<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Aniruddh Fataniya on Medium]]></title>
        <description><![CDATA[Stories by Aniruddh Fataniya on Medium]]></description>
        <link>https://medium.com/@aniruddhfataniya?source=rss-445e0e33ac7------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*NGH0aJioYveMiE2xLCAJzQ.jpeg</url>
            <title>Stories by Aniruddh Fataniya on Medium</title>
            <link>https://medium.com/@aniruddhfataniya?source=rss-445e0e33ac7------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 10:21:19 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@aniruddhfataniya/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Building an End-to-End Managed SFTP Architecture on AWS]]></title>
            <link>https://aws.plainenglish.io/building-an-end-to-end-managed-sftp-architecture-on-aws-98d477299635?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/98d477299635</guid>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[sftp]]></category>
            <category><![CDATA[transfer-family]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Tue, 17 Nov 2026 10:39:42 GMT</pubDate>
            <atom:updated>2026-11-17T10:39:42.987Z</atom:updated>
            <content:encoded><![CDATA[<p>SFTP (Secure File Transfer Protocol) is a system that lets you securely send, receive, and manage files over a network using the encrypted SSH (Secure Shell) protocol.</p><p>AWS Transfer Family is a managed SFTP service that uses either S3 or EFS for storage.</p><p><strong>When to Choose Amazon S3</strong></p><ul><li><strong>Cost-Sensitive Archiving and Ingestion:</strong> You want the lowest storage cost for large volumes of incoming or outgoing files.</li><li><strong>Event-Driven Data Pipelines:</strong> You need uploads to automatically trigger downstream processing via S3 Event Notifications, AWS Lambda, or AWS Glue.</li><li><strong>Data Lakes and Analytics:</strong> You plan to query or analyze the transferred files directly using Amazon Athena, Redshift, or QuickSight.</li><li><strong>Simple Write-Once Workflows:</strong> Your trading partners only upload complete files or download whole files without modifying them in place.</li></ul><p><strong>When to Choose Amazon EFS</strong></p><ul><li><strong>POSIX Permissions and Ownership:</strong> Your legacy applications or partners require strict Linux user/group IDs (UID/GID) and file-level permissions (chmod/chown).</li><li><strong>In-Place File Modifications:</strong> Your workflows append data to existing files, update specific byte ranges, or rely on file locking.</li><li><strong>Direct Compute Mounting:</strong> You have existing EC2 instances, containers, or batch jobs that expect to read and write files over a standard local directory mount.</li><li><strong>Hierarchical Directory Operations:</strong> Your partners frequently rename large folders or perform heavy directory-tree operations, which are instant in EFS but require copy-and-delete operations in S3.</li></ul><p>This article provides an end-to-end SFTP solution using AWS Transfer Family with EFS. The following is the architecture, and we will discuss the components one by one:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AV2_4ugxuUgCbK-af_dDBQ.png" /></figure><h3>Create an EFS file system</h3><p>The first step is to provision an EFS file system, which will be used as the storage location for the AWS Transfer Family.</p><ul><li>Set up an EFS file system in the private subnets with the required 2049 (NFS) traffic allowed from the Lambda SG in the attached security group</li><li>Create an Access Point for this file system, which will be used by the Lambda functions to create/manage users and storage</li></ul><h3>Create an AWS Transfer Family Server</h3><p>Create a server in the AWS Transfer Family with the following attributes:</p><ul><li>Protocol: SFTP</li><li>Identity provider type: Custom — AWS Lambda (to be created in the subsequent step)</li><li>Authentication methods: Password only</li><li>Endpoint type: Public</li><li>Domain: EFS</li></ul><p>You can also mention the Custom hostname if you want to use a custom domain name rather than using the Endpoint directly.</p><h3>DynamoDB tables</h3><p>There are two DynamoDB tables are required here:</p><ul><li>Storing user data</li><li>Keeping a record of the GID and UID for generating unique values for each user</li></ul><p>Create two tables for the same.</p><h3>Lambda Function</h3><p>Here we will create Lambda functions for different functionalities:</p><h4>Create user</h4><p>This Lambda function creates users by setting up a directory for the user in EFS and adding records to the DynamoDB tables.</p><p>This Lambda function requires connecting to EFS, so it needs to be set up in the VPC. Under Configurations, set up the VPC and select the private subnets that were selected for EFS.</p><p>The Security group of the Lambda should have a rule to allow outbound traffic on port 2049 (NFS) to the EFS SG.</p><p>The file system also needs to be mounted on the Lambda. Select the respective EFS file system and access point.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/291/1*kYxZzkywxxT0hr09t6d4Hg.png" /></figure><p>Make sure the Lambda function has the required permissions for EFS and DynamoDB. Also make sure the Lambda has enough timeout. Please find the code for the Lambda function as follows:</p><pre><br>import os<br>import json<br>import re<br>import base64<br>import secrets<br>import string<br>import hashlib<br>import logging<br>from datetime import datetime, timezone<br><br>import boto3<br>from botocore.exceptions import ClientError<br>from botocore.config import Config<br><br>logger = logging.getLogger()<br>logger.setLevel(logging.INFO)<br><br># --------------------------------------------------<br># Configuration<br># --------------------------------------------------<br><br>REGION = &quot;region&quot;<br><br>SERVER_ID = &quot;server_id&quot;<br><br>EFS_ID = &quot;efs_id&quot;<br>EFS_MOUNT_PATH = &quot;/mnt/efs&quot;<br><br>DYNAMODB_TABLE = &quot;user_table&quot;<br>COUNTER_TABLE_NAME = os.environ.get(<br>    &quot;COUNTER_TABLE&quot;,<br>    &quot;scounter_table&quot;<br>)<br><br>PBKDF2_ITERATIONS = 600000<br>PASSWORD_LENGTH = 20<br><br># --------------------------------------------------<br># AWS Clients<br># --------------------------------------------------<br><br>dynamodb = boto3.resource(<br>    &quot;dynamodb&quot;,<br>    region_name=REGION,<br>    config=Config(<br>        connect_timeout=5,<br>        read_timeout=10,<br>        retries={&quot;max_attempts&quot;: 3}<br>    )<br>)<br><br>table = dynamodb.Table(DYNAMODB_TABLE)<br>counter_table = dynamodb.Table(COUNTER_TABLE_NAME)<br><br># --------------------------------------------------<br># API Response<br># --------------------------------------------------<br><br>def response(status_code, body):<br>    return {<br>        &quot;statusCode&quot;: status_code,<br>        &quot;headers&quot;: {<br>            &quot;Content-Type&quot;: &quot;application/json&quot;,<br>            &quot;Cache-Control&quot;: &quot;no-store&quot;<br>        },<br>        &quot;body&quot;: json.dumps(body)<br>    }<br><br><br># --------------------------------------------------<br># Parse API Gateway Request<br># --------------------------------------------------<br><br>def get_request_body(event):<br><br>    if &quot;body&quot; in event:<br><br>        body = event[&quot;body&quot;]<br><br>        if body is None:<br>            raise ValueError(&quot;Request body is required&quot;)<br><br>        if event.get(&quot;isBase64Encoded&quot;, False):<br>            body = base64.b64decode(body).decode(&quot;utf-8&quot;)<br><br>        if isinstance(body, str):<br>            body = json.loads(body)<br><br>    else:<br>        # Direct Lambda invocation support<br>        body = event<br><br>    if not isinstance(body, dict):<br>        raise ValueError(&quot;Request body must be a JSON object&quot;)<br><br>    return body<br><br><br># --------------------------------------------------<br># Validate Username<br># --------------------------------------------------<br><br>def validate_user(username):<br><br>    if not isinstance(username, str):<br>        raise ValueError(&quot;Username must be a string&quot;)<br><br>    username = username.strip()<br><br>    if not re.fullmatch(r&quot;[A-Za-z0-9_-]{3,32}&quot;, username):<br>        raise ValueError(<br>            &quot;Username must contain 3-32 characters &quot;<br>            &quot;using letters, numbers, underscore or hyphen&quot;<br>        )<br><br>    return username<br><br><br># --------------------------------------------------<br># Generate Unique UID and GID<br># --------------------------------------------------<br><br>def generate_uid_gid():<br><br>    result = counter_table.update_item(<br>        Key={<br>            &quot;server_id&quot;: SERVER_ID<br>        },<br>        UpdateExpression=(<br>            &quot;SET last_id = if_not_exists(last_id, :start) + :increment&quot;<br>        ),<br>        ConditionExpression=(<br>            &quot;attribute_not_exists(last_id) OR last_id &lt; :max_id&quot;<br>        ),<br>        ExpressionAttributeValues={<br>            &quot;:start&quot;: 10000,<br>            &quot;:increment&quot;: 1,<br>            &quot;:max_id&quot;: 2147483647<br>        },<br>        ReturnValues=&quot;UPDATED_NEW&quot;<br>    )<br><br>    user_id = int(result[&quot;Attributes&quot;][&quot;last_id&quot;])<br><br>    logger.info(&quot;Generated UID/GID: %s&quot;, user_id)<br><br>    # Private group model: UID and GID are identical<br>    return user_id, user_id<br><br><br># --------------------------------------------------<br># Generate Secure Random Password<br># --------------------------------------------------<br><br>def generate_password(length=PASSWORD_LENGTH):<br><br>    if length &lt; 16:<br>        raise ValueError(&quot;Password length must be at least 16&quot;)<br><br>    uppercase = string.ascii_uppercase<br>    lowercase = string.ascii_lowercase<br>    digits = string.digits<br>    special = &quot;!@#$%^&amp;*()-_=+&quot;<br><br>    all_characters = uppercase + lowercase + digits + special<br><br>    password_characters = [<br>        secrets.choice(uppercase),<br>        secrets.choice(lowercase),<br>        secrets.choice(digits),<br>        secrets.choice(special)<br>    ]<br><br>    password_characters.extend(<br>        secrets.choice(all_characters)<br>        for _ in range(length - 4)<br>    )<br><br>    secrets.SystemRandom().shuffle(password_characters)<br><br>    return &quot;&quot;.join(password_characters)<br><br><br># --------------------------------------------------<br># Hash Password<br># --------------------------------------------------<br><br>def hash_password(password):<br><br>    salt = secrets.token_bytes(16)<br><br>    password_hash = hashlib.pbkdf2_hmac(<br>        &quot;sha256&quot;,<br>        password.encode(&quot;utf-8&quot;),<br>        salt,<br>        PBKDF2_ITERATIONS,<br>        dklen=32<br>    )<br><br>    return salt.hex(), password_hash.hex()<br><br><br># --------------------------------------------------<br># Create EFS User Directory<br># --------------------------------------------------<br><br>def create_efs_directory(username, uid, gid):<br><br>    if not os.path.ismount(EFS_MOUNT_PATH):<br>        raise RuntimeError(&quot;EFS is not mounted&quot;)<br><br>    user_directory = os.path.realpath(<br>        os.path.join(EFS_MOUNT_PATH, username)<br>    )<br><br>    efs_root = os.path.realpath(EFS_MOUNT_PATH)<br><br>    # Prevent directory traversal<br>    if os.path.dirname(user_directory) != efs_root:<br>        raise ValueError(&quot;Invalid EFS directory path&quot;)<br><br>    if os.path.exists(user_directory):<br><br>        if not os.path.isdir(user_directory):<br>            raise RuntimeError(<br>                &quot;User path exists but is not a directory&quot;<br>            )<br><br>        raise FileExistsError(&quot;EFS directory already exists&quot;)<br><br>    os.mkdir(user_directory, 0o750)<br><br>    try:<br>        os.chown(user_directory, uid, gid)<br>        os.chmod(user_directory, 0o750)<br><br>    except Exception:<br>        try:<br>            os.rmdir(user_directory)<br>        except OSError:<br>            logger.exception(<br>                &quot;Failed to clean up EFS directory&quot;<br>            )<br>        raise<br><br>    logger.info(&quot;Created EFS directory for user %s&quot;, username)<br><br>    return user_directory<br><br><br># --------------------------------------------------<br># Save User in DynamoDB<br># --------------------------------------------------<br><br>def save_user(username, uid, gid, password):<br><br>    salt, password_hash = hash_password(password)<br><br>    current_time = datetime.now(timezone.utc).isoformat()<br><br>    item = {<br>        &quot;server_id&quot;: SERVER_ID,<br>        &quot;username&quot;: username,<br>        &quot;password_salt&quot;: salt,<br>        &quot;password_hash&quot;: password_hash,<br>        &quot;uid&quot;: uid,<br>        &quot;gid&quot;: gid,<br>        &quot;status&quot;: &quot;ACTIVE&quot;,<br>        &quot;created_at&quot;: current_time,<br>        &quot;updated_at&quot;: current_time<br>    }<br><br>    table.put_item(<br>        Item=item,<br>        ConditionExpression=&quot;attribute_not_exists(username)&quot;<br>    )<br><br><br># --------------------------------------------------<br># Rollback DynamoDB Record<br># --------------------------------------------------<br><br>def delete_user_record(username):<br><br>    try:<br>        table.delete_item(<br>            Key={<br>                &quot;server_id&quot;: SERVER_ID,<br>                &quot;username&quot;: username<br>            }<br>        )<br><br>        logger.info(<br>            &quot;Rolled back DynamoDB record for %s&quot;,<br>            username<br>        )<br><br>    except Exception:<br>        logger.exception(<br>            &quot;Failed to roll back DynamoDB record for %s&quot;,<br>            username<br>        )<br><br><br># --------------------------------------------------<br># Rollback EFS Directory<br># --------------------------------------------------<br><br>def delete_efs_directory(user_directory, username):<br><br>    if not user_directory:<br>        return<br><br>    try:<br>        os.rmdir(user_directory)<br><br>        logger.info(<br>            &quot;Rolled back EFS directory for %s&quot;,<br>            username<br>        )<br><br>    except Exception:<br>        logger.exception(<br>            &quot;Failed to roll back EFS directory for %s&quot;,<br>            username<br>        )<br><br><br># --------------------------------------------------<br># Lambda Handler<br># --------------------------------------------------<br><br>def lambda_handler(event, context):<br><br>    username = None<br>    user_directory = None<br>    directory_created = False<br>    dynamodb_record_created = False<br><br>    try:<br><br>        # Parse request<br>        body = get_request_body(event)<br><br>        username = validate_user(body.get(&quot;username&quot;))<br><br>        # Check whether user already exists<br>        existing_user = table.get_item(<br>            Key={<br>                &quot;server_id&quot;: SERVER_ID,<br>                &quot;username&quot;: username<br>            },<br>            ConsistentRead=True<br>        )<br><br>        if &quot;Item&quot; in existing_user:<br><br>            return response(409, {<br>                &quot;message&quot;: &quot;SFTP user already exists&quot;,<br>                &quot;username&quot;: username<br>            })<br><br>        # Allocate unique UID/GID<br>        uid, gid = generate_uid_gid()<br><br>        # Create EFS directory<br>        user_directory = create_efs_directory(<br>            username,<br>            uid,<br>            gid<br>        )<br><br>        directory_created = True<br><br>        # Generate password<br>        generated_password = generate_password()<br><br>        # Save password hash and user details<br>        save_user(<br>            username,<br>            uid,<br>            gid,<br>            generated_password<br>        )<br><br>        dynamodb_record_created = True<br><br>        logger.info(<br>            &quot;SFTP user provisioned successfully: %s&quot;,<br>            username<br>        )<br><br>        # Return password only once<br>        return response(201, {<br>            &quot;message&quot;: &quot;SFTP user created successfully&quot;,<br>            &quot;username&quot;: username,<br>            &quot;password&quot;: generated_password,<br>            &quot;server_id&quot;: SERVER_ID,<br>            &quot;protocol&quot;: &quot;SFTP&quot;,<br>            &quot;port&quot;: 22,<br>            &quot;home_directory&quot;: &quot;/&quot;,<br>            &quot;home_directory_type&quot;: &quot;LOGICAL&quot;,<br>            &quot;uid&quot;: uid,<br>            &quot;gid&quot;: gid<br>        })<br><br>    except ValueError as e:<br><br>        return response(400, {<br>            &quot;message&quot;: str(e)<br>        })<br><br>    except FileExistsError:<br><br>        return response(409, {<br>            &quot;message&quot;: &quot;EFS directory already exists&quot;,<br>            &quot;username&quot;: username<br>        })<br><br>    except ClientError as e:<br><br>        error_code = e.response.get(<br>            &quot;Error&quot;, {}<br>        ).get(&quot;Code&quot;, &quot;&quot;)<br><br>        logger.error(<br>            &quot;AWS operation failed for user %s: %s&quot;,<br>            username,<br>            error_code<br>        )<br><br>        # Roll back resources created<br>        if dynamodb_record_created:<br>            delete_user_record(username)<br><br>        if directory_created:<br>            delete_efs_directory(user_directory, username)<br><br>        if error_code == &quot;ConditionalCheckFailedException&quot;:<br><br>            return response(409, {<br>                &quot;message&quot;: &quot;SFTP user already exists&quot;,<br>                &quot;username&quot;: username<br>            })<br><br>        return response(500, {<br>            &quot;message&quot;: &quot;SFTP user provisioning failed&quot;<br>        })<br><br>    except Exception:<br><br>        logger.exception(<br>            &quot;SFTP user provisioning failed for %s&quot;,<br>            username<br>        )<br><br>        if dynamodb_record_created:<br>            delete_user_record(username)<br><br>        if directory_created:<br>            delete_efs_directory(user_directory, username)<br><br>        return response(500, {<br>            &quot;message&quot;: &quot;SFTP user provisioning failed&quot;<br>        })</pre><h4>Authenticate user</h4><p>This Lambda function is set as the Custom Identity provider for user authentication. Make sure it has the required DynamoDB permissions attached to the IAM role. It also requires a Resource-based policy to be attached to the Lambda function, as it will be invoked by the AWS Transfer Family for authentication.