AWS SDK DynamoDB Examples: CRUD, Query, Batch and Conditional Writes (Python, PHP, Node.js)

By Ravindra Bagale, trainer · LinkedIn · Last updated: 24 October 2026

1. The big picture: a ledger with tabs

Why do we need this?

A shopkeeper keeps a ledger with one tab per customer (partition key) and lines sorted by date inside each tab (sort key). Finding "all of Tom's orders in October" is instant because you open one tab and read a range of dates.

What is it?

boto3 has a friendly Table resource that uses plain Python types. PHP has a Marshaler to convert arrays. Node has @aws-sdk/lib-dynamodb (DynamoDBDocumentClient) so you write plain JSON instead of {"S": "..."} types.

How to do it (step by step)

2. Create the table

Why do we need this?

Every order needs a home, and on-demand billing means no capacity planning for a small shop.

What is it?

Define the key attributes only (customer as HASH, orderDate as RANGE). Other attributes are schemaless.

How to do it (step by step)

1. Python: create and wait.

import boto3

ddb = boto3.resource("dynamodb", region_name="ap-south-1")
t = ddb.create_table(
    TableName="ShopOrders",
    KeySchema=[{"AttributeName": "customer", "KeyType": "HASH"}, {"AttributeName": "orderDate", "KeyType": "RANGE"}],
    AttributeDefinitions=[{"AttributeName": "customer", "AttributeType": "S"}, {"AttributeName": "orderDate", "AttributeType": "S"}],
    BillingMode="PAY_PER_REQUEST",
)
t.wait_until_exists()
print(t.table_status)

Expected output:

ACTIVE

2. Verify.

aws dynamodb describe-table --table-name ShopOrders --query "Table.[TableStatus,BillingModeSummary.BillingMode]" --output text

Expected output:

ACTIVE	PAY_PER_REQUEST

3. Put, get, update, delete

Why do we need this?

Fatima takes an order for 2 kg of biryani masala at ₹1,200; Tom sells a £3.40 flat white. Then quantities change and some orders are cancelled.

What is it?

PutItem writes a whole item; GetItem reads by full key; UpdateItem changes some attributes with an UpdateExpression; DeleteItem removes it. Use Decimal in Python for numbers.

How to do it (step by step)

1. Python: CRUD with the Table resource.

import boto3
from decimal import Decimal

t = boto3.resource("dynamodb", region_name="ap-south-1").Table("ShopOrders")
t.put_item(Item={"customer": "fatima-hyd", "orderDate": "2026-10-24#001", "item": "biryani masala", "kg": 2, "amount": Decimal("1200"), "currency": "INR"})
print(t.get_item(Key={"customer": "fatima-hyd", "orderDate": "2026-10-24#001"})["Item"]["item"])
r = t.update_item(Key={"customer": "fatima-hyd", "orderDate": "2026-10-24#001"},
    UpdateExpression="SET kg = :k, amount = :a", ExpressionAttributeValues={":k": 3, ":a": Decimal("1800")}, ReturnValues="UPDATED_NEW")
print(r["Attributes"])

Expected output:

biryani masala
{'kg': Decimal('3'), 'amount': Decimal('1800')}

2. PHP: Marshaler converts arrays to DynamoDB types.

<?php
require "vendor/autoload.php";
use Aws\DynamoDb\DynamoDbClient;
use Aws\DynamoDb\Marshaler;

$ddb = new DynamoDbClient(["region" => "ap-south-1"]);
$m = new Marshaler();
$key = $m->marshalItem(["customer" => "tom-bristol", "orderDate" => "2026-10-24#014"]);
$ddb->putItem(["TableName" => "ShopOrders", "Item" => $m->marshalItem(["customer" => "tom-bristol", "orderDate" => "2026-10-24#014", "item" => "flat white", "amount" => 3.40, "currency" => "GBP"])]);
$it = $m->unmarshalItem($ddb->getItem(["TableName" => "ShopOrders", "Key" => $key])["Item"]);
echo $it["item"], " £", $it["amount"], PHP_EOL;
$ddb->deleteItem(["TableName" => "ShopOrders", "Key" => $key]);
echo "deleted", PHP_EOL;

