Cloud Services Cross-Reference: Integration, Messaging & Migration

This document maps integration, messaging, migration, and IoT services across AWS, Azure, Oracle Cloud Infrastructure (OCI), and Google Cloud Platform (GCP). These categories are tightly related in practice: event-driven messaging underpins modern application integration, and migration tooling increasingly relies on streaming and event infrastructure. All four providers offer message queues, pub/sub, event buses, and workflow orchestration, but differ substantially in service granularity, pricing model, and depth of managed tooling. Migration and IoT services show the starkest gaps: OCI lacks a native server migration tool comparable to AWS MGN or Azure Migrate, GCP retired its IoT Core service in 2023, and both AWS Snowball Edge and OCI Data Transfer have closed to new customers in 2025 in favor of online transfer alternatives.

1. Integration & Messaging

1.1 Message Queues

Message queues decouple producers and consumers through an intermediary store-and-forward buffer. Messages are consumed by a single receiver (point-to-point), distinguishing queues from pub/sub topics.

AWS — Amazon SQS (Simple Queue Service) SQS is a fully managed, horizontally scalable message queue supporting Standard queues (at-least-once delivery, best-effort ordering, near-unlimited throughput) and FIFO queues (exactly-once processing, strict ordering, up to 30,000 transactions per second with batching). Dead-letter queues (DLQs) capture unconsumed messages after a configurable number of receive attempts. SQS has no message broker infrastructure to manage; it scales automatically.

Azure — Azure Queue Storage / Azure Service Bus Queues Azure offers two queue products serving different tiers. Azure Queue Storage is a simple, low-cost HTTP-accessible queue suitable for loose coupling in large workloads; it stores millions of messages accessible via authenticated HTTP calls. Azure Service Bus Queues are a brokered messaging service supporting AMQP 1.0, sessions (ordered processing of related messages), dead-lettering, duplicate detection, transactions, and scheduled delivery. For enterprise integration patterns, Service Bus Queues are the preferred choice over Queue Storage.

OCI — OCI Queue OCI Queue is a fully managed, serverless message queuing service that automatically scales to meet workload demand. It provides dead-letter queues by default, ensures at-least-once delivery, and supports STOMP protocol over WebSocket. The service is designed for application decoupling and asynchronous event-driven architectures across OCI services.

GCP — Cloud Tasks Cloud Tasks is a fully managed asynchronous task execution service. It manages the queuing, dispatching, and tracking of task execution against HTTP targets (including Cloud Functions and Cloud Run). Cloud Tasks is designed for task dispatch rather than generalized message queuing; for broader pub/sub messaging, GCP directs users to Pub/Sub.

Key differentiators:

• Azure Service Bus Queues offer the richest enterprise messaging features (AMQP, sessions, transactions, scheduled delivery).
• AWS SQS FIFO provides exactly-once processing, not available natively on the other platforms.
• OCI Queue's STOMP/WebSocket protocol support distinguishes it from the others; it is the newest of the four services.
• GCP Cloud Tasks is purpose-built for HTTP task dispatch, not general message queuing; GCP has no SQS-equivalent.

1.2 Pub/Sub (Publish/Subscribe)

Pub/Sub services fan out messages from a single publisher to multiple subscribers. Every subscriber receives every message, distinguishing this pattern from point-to-point queues.

AWS — Amazon SNS (Simple Notification Service) SNS is a managed pub/sub messaging service supporting topic-based fan-out to SQS queues, Lambda functions, HTTP/S endpoints, email addresses, SMS numbers, and mobile push notification endpoints. Subscription filter policies allow subscribers to receive only the messages matching specified attribute conditions. SNS supports FIFO topics (ordered, deduplicated) when paired with SQS FIFO queues.

Azure — Azure Service Bus Topics Service Bus Topics deliver pub/sub messaging within the Service Bus brokered messaging tier. A topic can have multiple subscriptions, each acting as an independent queue with its own filter rules and actions. This model allows complex message routing where different consumers receive different subsets of messages based on SQL-filter expressions.