</p><p>Make sure the Lambda function has enough timeout.</p><p>Find the Resource-based policy as follows:</p><pre>{<br>  &quot;Version&quot;: &quot;2012-10-17&quot;,<br>  &quot;Id&quot;: &quot;Invoke&quot;,<br>  &quot;Statement&quot;: [<br>    {<br>      &quot;Sid&quot;: &quot;AllowTransferFamilyInvoke&quot;,<br>      &quot;Effect&quot;: &quot;Allow&quot;,<br>      &quot;Principal&quot;: {<br>        &quot;Service&quot;: &quot;transfer.amazonaws.com&quot;<br>      },<br>      &quot;Action&quot;: &quot;lambda:InvokeFunction&quot;,<br>      &quot;Resource&quot;: &quot;lambda_arn&quot;,<br>      &quot;Condition&quot;: {<br>        &quot;ArnLike&quot;: {<br>          &quot;AWS:SourceArn&quot;: &quot;sftp_server_arn&quot;<br>        }<br>      }<br>    }<br>  ]<br>}</pre><p>Please find the code for the Lambda function:</p><pre><br>import os<br>import json<br>import hashlib<br>import hmac<br>import logging<br><br>import boto3<br>from botocore.config import Config<br><br>logger = logging.getLogger()<br>logger.setLevel(logging.INFO)<br><br>REGION = &quot;us-east-1&quot;<br>SERVER_ID = os.environ[&quot;SERVER_ID&quot;]<br>TABLE_NAME = os.environ[&quot;DYNAMODB_TABLE&quot;]<br>EFS_ID = os.environ[&quot;EFS_ID&quot;]<br>TRANSFER_USER_ROLE_ARN = os.environ[&quot;TRANSFER_USER_ROLE_ARN&quot;]<br><br>PBKDF2_ITERATIONS = 600000<br><br>dynamodb = boto3.resource(<br>    &quot;dynamodb&quot;,<br>    region_name=REGION,<br>    config=Config(<br>        connect_timeout=3,<br>        read_timeout=5,<br>        retries={&quot;max_attempts&quot;: 2}<br>    )<br>)<br><br>table = dynamodb.Table(TABLE_NAME)<br><br><br>def verify_password(password, salt_hex, stored_hash):<br>    salt = bytes.fromhex(salt_hex)<br><br>    calculated_hash = hashlib.pbkdf2_hmac(<br>        &quot;sha256&quot;,<br>        password.encode(&quot;utf-8&quot;),<br>        salt,<br>        PBKDF2_ITERATIONS,<br>        dklen=32<br>    ).hex()<br><br>    return hmac.compare_digest(calculated_hash, stored_hash)<br><br><br>def lambda_handler(event, context):<br><br>    # Never log the complete event because it contains the password.<br>    username = event.get(&quot;username&quot;, &quot;&quot;)<br>    password = event.get(&quot;password&quot;, &quot;&quot;)<br>    server_id = event.get(&quot;serverId&quot;, &quot;&quot;)<br>    protocol = event.get(&quot;protocol&quot;, &quot;&quot;)<br><br>    if server_id != SERVER_ID or protocol != &quot;SFTP&quot;:<br>        return {}<br><br>    if not username or not password:<br>        return {}<br><br>    try:<br>        response = table.get_item(<br>            Key={<br>                &quot;server_id&quot;: SERVER_ID,<br>                &quot;username&quot;: username<br>            },<br>            ConsistentRead=True<br>        )<br><br>        user = response.get(&quot;Item&quot;)<br><br>        if not user or user.get(&quot;status&quot;) != &quot;ACTIVE&quot;:<br>            logger.warning(&quot;Authentication rejected for user %s&quot;, username)<br>            return {}<br><br>        if not verify_password(<br>            password,<br>            user[&quot;password_salt&quot;],<br>            user[&quot;password_hash&quot;]<br>        ):<br>            logger.warning(&quot;Invalid password for user %s&quot;, username)<br>            return {}<br><br>        uid = int(user[&quot;uid&quot;])<br>        gid = int(user[&quot;gid&quot;])<br><br>        result = {<br>            &quot;Role&quot;: TRANSFER_USER_ROLE_ARN,<br>            &quot;HomeDirectoryType&quot;: &quot;LOGICAL&quot;,<br>            &quot;HomeDirectoryDetails&quot;: json.dumps([<br>                {<br>                    &quot;Entry&quot;: &quot;/&quot;,<br>                    &quot;Target&quot;: f&quot;/{EFS_ID}/{username}&quot;<br>                }<br>            ]),<br>            &quot;PosixProfile&quot;: {<br>                &quot;Uid&quot;: uid,<br>                &quot;Gid&quot;: gid<br>            }<br>        }<br><br>        logger.info(&quot;Authentication successful for user %s&quot;, username)<br><br>        return result<br><br>    except Exception:<br>        logger.exception(<br>            &quot;Authentication processing failed for user %s&quot;,<br>            username<br>        )<br>        return {}</pre><h4>Custom workflow Lambda for archiving files to S3</h4><p>This Lambda will help archive the files when they are uploaded to SFTP. This Lambda requires similar configurations as the Create user Lambda function, i.e., VPC and File system setup.</p><p>This Lambda function will create a prefix in the S3 bucket for the respective user, rename the file by appending a timestamp, and store it.</p><p>Make sure the Lambda function has enough timeout.</p><p>Please find the code:</p><pre><br>import os<br>import json<br>import logging<br>import time<br>from datetime import datetime, timezone<br><br>import boto3<br>from botocore.config import Config<br>from botocore.exceptions import ClientError<br>from boto3.s3.transfer import TransferConfig<br><br>logger = logging.getLogger()<br>logger.setLevel(logging.INFO)<br><br><br># ============================================================<br># Configuration<br># ============================================================<br><br>REGION = &quot;us-east-1&quot;<br><br>EFS_ID = &quot;efs_id&quot;<br>EFS_MOUNT_PATH = &quot;/mnt/efs&quot;<br><br>S3_BUCKET = &quot;s3_bucket&quot;<br><br><br># ============================================================<br># AWS Clients<br># ============================================================<br><br>transfer = boto3.client(<br>    &quot;transfer&quot;,<br>    region_name=REGION,<br>    