Cloud concept · Updated

Transport Layer Security - Transport Layer Security is a cryptographic protocol that protects network communications through encryption, integrity checking, and endpoint authentication.

In short

TLS protects data while it moves between clients, load balancers, applications, and other resources. Effective protection requires more than enabling an encrypted listener: engineers must enforce acceptable protocol versions, validate identities, and protect every required network hop.

Cloud Concepts Hub →

6 key aspects of Transport Layer Security

Effective TLS combines strong cryptography, validated endpoint identities, enforced connection policies, explicit hop coverage, and ongoing operational testing.

01
Confidentiality and integrity

TLS encrypts application traffic and detects unauthorized modification in transit. It does not protect data after a receiving endpoint decrypts it or replace encryption at rest.

02
Certificate-based endpoint authentication

Clients should validate the server certificate's trust chain, validity period, and hostname. Mutual TLS also authenticates clients, but permissions must still be enforced separately.

03
Protocol and cipher enforcement

Configure supported TLS versions and cipher suites through listener policies or application settings. Test that disallowed versions fail, and reject plaintext rather than relying solely on redirects.

04
Explicit termination boundaries

TLS protects a connection only until its termination point. Re-encrypt subsequent hops or use passthrough when the application must terminate the original TLS session.

05
Workload-specific internal coverage

Distributed processing, container networking, and training jobs need explicit inspection of their internal communication paths. Managed API encryption does not automatically cover worker-to-worker or pod-to-pod traffic.

06
Operational verification

Automate certificate renewal and monitor expiration, handshake failures, and client compatibility. Include negative tests for plaintext, obsolete protocols, and invalid certificates.

Cloud services for Transport Layer Security

Elastic Load Balancing can enforce TLS policies at an application entry point, while AWS Certificate Manager supplies certificates for supported integrations. Backend connections require separate encryption decisions. EMR, EKS, and SageMaker AI each have workload-specific controls for internal communications; enabling HTTPS access to their APIs does not establish complete worker-to-worker coverage. EC2 hosts can run application TLS, while qualifying Nitro network encryption adds protection within its supported scope. Combine these capabilities by documenting termination boundaries, configuring each required hop, managing certificate trust and renewal, and testing that insecure connections fail.

Amazon EMRAmazon EMR

EMR security configurations enable in-transit encryption for supported application traffic. Supply the required certificate material and verify release-specific coverage, including application-specific protection for distributed processing traffic.

Amazon EC2Amazon EC2

Applications running on EC2 can implement TLS directly or behind a load balancer. Supported instance types also offer automatic Nitro-based encryption for qualifying inter-instance traffic, but this is a separate network protection mechanism rather than TLS or application identity validation.

Transport Layer Security exam questions

Real practice exam questions on Transport Layer Security, with the correct answer and its explanation. Answer them with every option in the exam simulator.

SOA-C03

A company runs an application on Amazon EC2 instances. Each instance accepts only HTTPS connections on port 8443 and has a valid certificate for api.example.com. Clients must connect to api.example.com on port 443 through a load balancer. A security policy requires the application instances to terminate the original client TLS sessions and prohibits the load balancer from decrypting application traffic. Which load balancer configuration meets these requirements?

Show answerCORRECT ANSWER

Configure a Network Load Balancer with a TCP listener on port 443 and a TCP target group on port 8443.

The TCP listener and TCP target group forward the encrypted TLS traffic without terminating TLS. The application instance performs the TLS handshake with the client and decrypts the requests. Using TCP as the load balancer protocol does not make the application traffic plaintext.

SAP-C03

A software company operates customer-facing HTTPS endpoints on Amazon EC2 instances, Kubernetes containers, and on-premises servers. Each endpoint runs an existing web server that must terminate TLS locally. Customers use standard web browsers, and the company cannot install additional trust anchors on customer devices or move TLS termination to a managed proxy.

The company controls the public DNS zones for all endpoint names. It wants centralized certificate issuance and managed renewal, with the least custom certificate lifecycle logic. An existing deployment platform can securely distribute certificate files and reload web servers. Private keys must be encrypted during transfer, and renewed certificates must be installed before the currently deployed certificates expire.

Which solution best meets these requirements?

Show answerCORRECT ANSWER

Request ACM public certificates with export enabled and use DNS validation. Retain the validation records. Export each certificate, certificate chain, and passphrase-encrypted private key for secure installation through the deployment platform. Automate detection of completed ACM renewals and export and deploy the renewed material to the affected endpoints.