OCI — OCI Notifications OCI Notifications is a pub/sub service that broadcasts messages to distributed components via subscriptions. Supported subscription protocols include email, HTTPS (webhooks), PagerDuty, Slack, and OCI Functions. The service integrates natively with OCI Events and OCI Alarms to trigger notifications in response to resource state changes or metric threshold breaches.

GCP — Cloud Pub/Sub Cloud Pub/Sub is GCP's core asynchronous messaging service providing durable, high-throughput message delivery between independent applications. It supports push delivery (GCP calls the subscriber's endpoint) and pull delivery (subscribers poll). Pub/Sub Lite is a lower-cost, zonal alternative with reduced durability guarantees. Pub/Sub integrates with Dataflow, Eventarc, and BigQuery for streaming analytics pipelines.

Key differentiators:

• GCP Pub/Sub is the highest-throughput, most flexible pub/sub service; pull model allows consumers to pace their intake.
• AWS SNS integrates mobile push notification delivery directly; no separate service required.
• Azure Service Bus Topics offer the most powerful server-side message filtering (SQL expressions) of any of these services.
• OCI Notifications is tightly integrated with OCI resource events and alarms but has limited protocol support compared to SNS.

1.3 Event Bus

Event bus services route discrete events from sources (cloud services, applications, SaaS) to targets based on rules or patterns. Unlike message queues, events represent state changes rather than work items.

AWS — Amazon EventBridge EventBridge is a serverless event bus with three bus types: default (all AWS service events), custom (application events), and partner (SaaS events from 130+ partners). Events are matched against rules using JSON-based event patterns; matched events are routed to 40+ target types including Lambda, SQS, SNS, Step Functions, and API Gateway. EventBridge Pipes provides point-to-point integration with filtering and enrichment between source and target. EventBridge Scheduler enables scheduled and recurring event delivery.

Azure — Azure Event Grid Event Grid is a fully managed event routing service supporting both CloudEvents 1.0 and custom schemas. It routes events from Azure services (blob storage, resource groups, subscriptions), custom applications, and partner topics to subscribers via push delivery. Event Grid Namespaces (generally available 2024) introduced pull delivery, MQTT broker support (for IoT scenarios), and topic spaces—extending Event Grid beyond pure eventing into a hybrid IoT/event use case.

OCI — OCI Events OCI Events tracks resource state changes across OCI services and routes them to actions via rules. Supported actions are OCI Notifications (pub/sub fan-out), OCI Streaming (Kafka-compatible stream), and OCI Functions (serverless). Events comply with CNCF CloudEvents specification. The service is primarily designed for infrastructure-event-driven automation (e.g., trigger a Function when an object is uploaded to Object Storage).

GCP — Eventarc Eventarc is GCP's unified eventing layer, routing events from Google Cloud services, custom sources (via Pub/Sub), and third-party sources to Cloud Run, Cloud Functions, GKE, and Cloud Workflows targets. Events are delivered in CloudEvents format. Eventarc Standard supports trigger types from Cloud Storage, Pub/Sub, Cloud Logging, and direct event sources. Integration with Workflows enables multi-step event-driven orchestration.

Key differentiators:

• AWS EventBridge is the most feature-rich event bus: SaaS partners, scheduler, Pipes, and event replay are unique at this scope.
• Azure Event Grid Namespaces (2024) adds MQTT broker and pull delivery, bridging event-driven and IoT messaging.
• OCI Events is narrower in scope—primarily for OCI resource change events; custom application events use OCI Streaming instead.
• GCP Eventarc's native Workflows integration is the cleanest path to multi-step event-driven orchestration of the four providers.

1.4 Workflow Orchestration

Workflow orchestration services coordinate multi-step processes across services and functions, handling state, retries, branching, and error handling.

AWS — AWS Step Functions Step Functions is a serverless workflow service using Amazon States Language (ASL), a JSON-based language for defining state machines. It supports Standard workflows (long-running, auditable, exactly-once execution) and Express workflows (high-volume, short-duration, at-least-once). Step Functions integrates with 220+ AWS services via SDK integrations and supports parallel execution, map states (dynamic parallelism), wait states, and error handling. AWS Step Functions has no managed visual low-code designer; workflows are defined as JSON or using the Workflow Studio drag-and-drop builder in the console.