endpoint_url=&quot;https://transfer-fips.us-east-1.amazonaws.com&quot;,<br>    config=Config(<br>        connect_timeout=5,<br>        read_timeout=15,<br>        retries={<br>            &quot;mode&quot;: &quot;standard&quot;,<br>            &quot;max_attempts&quot;: 3<br>        }<br>    )<br>)<br><br>s3 = boto3.client(<br>    &quot;s3&quot;,<br>    region_name=REGION,<br>    config=Config(<br>        connect_timeout=5,<br>        read_timeout=30,<br>        retries={<br>            &quot;mode&quot;: &quot;standard&quot;,<br>            &quot;max_attempts&quot;: 3<br>        }<br>    )<br>)<br><br># S3 upload configuration<br>upload_config = TransferConfig(<br>    multipart_threshold=16 * 1024 * 1024,<br>    multipart_chunksize=8 * 1024 * 1024,<br>    max_concurrency=4,<br>    use_threads=True<br>)<br><br><br># ============================================================<br># Transfer Family Workflow Callback<br># ============================================================<br><br>def report_workflow_status(event, status):<br><br>    execution = event[&quot;serviceMetadata&quot;][&quot;executionDetails&quot;]<br><br>    logger.info(<br>        &quot;Reporting workflow status: workflowId=%s, &quot;<br>        &quot;executionId=%s, status=%s&quot;,<br>        execution[&quot;workflowId&quot;],<br>        execution[&quot;executionId&quot;],<br>        status<br>    )<br><br>    transfer.send_workflow_step_state(<br>        WorkflowId=execution[&quot;workflowId&quot;],<br>        ExecutionId=execution[&quot;executionId&quot;],<br>        Token=event[&quot;token&quot;],<br>        Status=status<br>    )<br><br>    logger.info(<br>        &quot;Transfer Family accepted workflow status: %s&quot;,<br>        status<br>    )<br><br><br>def report_failure_safely(event):<br><br>    try:<br>        report_workflow_status(event, &quot;FAILURE&quot;)<br><br>    except Exception:<br>        logger.exception(<br>            &quot;Unable to report FAILURE to Transfer Family. &quot;<br>            &quot;The workflow execution may already have expired.&quot;<br>        )<br><br><br># ============================================================<br># Source File Validation<br># ============================================================<br><br>def resolve_source_file(file_location):<br><br>    if file_location.get(&quot;domain&quot;) != &quot;EFS&quot;:<br>        raise ValueError(&quot;Expected an EFS source file&quot;)<br><br>    if file_location.get(&quot;fileSystemId&quot;) != EFS_ID:<br>        raise ValueError(&quot;Unexpected EFS filesystem ID&quot;)<br><br>    if not os.path.ismount(EFS_MOUNT_PATH):<br>        raise RuntimeError(&quot;EFS is not mounted&quot;)<br><br>    efs_root = os.path.realpath(EFS_MOUNT_PATH)<br><br>    source_path = file_location[&quot;path&quot;]<br><br>    source_file = os.path.realpath(<br>        os.path.join(<br>            efs_root,<br>            source_path.lstrip(&quot;/&quot;)<br>        )<br>    )<br><br>    if os.path.commonpath(<br>        [efs_root, source_file]<br>    ) != efs_root:<br>        raise ValueError(&quot;Invalid source file path&quot;)<br><br>    if not os.path.isfile(source_file):<br>        raise FileNotFoundError(<br>            f&quot;Uploaded file not found: {source_file}&quot;<br>        )<br><br>    return source_file<br><br><br># ============================================================<br># S3 Destination Key<br># ============================================================<br><br>def generate_s3_key(username, source_file):<br><br>    original_filename = os.path.basename(source_file)<br><br>    filename, extension = os.path.splitext(original_filename)<br><br>    # Use the EFS file modification time.<br>    # This remains stable across repeated Lambda invocations<br>    # for the same unchanged uploaded file.<br><br>    file_stat = os.stat(source_file)<br><br>    modified_time = datetime.fromtimestamp(<br>        file_stat.st_mtime,<br>        tz=timezone.utc<br>    )<br><br>    timestamp = datetime.now(timezone.utc).strftime(&quot;%Y%m%d_%H%M%S_%f&quot;)<br><br>    current_month = modified_time.strftime(&quot;%Y-%m&quot;)<br><br>    new_filename = (<br>        f&quot;{filename}_{timestamp}{extension}&quot;<br>    )<br><br>    s3_key = (<br>        f&quot;{username}/{current_month}/{new_filename}&quot;<br>    )<br><br>    return s3_key, file_stat.st_size, modified_time<br><br><br># ============================================================<br># S3 Upload<br># ============================================================<br><br>def archive_to_s3(source_file, s3_key):<br><br>    file_size = os.path.getsize(source_file)<br><br>    logger.info(<br>        &quot;Starting S3 upload: source=%s, size=%s bytes, &quot;<br>        &quot;destination=s3://%s/%s&quot;,<br>        source_file,<br>        file_size,<br>        S3_BUCKET,<br>        s3_key<br>    )<br><br>    upload_start = time.monotonic()<br><br>    relative_path = os.path.relpath(source_file, EFS_MOUNT_PATH)<br><br>    s3.upload_file(<br>        source_file,<br>        S3_BUCKET,<br>        s3_key,<br>        ExtraArgs={<br>            &quot;Metadata&quot;: {<br>                &quot;efs-relative-path&quot;: relative_path<br>            }<br>        },<br>        Config=upload_config<br>    )<br><br>    upload_duration = time.monotonic() - upload_start<br><br>    logger.info(<br>        &quot;S3 upload completed successfully in %.2f seconds&quot;,<br>        upload_duration<br>    )<br><br>    # Verify that the object exists and has the expected size.