Performing DynamoDB transactions in AWS AppSync

Note

We now primarily support the APPSYNC_JS runtime and its documentation. Please consider using the APPSYNC_JS runtime and its guides here.

AWS AppSync supports using Amazon DynamoDB transaction operations across one or more tables in a single region. Supported operations are TransactGetItems and TransactWriteItems. By using these features in AWS AppSync, you can perform tasks such as:

Pass a list of keys in a single query and return the results from a table
Read records from one or more tables in a single query
Write records in transaction to one or more tables in an all-or-nothing way
Execute transactions when some conditions are satisfied

Permissions

Like other resolvers, you need to create a data source in AWS AppSync and either create a role or use an existing one. Because transaction operations require different permissions on DynamoDB tables, you need to grant the configured role permissions for read or write actions:


{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Action": [
                "dynamodb:DeleteItem",
                "dynamodb:GetItem",
                "dynamodb:PutItem",
                "dynamodb:Query",
                "dynamodb:Scan",
                "dynamodb:UpdateItem"
            ],
            "Effect": "Allow",
            "Resource": [
                "arn:aws:dynamodb:region:accountId:table/TABLENAME",
                "arn:aws:dynamodb:region:accountId:table/TABLENAME/*"
            ]
        }
    ]
}

Note: Roles are tied to data sources in AWS AppSync, and resolvers on fields are invoked against a data source. Data sources configured to fetch against DynamoDB only have one table specified, to keep configuration simple. Therefore, when performing a transaction operation against multiple tables in a single resolver, which is a more advanced task, you must grant the role on that data source access to any tables the resolver will interact with. This would be done in the Resource field in the IAM policy above. Configuration of the transaction calls against the tables is done in the resolver template, which we describe below.

Data Source

For the sake of simplicity, we’ll use the same data source for all the resolvers used in this tutorial. On the Data sources tab, create a new DynamoDB data source and name it TransactTutorial. The table name can be anything because table names are specified as part of the request mapping template for transaction operations. We will give the table name empty.

We’ll have two tables called savingAccounts and checkingAccounts, both with accountNumber as partition key, and a transactionHistory table with transactionId as partition key.

For this tutorial, any role with the following inline policy will work. Replace region and accountId with your region and account ID:


{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Action": [
                "dynamodb:DeleteItem",
                "dynamodb:GetItem",
                "dynamodb:PutItem",
                "dynamodb:Query",
                "dynamodb:Scan",
                "dynamodb:UpdateItem"
            ],
            "Effect": "Allow",
            "Resource": [
                "arn:aws:dynamodb:region:accountId:table/savingAccounts",
                "arn:aws:dynamodb:region:accountId:table/savingAccounts/*",
                "arn:aws:dynamodb:region:accountId:table/checkingAccounts",
                "arn:aws:dynamodb:region:accountId:table/checkingAccounts/*",
                "arn:aws:dynamodb:region:accountId:table/transactionHistory",
                "arn:aws:dynamodb:region:accountId:table/transactionHistory/*"
            ]
        }
    ]
}

Transactions

For this example, the context is a classic banking transaction, where we’ll use TransactWriteItems to:

Transfer money from saving accounts to checking accounts
Generate new transaction records for each transaction

And then we’ll use TransactGetItems to retrieve details from saving accounts and checking accounts.

Warning

TransactWriteItems is not supported when used with conflict detection and resolution. These settings must be disabled to prevent possible errors.

We define our GraphQL schema as follows:


type SavingAccount {
    accountNumber: String!
    username: String
    balance: Float
}

type CheckingAccount {
    accountNumber: String!
    username: String
    balance: Float
}

type TransactionHistory {
    transactionId: ID!
    from: String
    to: String
    amount: Float
}

type TransactionResult {
    savingAccounts: [SavingAccount]
    checkingAccounts: [CheckingAccount]
    transactionHistory: [TransactionHistory]
}

input SavingAccountInput {
    accountNumber: String!
    username: String
    balance: Float
}

input CheckingAccountInput {
    accountNumber: String!
    username: String
    balance: Float
}

input TransactionInput {
    savingAccountNumber: String!
    checkingAccountNumber: String!
    amount: Float!
}

type Query {
    getAccounts(savingAccountNumbers: [String], checkingAccountNumbers: [String]): TransactionResult
}

type Mutation {
    populateAccounts(savingAccounts: [SavingAccountInput], checkingAccounts: [CheckingAccountInput]): TransactionResult
    transferMoney(transactions: [TransactionInput]): TransactionResult
}

schema {
    query: Query
    mutation: Mutation
}

TransactWriteItems - Populate Accounts

In order to transfer money between accounts, we need to populate the table with the details. We’ll use the GraphQL operation Mutation.populateAccounts to do so.