Azure — Azure Logic Apps / Durable Functions Azure provides two workflow orchestration approaches. Azure Logic Apps is a low-code integration platform with a visual designer, 1,000+ connectors, and built-in triggers for Azure services, SaaS, and on-premises systems (via Data Gateway). Logic Apps Standard runs on the Azure Functions v4 runtime and supports local development and CI/CD. Azure Durable Functions is a code-first (C#, Python, JavaScript, Java) framework for stateful orchestration within the Azure Functions programming model, using the orchestrator/activity pattern. Both converge on the same runtime in 2025. For enterprise low-code integration, Logic Apps is the primary choice; for developer-code orchestration, Durable Functions is preferred.

OCI — Oracle Integration Cloud (OIC) Oracle Integration Cloud is a comprehensive iPaaS (Integration Platform as a Service) supporting REST/SOAP adapters, 200+ prebuilt connectors (Oracle ERP, HCM, SCM, Salesforce, SAP), visual workflow designer, process automation (BPMN 2.0-based), and B2B EDI capabilities. OIC is more fully featured than Step Functions or Workflows for enterprise application integration scenarios. OIC 3 (current generation) runs on OCI and supports event-driven architectures, AI document understanding (2025), and OCI Functions integration.

GCP — Cloud Workflows Cloud Workflows is a fully managed, serverless orchestration service using a YAML/JSON syntax. It supports HTTP step execution, parallel steps, conditional branching, retry policies, and callbacks (waiting for external signals). Workflows integrates with Eventarc for event-triggered execution. It is positioned between the simplicity of a task queue and the complexity of a full integration platform—suited for cloud-service orchestration rather than enterprise application integration.

Key differentiators:

• OCI Integration Cloud is the most fully featured enterprise iPaaS of the four, with native B2B, BPMN, and deep Oracle application connectors.
• Azure Logic Apps has the largest connector catalog (1,000+) and is the only one with a dedicated no-code audience path (Power Automate integration).
• AWS Step Functions has the deepest native AWS service integration with 220+ SDK-native integrations requiring no custom HTTP calls.
• GCP Cloud Workflows is the leanest service—ideal for orchestrating cloud APIs and serverless functions but not enterprise application integration.

1.5 Email Services

Managed email delivery services handle transactional and marketing email sending, reputation management, DKIM/SPF configuration, and delivery analytics.

AWS — Amazon SES (Simple Email Service) SES is a high-volume transactional and marketing email service supporting SMTP, AWS SDK, and the SES API. It provides bounce/complaint handling, email receiving (inbound), dedicated IP address management, configuration sets for delivery analytics, and a virtual deliverability manager for automated reputation guidance. SES is widely used for application-generated emails (order confirmations, password resets, notifications) at very low per-message cost ($0.10 per 1,000 emails).

Azure — Azure Communication Services (Email) Azure Communication Services includes an Email channel for transactional email sending. It provides managed sender domain provisioning, DKIM/SPF setup, bounce/complaint event handling via Event Grid, and delivery analytics. The service supports Azure-managed domains (for rapid start) or customer-managed custom domains. It is comparable to SES in scope, positioned as a developer-focused transactional email API integrated into the Azure Communication Services SDK.

OCI — OCI Email Delivery OCI Email Delivery is a transactional email service for OCI-hosted applications. It supports SMTP relay and API-based sending, approved sender management, suppression list management, and SPF/DKIM configuration. Delivery metrics are accessible via OCI Monitoring. It is comparable in scope to SES and Azure Communication Services Email.

GCP — No native equivalent GCP does not offer a first-party email delivery service. Google recommends SendGrid (owned by Twilio, available on Google Cloud Marketplace), Mailgun, or Mailjet for transactional email from GCP workloads.

Key differentiator: AWS SES is the only first-party email service with inbound email processing. GCP has no native email delivery offering.

1.6 Notification Services (Mobile Push)

Mobile push notification services deliver messages to iOS, Android, and web browser clients at scale, abstracting platform-specific push notification services (APNs, FCM, WNS).