<br><br>    response = s3.head_object(<br>        Bucket=S3_BUCKET,<br>        Key=s3_key<br>    )<br><br>    uploaded_size = response[&quot;ContentLength&quot;]<br><br>    if uploaded_size != file_size:<br>        raise RuntimeError(<br>            f&quot;S3 object size mismatch. &quot;<br>            f&quot;Source={file_size}, S3={uploaded_size}&quot;<br>        )<br><br>    logger.info(<br>        &quot;S3 archive verified: s3://%s/%s, size=%s bytes&quot;,<br>        S3_BUCKET,<br>        s3_key,<br>        uploaded_size<br>    )<br><br>    return uploaded_size<br><br><br># ============================================================<br># Lambda Handler<br># ============================================================<br><br>def lambda_handler(event, context):<br><br>    logger.info(<br>        &quot;Received workflow event: %s&quot;,<br>        json.dumps(event)<br>    )<br><br>    execution = event[&quot;serviceMetadata&quot;][&quot;executionDetails&quot;]<br><br>    transfer_details = event[&quot;serviceMetadata&quot;][&quot;transferDetails&quot;]<br><br>    username = transfer_details[&quot;userName&quot;]<br><br>    source_path = event[&quot;fileLocation&quot;][&quot;path&quot;]<br><br>    workflow_id = execution[&quot;workflowId&quot;]<br><br>    execution_id = execution[&quot;executionId&quot;]<br><br>    session_id = transfer_details[&quot;sessionId&quot;]<br><br>    logger.info(<br>        &quot;Workflow execution details: workflowId=%s, &quot;<br>        &quot;executionId=%s, sessionId=%s, username=%s, source=%s&quot;,<br>        workflow_id,<br>        execution_id,<br>        session_id,<br>        username,<br>        source_path<br>    )<br><br>    # --------------------------------------------------------<br>    # Phase 1: Validate and archive the file<br>    # --------------------------------------------------------<br><br>    try:<br><br>        source_file = resolve_source_file(<br>            event[&quot;fileLocation&quot;]<br>        )<br><br>        s3_key, file_size, modified_time = generate_s3_key(<br>            username,<br>            source_file<br>        )<br><br>        logger.info(<br>            &quot;Source file details: path=%s, size=%s, &quot;<br>            &quot;lastModifiedUTC=%s&quot;,<br>            source_file,<br>            file_size,<br>            modified_time.isoformat()<br>        )<br><br>        logger.info(<br>            &quot;Archive destination: s3://%s/%s&quot;,<br>            S3_BUCKET,<br>            s3_key<br>        )<br><br>        uploaded_size = archive_to_s3(<br>            source_file,<br>            s3_key<br>        )<br><br>        logger.info(<br>            &quot;S3 archival successful for execution %s&quot;,<br>            execution_id<br>        )<br><br>    except Exception:<br><br>        logger.exception(<br>            &quot;S3 archival operation failed for execution %s&quot;,<br>            execution_id<br>        )<br><br>        report_failure_safely(event)<br><br>        raise<br><br>    # --------------------------------------------------------<br>    # Phase 2: Report SUCCESS to Transfer Family<br>    # --------------------------------------------------------<br><br>    try:<br><br>        report_workflow_status(<br>            event,<br>            &quot;SUCCESS&quot;<br>        )<br><br>    except ClientError as error:<br><br>        error_code = error.response.get(<br>            &quot;Error&quot;, {}<br>        ).get(&quot;Code&quot;)<br><br>        if error_code == &quot;ResourceNotFoundException&quot;:<br><br>            logger.error(<br>                &quot;S3 archive succeeded, but Transfer Family &quot;<br>                &quot;returned Unknown execution. &quot;<br>                &quot;workflowId=%s, executionId=%s, &quot;<br>                &quot;destination=s3://%s/%s&quot;,<br>                workflow_id,<br>                execution_id,<br>                S3_BUCKET,<br>                s3_key<br>            )<br><br>            # Do not report FAILURE here.<br>            # S3 archival has already succeeded and the<br>            # workflow execution is no longer recognized.<br><br>            return {<br>                &quot;statusCode&quot;: 200,<br>                &quot;body&quot;: json.dumps({<br>                    &quot;message&quot;: &quot;S3 archive succeeded, but &quot;<br>                               &quot;Transfer Family execution &quot;<br>                               &quot;was no longer available.&quot;,<br>                    &quot;workflowId&quot;: workflow_id,<br>                    &quot;executionId&quot;: execution_id,<br>                    &quot;destination&quot;: f&quot;s3://{S3_BUCKET}/{s3_key}&quot;,<br>                    &quot;workflowCallbackAccepted&quot;: False<br>                })<br>            }<br><br>        logger.exception(<br>            &quot;S3 archive succeeded, but workflow SUCCESS &quot;<br>            &quot;callback failed.&quot;<br>        )<br><br>        raise<br><br>    except Exception:<br><br>        logger.exception(<br>            &quot;S3 archive succeeded, but workflow SUCCESS &quot;<br>            &quot;callback failed.&quot;<br>        )<br><br>        raise<br><br>    # --------------------------------------------------------<br>    # Phase 3: Return result<br>    # --------------------------------------------------------<br><br>    logger.info(<br>        &quot;File archival and workflow completion successful: &quot;<br>        &quot;s3://%s/%s&quot;,<br>        S3_BUCKET,<br>        s3_key<br>    )<br><br>    return {<br>        &quot;statusCode&quot;: 200,<br>        &quot;body&quot;: json.dumps({<br>            &quot;message&quot;: &quot;File archived successfully&quot;,<br>            &quot;username&quot;: username,<br>            &quot;source_file&quot;: source_path,<br>            &quot;destination&quot;: f&quot;s3://{S3_BUCKET}/{s3_key}&quot;,<br>            &quot;file_size&quot;: uploaded_size,<br>            &quot;workflowId&quot;: workflow_id,<br>            &quot;executionId&quot;: execution_id,<br>            &quot;workflowCallbackAccepted&quot;: True<br>        })<br>    }</pre><p>You can also create other Lambda functions to list users, update passwords, and activate/deactivate users.</p><h4>Set up Workflow</h4><p>You can set up a workflow to archive the files when they are uploaded to SFTP.</p><p>In AWS Transfer Family, set up a workflow with the following parameters and attach it to the Server.</p><ul><li>Type: CUSTOM</li><li>Target: Lambda function ARN</li><li>Source File Location: ${original.file}</li><li>TimeoutSeconds: 1200</li></ul><h4>API Gateway (Optional)</h4><p>You can deploy API Gateway APIs with an API key to proxy Lambda functions for Create user and other user-management-related Lambda functions.</p><p>The setup is complete.</p><p>Now, you can create user(s) and test the connection and functionality using tools like WinSCP.