Expected output:

flat white £3.4
deleted

3. Node.js: DynamoDBDocumentClient with plain JSON.

// npm install @aws-sdk/client-dynamodb @aws-sdk/lib-dynamodb
import { DynamoDBClient } from "@aws-sdk/client-dynamodb";
import { DynamoDBDocumentClient, PutCommand, GetCommand, UpdateCommand } from "@aws-sdk/lib-dynamodb";

const doc = DynamoDBDocumentClient.from(new DynamoDBClient({ region: "ap-south-1" }));
const Key = { customer: "lucas-paris", orderDate: "2026-10-24#007" };
await doc.send(new PutCommand({ TableName: "ShopOrders", Item: { ...Key, item: "croissant box", amount: 18.5, currency: "EUR" } }));
const { Item } = await doc.send(new GetCommand({ TableName: "ShopOrders", Key }));
console.log(Item.item, Item.amount);
const u = await doc.send(new UpdateCommand({ TableName: "ShopOrders", Key, UpdateExpression: "SET #s = :s", ExpressionAttributeNames: { "#s": "status" }, ExpressionAttributeValues: { ":s": "PAID" }, ReturnValues: "ALL_NEW" }));
console.log(u.Attributes.status);

Expected output:

croissant box 18.5
PAID

4. Verify: read the item with the CLI.

aws dynamodb get-item --table-name ShopOrders --key '{"customer":{"S":"fatima-hyd"},"orderDate":{"S":"2026-10-24#001"}}' --query Item.kg.N --output text

Expected output:

3

4. Query and batch write

Why do we need this?

At month end Fatima needs all her October orders, and loading 500 old orders one by one would be slow.

What is it?

Query needs the partition key and can filter the sort key with begins_with or BETWEEN. BatchWriteItem takes up to 25 items; boto3 batch_writer and Node retries of UnprocessedItems handle the rest.

How to do it (step by step)

1. Python: batch_writer then query.

import boto3
from boto3.dynamodb.conditions import Key
from decimal import Decimal

t = boto3.resource("dynamodb", region_name="ap-south-1").Table("ShopOrders")
with t.batch_writer() as bw:
    for n in range(1, 101):
        bw.put_item(Item={"customer": "fatima-hyd", "orderDate": f"2026-10-{(n % 28) + 1:02d}#{n:03d}", "amount": Decimal(n * 50)})
r = t.query(KeyConditionExpression=Key("customer").eq("fatima-hyd") & Key("orderDate").begins_with("2026-10"))
print("October orders:", r["Count"])

Expected output:

October orders: 101

2. PHP: query with ExpressionAttributeValues.

<?php
require "vendor/autoload.php";
$ddb = new Aws\DynamoDb\DynamoDbClient(["region" => "ap-south-1"]);
$r = $ddb->query([
    "TableName" => "ShopOrders",
    "KeyConditionExpression" => "customer = :c AND begins_with(orderDate, :m)",
    "ExpressionAttributeValues" => [":c" => ["S" => "fatima-hyd"], ":m" => ["S" => "2026-10"]],
    "Select" => "COUNT",
]);
echo "October orders: ", $r["Count"], PHP_EOL;

Expected output:

October orders: 101

3. Node.js: BatchWriteCommand in chunks of 25.

import { DynamoDBClient } from "@aws-sdk/client-dynamodb";
import { DynamoDBDocumentClient, BatchWriteCommand, QueryCommand } from "@aws-sdk/lib-dynamodb";

const doc = DynamoDBDocumentClient.from(new DynamoDBClient({ region: "ap-south-1" }));
const items = Array.from({ length: 60 }, (_, i) => ({ customer: "tom-bristol", orderDate: `2026-10-${String(i % 28 + 1).padStart(2, "0")}#${i}`, amount: 3.4 }));
for (let i = 0; i < items.length; i += 25) {
  let req = { ShopOrders: items.slice(i, i + 25).map(Item => ({ PutRequest: { Item } })) };
  while (req && Object.keys(req).length) {
    const r = await doc.send(new BatchWriteCommand({ RequestItems: req }));
    req = r.UnprocessedItems;
  }
}
const q = await doc.send(new QueryCommand({ TableName: "ShopOrders", KeyConditionExpression: "customer = :c", ExpressionAttributeValues: { ":c": "tom-bristol" }, Select: "COUNT" }));
console.log("Tom orders:", q.Count);

Expected output:

Tom orders: 60

4. Verify: count with the CLI.

aws dynamodb query --table-name ShopOrders --key-condition-expression "customer = :c" --expression-attribute-values '{":c":{"S":"tom-bristol"}}' --select COUNT --query Count

Expected output:

60

5. Conditional writes and counters

Why do we need this?