AWS — Amazon SNS (Simple Notification Service) SNS handles mobile push directly as one of its subscription endpoint types. SNS integrates with APNs (iOS), FCM/APNS (Android), Amazon Device Messaging (Kindle), and Windows Push Notification Services (WNS). SNS manages device token registration, platform-specific credential management, and delivery at scale without a separate service.

Azure — Azure Notification Hubs Notification Hubs is a dedicated mobile push notification engine, separate from Service Bus and Event Grid. It abstracts APNs, FCM (using OAuth 2.0 service account credentials as of 2024, replacing the legacy GCM server key), WNS, Baidu, and Amazon ADM. Notification Hubs supports tag-based targeting (send to all users with tag "sports"), template notifications (localized messages), and telemetry for delivery tracking. It is available in Free, Basic, and Standard tiers.

OCI — OCI Notifications OCI Notifications serves as the general-purpose notification service for OCI. It does not natively support mobile push to APNs or FCM. For mobile push from OCI, applications must integrate directly with Firebase Cloud Messaging, APNs, or a third-party service.

GCP — Firebase Cloud Messaging (FCM) FCM is GCP's cross-platform mobile push notification service, covering Android, iOS, and web. FCM delivers notification messages and data messages, supports topic subscriptions (fan-out to groups of devices), single-device targeting, and device group messaging. FCM is free of charge and deeply integrated with Google Analytics for engagement tracking. FCM serves as the underlying push infrastructure for multiple other cloud providers' notification services (Azure Notification Hubs, AWS SNS for Android both call FCM).

Key differentiators:

• FCM is the backend for Android push across all providers; AWS SNS and Azure Notification Hubs both call FCM for Android delivery.
• Azure Notification Hubs is the most fully featured dedicated mobile push service with tag-based audience targeting and templates.
• OCI has no native mobile push capability; applications must call FCM or APNs directly.

2. Migration

2.1 Database Migration

Managed database migration services orchestrate schema conversion, data replication, and cutover for both homogeneous (same engine) and heterogeneous (different engine) database migrations.

AWS — AWS Database Migration Service (DMS) AWS DMS migrates databases from on-premises, other clouds, or between AWS regions. It supports Oracle, SQL Server, PostgreSQL, MySQL, MongoDB, MariaDB, SAP ASE, and 20+ additional sources and targets. DMS Schema Conversion (formerly AWS SCM) handles heterogeneous schema and code conversion. Continuous Data Replication (CDC) via AWS DMS keeps the source database online during migration, enabling minimal-downtime cutovers. DMS Fleet Advisor inventories on-premises database estates and recommends migration targets.

Azure — Azure Database Migration Service Azure DMS provides both offline and online (minimal-downtime) migration paths for SQL Server to Azure SQL Database, Azure SQL Managed Instance, and SQL Server on Azure VMs; MySQL, PostgreSQL, and MongoDB to Azure PaaS database targets. Azure Migrate integrates DMS as the database migration component within the broader migration platform. The service also handles SQL Server agent jobs, logins, and SSIS packages as part of migration.

OCI — OCI Database Migration OCI Database Migration is a free managed service for Oracle Database (any version, any OS) and MySQL HeatWave migrations. For Oracle database migrations, it transparently orchestrates Zero Downtime Migration (ZDM), Cloud Premigration Advisor Tool (CPAT), Data Pump, and GoldenGate. It supports logical offline migration (downtime required) and logical online migration (CDC replication, near-zero downtime). The MySQL HeatWave Migration Assistant provides parallel data streaming to OCI Object Storage concurrent with HeatWave instance provisioning, minimizing total migration time.

GCP — Database Migration Service GCP Database Migration Service handles homogeneous migrations (MySQL to Cloud SQL MySQL, PostgreSQL to Cloud SQL PostgreSQL or AlloyDB, SQL Server to Cloud SQL SQL Server) at no additional charge. Heterogeneous migrations (Oracle to PostgreSQL) are available but billed separately. The service uses CDC (via Change Data Capture) for minimal-downtime migrations. BigQuery Migration Service is a separate offering for data warehouse migrations (Teradata, Redshift, Snowflake, Netezza to BigQuery).