</p><p>I hope this artical provides insights on CloudWatch Omni. If this post was helpful, please do follow and click the clap 👏 button below to show your support.</p><p>Thanks for reading!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=98d477299635" width="1" height="1" alt=""><hr><p><a href="https://aws.plainenglish.io/building-an-end-to-end-managed-sftp-architecture-on-aws-98d477299635">Building an End-to-End Managed SFTP Architecture on AWS</a> was originally published in <a href="https://aws.plainenglish.io/">AWS in Plain English</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[CloudWatch Omni: Let AI Troubleshoot Your AWS Infrastructure and Applications]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://aws.plainenglish.io/cloudwatch-omni-let-ai-troubleshoot-your-aws-infrastructure-and-applications-26afec35215a?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/607/1*SfpxB2wrAFEYPV7vGeiJmQ.png" width="607"></a></p><p class="medium-feed-snippet">It is very challenging to troubleshoot issues in the application by checking logs and each component of the infrastructure. It takes ample&#x2026;</p><p class="medium-feed-link"><a href="https://aws.plainenglish.io/cloudwatch-omni-let-ai-troubleshoot-your-aws-infrastructure-and-applications-26afec35215a?source=rss-445e0e33ac7------2">Continue reading on AWS in Plain English »</a></p></div>]]></description>
            <link>https://aws.plainenglish.io/cloudwatch-omni-let-ai-troubleshoot-your-aws-infrastructure-and-applications-26afec35215a?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/26afec35215a</guid>
            <category><![CDATA[cloudwatch]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[observability]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Wed, 23 Sep 2026 10:41:01 GMT</pubDate>
            <atom:updated>2026-11-17T11:14:26.147Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[How to Set Up Retrieval-Augmented Generation (RAG) in Amazon Bedrock?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://aws.plainenglish.io/how-to-set-up-retrieval-augmented-generation-rag-in-amazon-bedrock-69b8fa4d7189?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/611/1*zHFVqz2ztfh3KQWJiOO_XA.png" width="611"></a></p><p class="medium-feed-snippet">Retrieval-augmented generation (RAG) enables you to customize the LLM models without fine-tuning. It can be helpful when you want your&#x2026;</p><p class="medium-feed-link"><a href="https://aws.plainenglish.io/how-to-set-up-retrieval-augmented-generation-rag-in-amazon-bedrock-69b8fa4d7189?source=rss-445e0e33ac7------2">Continue reading on AWS in Plain English »</a></p></div>]]></description>
            <link>https://aws.plainenglish.io/how-to-set-up-retrieval-augmented-generation-rag-in-amazon-bedrock-69b8fa4d7189?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/69b8fa4d7189</guid>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[amazon-bedrock]]></category>
            <category><![CDATA[rags]]></category>
            <category><![CDATA[retrieval-augmented-gen]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Mon, 24 Nov 2025 17:37:45 GMT</pubDate>
            <atom:updated>2025-11-27T20:12:24.882Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Simplifying Infrastructure as Code with AWS Infrastructure Composer]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://aws.plainenglish.io/simplifying-infrastructure-as-code-with-aws-infrastructure-composer-c394d072fbef?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/1024/1*V8VwG-NYC1OM8-4WOi9czg.png" width="1024"></a></p><p class="medium-feed-snippet">Writing Infrastructure as Code scripts has never been easy to create complex infrastructure. It is also required to keep in the mind all&#x2026;</p><p class="medium-feed-link"><a href="https://aws.plainenglish.io/simplifying-infrastructure-as-code-with-aws-infrastructure-composer-c394d072fbef?source=rss-445e0e33ac7------2">Continue reading on AWS in Plain English »</a></p></div>]]></description>
            <link>https://aws.plainenglish.io/simplifying-infrastructure-as-code-with-aws-infrastructure-composer-c394d072fbef?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/c394d072fbef</guid>
            <category><![CDATA[cloudformation]]></category>
            <category><![CDATA[iac]]></category>
            <category><![CDATA[infrastructure-as-code]]></category>
            <category><![CDATA[infrastructure-composer]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Wed, 29 Oct 2025 16:02:48 GMT</pubDate>
            <atom:updated>2025-11-09T15:51:20.569Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Connect Azure DevOps with AWS]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://aws.plainenglish.io/connect-azure-devops-with-aws-f8dc03cec584?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/1024/1*HKfKm1tRrX3RZ9XfJ4Qdnw.png" width="1024"></a></p><p class="medium-feed-snippet">Azure DevOps is a tool for the pipeline automation where you can create pipeline to build and deploy your code to the respective&#x2026;</p><p class="medium-feed-link"><a href="https://aws.plainenglish.io/connect-azure-devops-with-aws-f8dc03cec584?source=rss-445e0e33ac7------2">Continue reading on AWS in Plain English »</a></p></div>]]></description>
            <link>https://aws.plainenglish.io/connect-azure-devops-with-aws-f8dc03cec584?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/f8dc03cec584</guid>
            <category><![CDATA[service-connection]]></category>
            <category><![CDATA[azure-devops]]></category>
            <category><![CDATA[deployment]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[cicd]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Fri, 01 Aug 2025 12:51:26 GMT</pubDate>
