

Process Table

Step	Action	Index Affected	Resulting Index State	Query Capability
1	Create table with PK userId and LSI on orderDate	Table, LSI	Table created with primary key userId and LSI userId + orderDate	Query by userId or userId + orderDate range
2	Add GSI on status	GSI	GSI created with status as PK	Query by userId + orderDate range or status independently
3	Write item {userId:1, orderDate:'2024-06-01', status:'shipped'}	Table, LSI, GSI	Item indexed in table, LSI, and GSI	Item retrievable by userId, userId+orderDate, or status
4	Query by GSI status='shipped'	GSI	Returns items with status shipped	Query successful using GSI
5	Query by LSI userId=1 and orderDate='2024-06-01'	LSI	Returns items matching userId and orderDate	Query successful using LSI
6	Query by primary key userId=1	Primary Key	Returns item with userId=1	Query successful using primary key
7	Attempt to add LSI after table creation	LSI	Error: LSI can only be added at table creation	Operation fails
8	Attempt to query GSI with non-indexed attribute	GSI	No results	Query returns empty
9	End of operations	N/A	All indexes created and used correctly	Execution complete

💡 Execution stops after all index operations and queries are demonstrated.

Status Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	Final
Table	None	Created with PK userId and LSI on orderDate	Same	Contains 1 item	Same	Contains 1 item
LSI	None	Created on orderDate	Same	Indexes item by userId+orderDate	Same	Exists with 1 indexed item
GSI	None	None	Created on status	Indexes item by status	Same	Exists with 1 indexed item
Item Count	0	0	0	1	1	1

Key Moments - 3 Insights

Why can't we add a Local Secondary Index (LSI) after the table is created?

LSIs must be defined when the table is created because they share the same partition key as the table and affect the table's storage structure. This is shown in execution_table step 7 where adding LSI later causes an error.

How does a Global Secondary Index (GSI) differ in querying compared to the primary key?

A GSI can have a different partition key and sort key, allowing queries on attributes other than the primary key. Execution_table steps 4 and 6 show querying by GSI and primary key returning different results.

What happens to data when you write an item to a table with LSI and GSI?

The item is automatically indexed in the main table, the LSI, and the GSI if it has the indexed attributes. Execution_table step 3 shows the item being indexed in all relevant indexes.

Visual Quiz - 3 Questions

Look at the execution_table at step 3. After writing the item, which indexes contain the new item?

A Only the main table

B Main table, LSI, and GSI

C Main table and LSI only

D Main table and GSI only

💡 Hint

Check the 'Index Affected' and 'Resulting Index State' columns at step 3.

At which step does querying by the GSI 'status' attribute happen?

A Step 4

B Step 2

C Step 6

D Step 8

💡 Hint

Look for the step mentioning 'Query by GSI status' in the 'Action' column.

If you try to add an LSI after table creation, what will happen according to the execution_table?

A LSI is added successfully

B Table is deleted

C Error occurs and operation fails

D GSI is added instead

💡 Hint

Refer to step 7 in the execution_table for the result of adding LSI after creation.

Concept Snapshot

Secondary Indexes in DynamoDB:
- LSI shares the table's partition key, adds a sort key, must be created with the table.
- GSI has its own partition and optional sort key, can be added anytime.
- Both indexes speed up queries on non-primary key attributes.
- Data is automatically copied to indexes when written.
- LSI max 5 per table; GSI max 20 per table.

Full Transcript

This visual execution trace shows how secondary indexes work in AWS DynamoDB. First, a table is created with a primary key and LSI. Then, a Global Secondary Index (GSI) is added with a different partition key. When data is written, it is automatically indexed in the main table, LSI, and GSI if it contains the indexed attributes. Queries can then use the primary key, LSI, or GSI to retrieve data efficiently. The trace also shows that LSIs cannot be added after table creation, while GSIs can. This helps beginners understand how indexes improve query flexibility and performance in DynamoDB.