Key differentiators:

• OCI Database Migration is the strongest Oracle-to-Oracle migration service, with ZDM and GoldenGate orchestration built in.
• Azure DMS is the strongest SQL Server migration service, handling SQL Agent jobs, logins, and SSIS packages.
• GCP offers homogeneous migration at no charge; this is attractive for MySQL/PostgreSQL lift-and-shift scenarios.
• AWS DMS supports the broadest source/target matrix across all four providers.

2.2 Server / VM Migration

Server migration services automate the replication and conversion of physical or virtual machines to cloud compute instances with minimal downtime.

AWS — AWS Application Migration Service (MGN) AWS MGN (formerly CloudEndure Migration) is the primary AWS lift-and-shift tool. It installs a lightweight replication agent on source machines, continuously replicates data to AWS, and launches test or cutover instances. Source environments can be physical, VMware, Hyper-V, or any cloud. MGN minimizes downtime by maintaining a continuously replicated staging area; cutover requires only minutes of downtime. AWS Migration Hub provides centralized tracking across MGN and DMS migrations.

Azure — Azure Migrate Azure Migrate is a unified migration hub covering discovery and assessment of on-premises VMs (VMware, Hyper-V, physical), databases, web apps, and VDI. The Azure Migrate: Server Migration tool handles replication and cutover of VMware VMs (agentless), Hyper-V VMs, and physical/other-cloud servers (agent-based) to Azure VMs. Azure Migrate integrates with Azure Database Migration Service, Azure App Service Migration Assistant, and Azure Data Box for a complete migration factory experience.

OCI — Oracle Cloud Migrations Oracle Cloud Migrations is OCI's native, free VM migration service. It automates end-to-end migration of VMware VMs (using an on-premises agent) and AWS EC2 instances (agentless from OCI) to OCI Compute VMs. The service includes asset discovery, migration planning with automated target configuration recommendations, secure data replication, and launch automation. Oracle Cloud Migrations does not support Hyper-V source environments or physical-to-cloud (P2C) migrations; those scenarios rely on partner tools (RackWare, ZConverter, Cristie CloneManager).

GCP — Migrate to Virtual Machines Migrate to Virtual Machines (formerly Migrate for Compute Engine, formerly Velostrata) replicates VMs from VMware, AWS, Azure, and physical servers to GCP Compute Engine. It performs live migration with minimal downtime, automatically converts VM disk formats, and supports continuous replication. Migrate to Containers is a related service for modernizing applications into containerized form for GKE or Cloud Run during migration.

Key differentiators:

• Azure Migrate has the broadest source coverage and the most integrated hub experience (databases, VMs, web apps, VDI in one place).
• AWS MGN is the most battle-tested lift-and-shift agent for heterogeneous physical/virtual/cloud sources.
• OCI Cloud Migrations is limited to VMware and AWS sources; Hyper-V and P2C require third-party partner tools.
• GCP Migrate to VMs supports the agentless VMware path for minimal source-side footprint.

2.3 Data Transfer Appliances

Physical data transfer appliances ship encrypted storage devices between customer premises and cloud provider facilities when network transfer is impractical for large datasets (petabyte-scale, limited bandwidth, regulatory constraints).

AWS — AWS Snowball The AWS Snow family consists of Snowball Edge (Storage Optimized: 80 TB usable / 210 TB NVMe raw; Compute Optimized: 28 TB NVMe + EC2 instance capacity), Snowcone (8–14 TB, portable), and the discontinued Snowmobile (exabyte-scale). AWS Snowball Edge is no longer available to new customers effective November 7, 2025. AWS directs new requirements to AWS DataSync (online), AWS Data Transfer Terminal (secure physical courier), or AWS Partner solutions.

Azure — Azure Data Box Azure Data Box family includes Data Box Disk (up to 35 TB, SSD-based, simple import), Data Box (100 TB, ruggedized appliance with NFS/SMB), Data Box Heavy (1 PB, wheeled device for large migrations), and Data Box Offline Export (exporting from Azure to on-premises). Data Box remains actively available to new customers and is integrated with Azure Migrate for large dataset offline migration.