            <atom:updated>2025-08-03T01:28:30.525Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[How to Set Up AWS Network Firewall in Your VPC: A Practical Walkthrough]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://aws.plainenglish.io/how-to-set-up-aws-network-firewall-in-your-vpc-a-practical-walkthrough-df7a3911ea68?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/1280/0*oXAMkw8JeJ8fwbMZ.jpg" width="1280"></a></p><p class="medium-feed-snippet">With AWS Network Firewall, you can create firewall rules that provide fine-grained control over network traffic and easily deploy firewall&#x2026;</p><p class="medium-feed-link"><a href="https://aws.plainenglish.io/how-to-set-up-aws-network-firewall-in-your-vpc-a-practical-walkthrough-df7a3911ea68?source=rss-445e0e33ac7------2">Continue reading on AWS in Plain English »</a></p></div>]]></description>
            <link>https://aws.plainenglish.io/how-to-set-up-aws-network-firewall-in-your-vpc-a-practical-walkthrough-df7a3911ea68?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/df7a3911ea68</guid>
            <category><![CDATA[aws-network-firewall]]></category>
            <category><![CDATA[rules]]></category>
            <category><![CDATA[security]]></category>
            <category><![CDATA[ec2]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Thu, 23 Jan 2025 10:15:37 GMT</pubDate>
            <atom:updated>2025-01-27T08:20:34.396Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Simplify managing multiple EC2 instances using Ansible]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://awstip.com/simplify-managing-multiple-ec2-instances-using-ansible-cc8b7a365728?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/600/1*_AmZGr37XO-mLDKs2T9o-w.png" width="600"></a></p><p class="medium-feed-snippet">While working with large projects having many EC2 instances, it is difficult to manage EC2 instances individually.</p><p class="medium-feed-link"><a href="https://awstip.com/simplify-managing-multiple-ec2-instances-using-ansible-cc8b7a365728?source=rss-445e0e33ac7------2">Continue reading on AWS Tip »</a></p></div>]]></description>
            <link>https://awstip.com/simplify-managing-multiple-ec2-instances-using-ansible-cc8b7a365728?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/cc8b7a365728</guid>
            <category><![CDATA[ansible-playbook]]></category>
            <category><![CDATA[ansible-tutorial]]></category>
            <category><![CDATA[ec2]]></category>
            <category><![CDATA[ansible]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Wed, 06 Nov 2024 08:54:56 GMT</pubDate>
            <atom:updated>2024-11-25T08:36:26.614Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Connect On-Premise site to AWS using Site-to-Site VPN]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://awstip.com/connect-on-premise-site-to-aws-using-site-to-site-vpn-fa360727d9a3?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/899/1*DfXbW5J1ATH2jr_WyJ1_WA.png" width="899"></a></p><p class="medium-feed-snippet">In cloud computing era, sometimes there is a requirement for a Hybrid infrastructure setup. To connect an On-Premise network to AWS&#x2026;</p><p class="medium-feed-link"><a href="https://awstip.com/connect-on-premise-site-to-aws-using-site-to-site-vpn-fa360727d9a3?source=rss-445e0e33ac7------2">Continue reading on AWS Tip »</a></p></div>]]></description>
            <link>https://awstip.com/connect-on-premise-site-to-aws-using-site-to-site-vpn-fa360727d9a3?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/fa360727d9a3</guid>
            <category><![CDATA[hybrid-cloud]]></category>
            <category><![CDATA[site-to-site-vpn]]></category>
            <category><![CDATA[on-premise]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Fri, 27 Sep 2024 13:36:29 GMT</pubDate>
            <atom:updated>2024-11-25T08:36:51.296Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Cross-cloud migration — Google Cloud storage to Amazon S3]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://aws.plainenglish.io/cross-cloud-migration-google-cloud-storage-to-amazon-s3-f78a146f1923?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/1235/1*1oMiF06IkaiORQ8GVrnmVw.png" width="1235"></a></p><p class="medium-feed-snippet">Sometimes, there is a need to migrate data from one cloud to another for various reasons.</p><p class="medium-feed-link"><a href="https://aws.plainenglish.io/cross-cloud-migration-google-cloud-storage-to-amazon-s3-f78a146f1923?source=rss-445e0e33ac7------2">Continue reading on AWS in Plain English »</a></p></div>]]></description>
            <link>https://aws.plainenglish.io/cross-cloud-migration-google-cloud-storage-to-amazon-s3-f78a146f1923?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/f78a146f1923</guid>
            <category><![CDATA[amazon-s3]]></category>
            <category><![CDATA[migration]]></category>
            <category><![CDATA[google-cloud-storage]]></category>
            <category><![CDATA[aws-datasync]]></category>
            <category><![CDATA[cross-cloud-migration]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Fri, 23 Aug 2024 06:51:55 GMT</pubDate>
            <atom:updated>2024-11-25T08:37:16.881Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Simplify ETL operations with AWS Glue DataBrew]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://awstip.com/simplify-etl-operations-with-aws-glue-databrew-761a220f947d?source=rss-445e0e33ac7------2"><img src="https://cdn-images-1.medium.com/max/600/1*k1l_PaTKgbnjEWGJ-Y06Dw.png" width="600"></a></p><p class="medium-feed-snippet">ELT is one of the day-to-day tasks of a data scientist. ETL combines data from different sources into a single place with some&#x2026;</p><p class="medium-feed-link"><a href="https://awstip.com/simplify-etl-operations-with-aws-glue-databrew-761a220f947d?source=rss-445e0e33ac7------2">Continue reading on AWS Tip »</a></p></div>]]></description>
            <link>https://awstip.com/simplify-etl-operations-with-aws-glue-databrew-761a220f947d?source=rss-445e0e33ac7------2</link>
            <guid isPermaLink="false">https://medium.com/p/761a220f947d</guid>
            <category><![CDATA[glue-databrew]]></category>
            <category><![CDATA[glue]]></category>
            <category><![CDATA[amazon-s3]]></category>
            <category><![CDATA[etl]]></category>
            <dc:creator><![CDATA[Aniruddh Fataniya]]></dc:creator>
            <pubDate>Fri, 16 Aug 2024 07:19:20 GMT</pubDate>
            <atom:updated>2024-11-25T08:37:55.181Z</atom:updated>
        </item>
    </channel>
</rss>