Two people book the last table at Sofia's restaurant in Rome at the same second. Only one must win. And a page-view counter must never lose updates.

What is it?

ConditionExpression makes the write succeed only if a rule holds, such as attribute_not_exists(customer). If it fails you get ConditionalCheckFailedException. ADD increments a number atomically.

How to do it (step by step)

1. Python: book only if not already booked.

import boto3
from botocore.exceptions import ClientError

t = boto3.resource("dynamodb", region_name="ap-south-1").Table("ShopOrders")
for who in ["giulia", "marco"]:
    try:
        t.put_item(Item={"customer": "sofia-rome-table7", "orderDate": "2026-10-31T20:00", "bookedBy": who},
                   ConditionExpression="attribute_not_exists(customer)")
        print(who, "booked")
    except ClientError as e:
        if e.response["Error"]["Code"] == "ConditionalCheckFailedException":
            print(who, "too late")

Expected output:

giulia booked
marco too late

2. PHP: atomic counter with ADD.

<?php
require "vendor/autoload.php";
$ddb = new Aws\DynamoDb\DynamoDbClient(["region" => "ap-south-1"]);
for ($i = 0; $i < 3; $i++) {
    $r = $ddb->updateItem([
        "TableName" => "ShopOrders",
        "Key" => ["customer" => ["S" => "stats"], "orderDate" => ["S" => "views"]],
        "UpdateExpression" => "ADD hits :one",
        "ExpressionAttributeValues" => [":one" => ["N" => "1"]],
        "ReturnValues" => "UPDATED_NEW",
    ]);
}
echo "hits: ", $r["Attributes"]["hits"]["N"], PHP_EOL;

Expected output:

hits: 3

3. Node.js: optimistic locking with a version number.

import { DynamoDBClient } from "@aws-sdk/client-dynamodb";
import { DynamoDBDocumentClient, UpdateCommand } from "@aws-sdk/lib-dynamodb";

const doc = DynamoDBDocumentClient.from(new DynamoDBClient({ region: "ap-south-1" }));
try {
  await doc.send(new UpdateCommand({
    TableName: "ShopOrders", Key: { customer: "lucas-paris", orderDate: "2026-10-24#007" },
    UpdateExpression: "SET amount = :a, ver = :new", ConditionExpression: "attribute_not_exists(ver) OR ver = :old",
    ExpressionAttributeValues: { ":a": 20, ":new": 2, ":old": 1 },
  }));
  console.log("updated");
} catch (e) {
  console.log(e.name);
}

Expected output:

updated

4. Verify: only Giulia holds table 7.

aws dynamodb get-item --table-name ShopOrders --key '{"customer":{"S":"sofia-rome-table7"},"orderDate":{"S":"2026-10-31T20:00"}}' --query Item.bookedBy.S --output text

Expected output:

giulia

6. The same call in Java, Go, .NET and the CLI

Why do we need this?

Teams often mix languages. Once you know the call in Python, PHP and Node.js, the other SDKs are the same API with different syntax.

What is it?

All of them call the PutItem API. Use the CLI line to test it quickly from a terminal.

How to do it (step by step)

7. Clean up so nothing keeps costing

Why do we need this?

Lab resources left running cost money and can be a security risk.

What is it?

Delete in reverse order: children first, then parents.

How to do it (step by step)

1. Delete the table.

aws dynamodb delete-table --table-name ShopOrders --query TableDescription.TableStatus --output text
aws dynamodb wait table-not-exists --table-name ShopOrders

Expected output:

DELETING

8. Interview questions and answers

Revision, comparison and interview cards

Revise this lesson in a few minutes. Save the cards on your phone or download them as PNG files. The QR code on each card opens this lesson.

More cards: Ravi Sir's Cards.