how to check if email and username is unique in dynamodb table design - nosql

Question

I am creating a leads table where each lead has a id ( uuid) , which is the patrtision key and createdAt the rangekey . i need to implement duplicate check each time i insert new item to leads table
on running the following command
C:\Users\user>aws dynamodb describe-table --table-name serverless-isd-app-leads-dev --endpoint-url http://localhost:8008
this is the result shown
{
"Table": {
"AttributeDefinitions": [
{
"AttributeName": "id",
"AttributeType": "S"
},
{
"AttributeName": "email",
"AttributeType": "S"
},
{
"AttributeName": "phone",
"AttributeType": "S"
},
{
"AttributeName": "createdAt",
"AttributeType": "N"
}
],
"TableName": "serverless-isd-app-leads-dev",
"KeySchema": [
{
"AttributeName": "id",
"KeyType": "HASH"
},
{
"AttributeName": "createdAt",
"KeyType": "RANGE"
}
],
"TableStatus": "ACTIVE",
"CreationDateTime": "2021-12-18T20:54:11.940000+05:30",
"ProvisionedThroughput": {
"LastIncreaseDateTime": "1970-01-01T05:30:00+05:30",
"LastDecreaseDateTime": "1970-01-01T05:30:00+05:30",
"NumberOfDecreasesToday": 0,
"ReadCapacityUnits": 1,
"WriteCapacityUnits": 1
},
"TableSizeBytes": 0,
"ItemCount": 0,
"TableArn": "arn:aws:dynamodb:ddblocal:000000000000:table/serverless-isd-app-leads-dev",
"GlobalSecondaryIndexes": [
{
"IndexName": "emai_phone_index",
"KeySchema": [
{
"AttributeName": "email",
"KeyType": "HASH"
},
{
"AttributeName": "phone",
"KeyType": "RANGE"
}
],
"Projection": {
"ProjectionType": "ALL"
},
"IndexStatus": "ACTIVE",
"ProvisionedThroughput": {
"ReadCapacityUnits": 1,
"WriteCapacityUnits": 1
},
"IndexSizeBytes": 0,
"ItemCount": 0,
"IndexArn": "arn:aws:dynamodb:ddblocal:000000000000:table/serverless-isd-app-leads-dev/index/emai_phone_index"
}
]
}
}
and this is my query params
const params = {
TableName: process.env.LEADS_TABLE,
IndexName: "emai_phone_index",
Item: {
id: data.id,
email: data.email,
phone: data.phone,
firstName: data.firstName,
lastName: data.lastName,
createdAt: data.createdAt,
updatedAt: data.updatedAt,
},
ConditionExpression:
"attribute_not_exists(#email) AND attribute_not_exists(#phone)",
ExpressionAttributeNames: {
"#email": "email",
"#phone": "phone",
},
};
but since the email and phone is on global secondary index i am not able to find if its duplicate or not .
How can i solve this ? or is there any other way of designing a better table

Answer 1

Instead of creating global secondary index, you can achieve the uniqueness by adding two extra items of "email" and "phone" at the time of writing new record. For these two records, your primary key will be set to those attribute values (email and phone) separated by hash sign. So writing will involve three records in total and you also have to delete these three records at the time of deletion.
Something similar to this:
'[
{
"Put": {
"TableName" : "serverless-isd-app-leads-dev",
"ConditionExpression": "attribute_not_exists(id)",
"Item" : {
"id":{"S":"b201c1f2-238e-461f-88e6-0e606fbc3c51"},
"email":{"S":"bobby.tables#gmail.com"},
"firstName":{"S":"Bobby"},
"lastName":{"S":"Tables"},
"phone":{"S":"[phone]"}
}
}
},
{
"Put": {
"TableName" : "serverless-isd-app-leads-dev",
"ConditionExpression": "attribute_not_exists(id)",
"Item" : {
"id":{"S":"phone#[phone]"}
}
}
},
{
"Put": {
"TableName" : "serverless-isd-app-leads-dev",
"ConditionExpression": "attribute_not_exists(id)",
"Item" : {
"id":{"S":"email#bobby.tables#gmail.com"}
}
}
}
]'
Please read below link for detailed understanding which has similar use case:
Simulating Amazon DynamoDB unique constraints using transactions