In the Schema section, click on Attach next to the Mutation.populateAccounts operation. Go to VTL Unit Resolvers, then choose the same TransactTutorial data source.

Now use the following request mapping template:

Request Mapping Template


#set($savingAccountTransactPutItems = [])
#set($index = 0)
#foreach($savingAccount in ${ctx.args.savingAccounts})
    #set($keyMap = {})
    $util.qr($keyMap.put("accountNumber", $util.dynamodb.toString($savingAccount.accountNumber)))
    #set($attributeValues = {})
    $util.qr($attributeValues.put("username", $util.dynamodb.toString($savingAccount.username)))
    $util.qr($attributeValues.put("balance", $util.dynamodb.toNumber($savingAccount.balance)))
    #set($index = $index + 1)
    #set($savingAccountTransactPutItem = {"table": "savingAccounts",
        "operation": "PutItem",
        "key": $keyMap,
        "attributeValues": $attributeValues})
    $util.qr($savingAccountTransactPutItems.add($savingAccountTransactPutItem))
#end

#set($checkingAccountTransactPutItems = [])
#set($index = 0)
#foreach($checkingAccount in ${ctx.args.checkingAccounts})
    #set($keyMap = {})
    $util.qr($keyMap.put("accountNumber", $util.dynamodb.toString($checkingAccount.accountNumber)))
    #set($attributeValues = {})
    $util.qr($attributeValues.put("username", $util.dynamodb.toString($checkingAccount.username)))
    $util.qr($attributeValues.put("balance", $util.dynamodb.toNumber($checkingAccount.balance)))
    #set($index = $index + 1)
    #set($checkingAccountTransactPutItem = {"table": "checkingAccounts",
        "operation": "PutItem",
        "key": $keyMap,
        "attributeValues": $attributeValues})
    $util.qr($checkingAccountTransactPutItems.add($checkingAccountTransactPutItem))
#end

#set($transactItems = [])
$util.qr($transactItems.addAll($savingAccountTransactPutItems))
$util.qr($transactItems.addAll($checkingAccountTransactPutItems))

{
    "version" : "2018-05-29",
    "operation" : "TransactWriteItems",
    "transactItems" : $util.toJson($transactItems)
}

And the following response mapping template:

Response Mapping Template


#if ($ctx.error)
    $util.appendError($ctx.error.message, $ctx.error.type, null, $ctx.result.cancellationReasons)
#end

#set($savingAccounts = [])
#foreach($index in [0..2])
    $util.qr($savingAccounts.add(${ctx.result.keys[$index]}))
#end

#set($checkingAccounts = [])
#foreach($index in [3..5])
    $util.qr($checkingAccounts.add(${ctx.result.keys[$index]}))
#end

#set($transactionResult = {})
$util.qr($transactionResult.put('savingAccounts', $savingAccounts))
$util.qr($transactionResult.put('checkingAccounts', $checkingAccounts))

$util.toJson($transactionResult)

Save the resolver and navigate to the Queries section of the AWS AppSync console to populate the accounts.

Execute the following mutation:


mutation populateAccounts {
  populateAccounts (
    savingAccounts: [
      {accountNumber: "1", username: "Tom", balance: 100},
      {accountNumber: "2", username: "Amy", balance: 90},
      {accountNumber: "3", username: "Lily", balance: 80},
    ]
    checkingAccounts: [
      {accountNumber: "1", username: "Tom", balance: 70},
      {accountNumber: "2", username: "Amy", balance: 60},
      {accountNumber: "3", username: "Lily", balance: 50},
    ]) {
    savingAccounts {
      accountNumber
    }
    checkingAccounts {
      accountNumber
    }
  }
}

We populated 3 saving accounts and 3 checking accounts in one mutation.