ACM exportable public certificates support deployment on EC2 instances, containers, and on-premises hosts and provide public browser trust. The workflow uses DNS validation and exports the certificate, private key, and certificate chain. ACM manages eligible certificate renewals, but the company must automate exporting and installing renewed material on these non-integrated endpoints. The existing deployment platform can perform that work.

DEA-C01

A company's Python ingestion application uploads sensitive records to Amazon S3 through a mandatory corporate TLS-inspection proxy. The proxy is authorized to inspect the records and establishes a separate HTTPS connection to S3 with server certificate validation enabled. For the application's connection, the proxy presents an unexpired certificate with the correct S3 hostname, signed by the company's private certificate authority (CA).

The application's AWS SDK uses a public CA trust bundle, and uploads fail with a certificate verification error. Company policy requires encryption on both connections and requires the application to validate the certificate presented by the approved proxy. The application cannot bypass the proxy.

Which change should a data engineer make?

Show answerCORRECT ANSWER

Configure the SDK to use a trusted CA bundle that includes the company's private CA certificate, while retaining certificate and hostname verification.

The application cannot build a trusted certificate chain because its current trust bundle does not include the company's private CA. Adding the approved CA to the SDK's trust bundle allows certificate validation without disabling hostname or certificate checks. The proxy already protects and validates its separate connection to S3.

Practice Transport Layer Security in a full exam

Questions on Transport Layer Security in each practice exam, with every answer option and an explanation for each:

Start a free practice exam

What is Transport Layer Security

3 min read

Transport Layer Security, usually called TLS, establishes a protected communication channel over an untrusted network. During the handshake, endpoints negotiate protocol parameters and establish cryptographic keys. In a typical HTTPS connection, the server presents a certificate, and the client verifies its trust chain, validity period, and hostname. Session keys then protect application traffic against disclosure and undetected modification. Mutual TLS additionally authenticates the client using a certificate, but application authorization remains a separate responsibility.

In cloud architectures, start by mapping where connections originate, terminate, and restart. A browser might connect over HTTPS to a load balancer, which then opens another connection to an application. TLS on the public listener protects only the first connection. If the application connection uses HTTP, that second hop is not protected by TLS. Configure HTTPS targets when backend encryption is required, or use a TCP listener to pass TLS through to the application when the application must own the TLS session.

Requiring encryption is different from merely supporting it. Select listener security policies that permit the required modern protocols and cipher suites, typically TLS 1.2 and TLS 1.3 where supported. Disable obsolete protocols and reject plaintext connections where strict encryption is required. An HTTP-to-HTTPS redirect helps clients find the secure endpoint, but the original HTTP request has already traveled without TLS. Security groups restrict network reachability and ports; allowing port 443 alone does not verify that traffic actually uses TLS.

Certificate validation is as important as encryption. A client that ignores certificate errors can establish an encrypted connection to an attacker rather than the intended server. Maintain appropriate trust stores, protect private keys, automate certificate renewal, and test hostname validation. Do not assume every managed proxy validates backend certificates: an Application Load Balancer encrypts connections to HTTPS targets but does not validate their server certificates. If authenticated backend identity is required, choose an architecture that explicitly supplies that verification.

Inter-resource protection also applies to distributed workloads. Amazon EMR security configurations can enable in-transit encryption for supported application traffic, with certificate material and application-specific settings appropriate to the release. Check coverage for each framework and traffic path rather than assuming every cluster connection uses TLS. In Amazon EKS, the Kubernetes API endpoint supports TLS, but that does not automatically encrypt arbitrary application traffic between pods. Applications or appropriately configured proxies must provide TLS or mutual TLS for those connections.

For supported distributed training workloads in Amazon SageMaker AI, enable inter-container traffic encryption to protect communications between training containers on different instances. Treat this separately from HTTPS access to service APIs and from encryption of training data at rest. Similarly, supported Amazon EC2 instance types provide automatic Nitro-based encryption for qualifying traffic between instances. Nitro network encryption is not TLS, does not provide application certificate authentication, and has documented instance and network-path constraints. It does not replace TLS for connections outside its coverage.

TLS introduces handshake work, certificate lifecycle management, and potential compatibility issues with older clients. Connection reuse and session resumption can reduce overhead. Termination at a managed load balancer centralizes certificate operations and traffic inspection, while passthrough keeps TLS ownership with the application and limits intermediary visibility. Validate the design with successful secure connections, rejected plaintext and obsolete-protocol attempts, certificate-error tests, and checks of every backend hop. Encryption at rest, private subnets, and private connectivity are complementary controls, not substitutes for authenticated encryption in transit.

Learn Transport Layer Security in lessons

9 short lessons on Transport Layer Security in 5 certification courses.

and 3 more lessons

Open the lessons