I have a table like this in dynamodb
| name(PK) | key(SK) | description |
|---|---|---|
| read | user | Read User |
| read | account | Read Account |
| write | user | Write User |
I want to list all data from the table, but I don't want to perform a scan on the table. So I added an index on the table allIndex. It's value will always be same(all in example)
| name(PK) | key(SK) | description | allIndex |
|---|---|---|---|
| read | user | Read User | all |
| read | account | Read Account | all |
| write | user | Write User | all |
Now I can query using allIndex to list all data from a table. It's benefit would be all data will reside in the same partition in dynamodb's GSI. I wanted to know if this is a good idea to do so? or will it lead to a hot partition on GSI?
scan is an appropriate option for your use case.
query is preferred over scan when you are filtering a subset of the data in your application. The scan operation can be expensive because it has to read every item in your database. This is a bad idea if you're just fetching a few items. However, if your access pattern requires you to read the entire database, scan is the best option.