Use the DynamoDB console to validate that data shows up in both the savingAccounts and checkingAccounts tables.

TransactWriteItems - Transfer Money

Attach a resolver to the transferMoney mutation with the following Request Mapping Template. Note the values of amounts, savingAccountNumbers, and checkingAccountNumbers are the same.


#set($amounts = [])
#foreach($transaction in ${ctx.args.transactions})
    #set($attributeValueMap = {})
    $util.qr($attributeValueMap.put(":amount", $util.dynamodb.toNumber($transaction.amount)))
    $util.qr($amounts.add($attributeValueMap))
#end

#set($savingAccountTransactUpdateItems = [])
#set($index = 0)
#foreach($transaction in ${ctx.args.transactions})
    #set($keyMap = {})
    $util.qr($keyMap.put("accountNumber", $util.dynamodb.toString($transaction.savingAccountNumber)))
    #set($update = {})
    $util.qr($update.put("expression", "SET balance = balance - :amount"))
    $util.qr($update.put("expressionValues", $amounts[$index]))
    #set($index = $index + 1)
    #set($savingAccountTransactUpdateItem = {"table": "savingAccounts",
        "operation": "UpdateItem",
        "key": $keyMap,
        "update": $update})
    $util.qr($savingAccountTransactUpdateItems.add($savingAccountTransactUpdateItem))
#end

#set($checkingAccountTransactUpdateItems = [])
#set($index = 0)
#foreach($transaction in ${ctx.args.transactions})
    #set($keyMap = {})
    $util.qr($keyMap.put("accountNumber", $util.dynamodb.toString($transaction.checkingAccountNumber)))
    #set($update = {})
    $util.qr($update.put("expression", "SET balance = balance + :amount"))
    $util.qr($update.put("expressionValues", $amounts[$index]))
    #set($index = $index + 1)
    #set($checkingAccountTransactUpdateItem = {"table": "checkingAccounts",
        "operation": "UpdateItem",
        "key": $keyMap,
        "update": $update})
    $util.qr($checkingAccountTransactUpdateItems.add($checkingAccountTransactUpdateItem))
#end

#set($transactionHistoryTransactPutItems = [])
#foreach($transaction in ${ctx.args.transactions})
    #set($keyMap = {})
    $util.qr($keyMap.put("transactionId", $util.dynamodb.toString(${utils.autoId()})))
    #set($attributeValues = {})
    $util.qr($attributeValues.put("from", $util.dynamodb.toString($transaction.savingAccountNumber)))
    $util.qr($attributeValues.put("to", $util.dynamodb.toString($transaction.checkingAccountNumber)))
    $util.qr($attributeValues.put("amount", $util.dynamodb.toNumber($transaction.amount)))
    #set($transactionHistoryTransactPutItem = {"table": "transactionHistory",
        "operation": "PutItem",
        "key": $keyMap,
        "attributeValues": $attributeValues})
    $util.qr($transactionHistoryTransactPutItems.add($transactionHistoryTransactPutItem))
#end

#set($transactItems = [])
$util.qr($transactItems.addAll($savingAccountTransactUpdateItems))
$util.qr($transactItems.addAll($checkingAccountTransactUpdateItems))
$util.qr($transactItems.addAll($transactionHistoryTransactPutItems))

{
    "version" : "2018-05-29",
    "operation" : "TransactWriteItems",
    "transactItems" : $util.toJson($transactItems)
}

We will have 3 banking transactions in a single TransactWriteItems operation. Use the following Response Mapping Template:


#if ($ctx.error)
    $util.appendError($ctx.error.message, $ctx.error.type, null, $ctx.result.cancellationReasons)
#end

#set($savingAccounts = [])
#foreach($index in [0..2])
    $util.qr($savingAccounts.add(${ctx.result.keys[$index]}))
#end

#set($checkingAccounts = [])
#foreach($index in [3..5])
    $util.qr($checkingAccounts.add(${ctx.result.keys[$index]}))
#end

#set($transactionHistory = [])
#foreach($index in [6..8])
    $util.qr($transactionHistory.add(${ctx.result.keys[$index]}))
#end

#set($transactionResult = {})
$util.qr($transactionResult.put('savingAccounts', $savingAccounts))
$util.qr($transactionResult.put('checkingAccounts', $checkingAccounts))
$util.qr($transactionResult.put('transactionHistory', $transactionHistory))

$util.toJson($transactionResult)

Now navigate to the Queries section of the AWS AppSync console and execute the transferMoney mutation as follows:


mutation write {
  transferMoney(
    transactions: [
      {savingAccountNumber: "1", checkingAccountNumber: "1", amount: 7.5},
      {savingAccountNumber: "2", checkingAccountNumber: "2", amount: 6.0},
      {savingAccountNumber: "3", checkingAccountNumber: "3", amount: 3.3}
    ]) {
    savingAccounts {
      accountNumber
    }
    checkingAccounts {
      accountNumber
    }
    transactionHistory {
      transactionId
    }
  }
}

We sent 2 banking transactions in one mutation. Use the DynamoDB console to validate that data shows up in the savingAccounts, checkingAccounts, and transactionHistory tables.

TransactGetItems - Retrieve Accounts

In order to retrieve the details from saving accounts and checking accounts in a single transactional request we’ll attach a resolver to the Query.getAccounts GraphQL operation on our schema. Select Attach, go to VTL Unit Resolvers, then on the next screen, pick the same TransactTutorial data source created at the beginning of the tutorial. Configure the templates as follows:

Request Mapping Template


#set($savingAccountsTransactGets = [])
#foreach($savingAccountNumber in ${ctx.args.savingAccountNumbers})
    #set($savingAccountKey = {})
    $util.qr($savingAccountKey.put("accountNumber", $util.dynamodb.toString($savingAccountNumber)))
    #set($savingAccountTransactGet = {"table": "savingAccounts", "key": $savingAccountKey})
    $util.qr($savingAccountsTransactGets.add($savingAccountTransactGet))
#end

#set($checkingAccountsTransactGets = [])
#foreach($checkingAccountNumber in ${ctx.args.checkingAccountNumbers})
    #set($checkingAccountKey = {})
    $util.qr($checkingAccountKey.put("accountNumber", $util.dynamodb.toString($checkingAccountNumber)))
    #set($checkingAccountTransactGet = {"table": "checkingAccounts", "key": $checkingAccountKey})
    $util.qr($checkingAccountsTransactGets.add($checkingAccountTransactGet))
#end

#set($transactItems = [])
$util.qr($transactItems.addAll($savingAccountsTransactGets))
$util.qr($transactItems.addAll($checkingAccountsTransactGets))

{
    "version" : "2018-05-29",
    "operation" : "TransactGetItems",
    "transactItems" : $util.toJson($transactItems)
}

Response Mapping Template


#if ($ctx.error)
    $util.appendError($ctx.error.message, $ctx.error.type, null, $ctx.result.cancellationReasons)
#end

#set($savingAccounts = [])
#foreach($index in [0..2])
    $util.qr($savingAccounts.add(${ctx.result.items[$index]}))
#end

#set($checkingAccounts = [])
#foreach($index in [3..4])
    $util.qr($checkingAccounts.add($ctx.result.items[$index]))
#end

#set($transactionResult = {})
$util.qr($transactionResult.put('savingAccounts', $savingAccounts))
$util.qr($transactionResult.put('checkingAccounts', $checkingAccounts))

$util.toJson($transactionResult)

Save the resolver and navigate to the Queries sections of the AWS AppSync console. In order to retrieve the saving accounts and checing accounts, execute the following query:


query getAccounts {
  getAccounts(
    savingAccountNumbers: ["1", "2", "3"],
    checkingAccountNumbers: ["1", "2"]
  ) {
    savingAccounts {
      accountNumber
      username
      balance
    }
    checkingAccounts {
      accountNumber
      username
      balance
    }
  }
}

We have successfully demonstrated the use of DynamoDB transactions using AWS AppSync.

Warning Javascript is disabled or is unavailable in your browser.

To use the Amazon Web Services Documentation, Javascript must be enabled. Please refer to your browser's Help pages for instructions.

Document Conventions

Using DynamoDB batch operations

Using HTTP resolvers