OCI — OCI Data Transfer (EOL) OCI Data Transfer Service reached end-of-distribution and end-of-life effective February 6, 2025. No new customer orders are accepted. Oracle recommends OCI Object Storage sync, Rclone, Roving Edge as Data Transfer Gateway (up to 45 TB), or Seagate Lyve for physical data transfer requirements.

GCP — Transfer Appliance Google Cloud Transfer Appliance is available in 100 TB and 480 TB capacities. Data is loaded on-premises, the device is shipped to a Google facility, and the data is uploaded to Cloud Storage. The service remains available for new customers and is suitable for petabyte-scale offline migration scenarios.

Key differentiators:

• Azure Data Box has the widest active product family (disk, appliance, heavy) and the only petabyte-class appliance still accepting new customers.
• AWS Snowball Edge had a unique edge compute capability (running EC2 workloads on the device in disconnected environments); this capability has no equivalent in the other providers' appliances.
• Both AWS Snowball Edge and OCI Data Transfer are closed to new customers as of 2025; GCP Transfer Appliance and Azure Data Box remain the active choices.

2.4 Application Migration Tools

Application migration tooling helps assess, refactor, or re-platform web applications, PaaS workloads, and legacy application stacks.

Key differentiators:

• Azure Migrate is the most integrated assessment hub, covering servers, databases, web apps, and VDI in a single pane of glass.
• OCI provides the deepest Oracle application migration tooling (WebLogic, E-Business Suite Cloud Manager, JD Edwards) of any provider.
• GCP's Kf is unique in providing a supported Cloud Foundry-to-Kubernetes migration path.
• AWS Mainframe Modernization is the most comprehensive managed mainframe re-platforming service available on a public cloud.

3. IoT

3.1 IoT Platform

IoT platforms provide device connectivity, device management (provisioning, OTA updates, twin/shadow state), telemetry ingestion, and routing to backend analytics or storage.

AWS — AWS IoT Core AWS IoT Core is the primary managed IoT connectivity service, providing a scalable MQTT broker (MQTT 3.1.1 and MQTT 5, including WebSocket), HTTPS, and LoRaWAN connectivity. It includes a Rules Engine for routing messages to Lambda, SQS, SNS, Kinesis, S3, DynamoDB, and other services. IoT Core Device Shadows maintain a persistent JSON representation of device state. AWS IoT Device Management handles fleet indexing, group management, OTA jobs, and tunneling. In March 2025, AWS launched managed integrations for IoT Device Management, supporting ZigBee, Z-Wave, and Wi-Fi protocols with 80+ device data models.

Azure — Azure IoT Hub Azure IoT Hub is a managed backend for bidirectional device-cloud and cloud-device communication. It supports MQTT 3.1.1, AMQP, and HTTPS protocols. IoT Hub Device Twins maintain desired and reported property states, and Direct Methods enable synchronous device invocation. IoT Hub integrates with Azure Event Hubs for telemetry ingestion at scale and Azure Stream Analytics, Time Series Insights, and Azure Digital Twins for downstream processing. Azure IoT Operations (2025) is a new arc-enabled service that extends cloud capabilities to on-premises industrial environments using MQTT broker, data processor pipelines, and OPC UA connectivity.

OCI — OCI IoT Platform OCI IoT Platform is a PaaS service providing device agnostic data ingestion via MQTT, HTTP, and MQTT over WebSocket. It uses digital twin models and adapters to normalize device data into a standardized format stored in Oracle Autonomous AI Database. The platform supports GenAI integration for real-time IoT data processing and targets industrial, smart building, and connected device use cases. It is a newer service compared to AWS IoT Core and Azure IoT Hub.

GCP — Cloud IoT Core (Retired August 2023) Google Cloud IoT Core was retired on August 16, 2023. Google cited the inability to adequately serve customers with specific IoT needs and directed customers to ecosystem partners. GCP currently has no first-party managed IoT platform. Google recommends EMQX, HiveMQ, or other partner MQTT brokers on GCP, combined with Pub/Sub, Dataflow, and BigQuery for telemetry processing.

Key differentiator: GCP is the only major cloud provider without an active first-party IoT platform. AWS IoT Core is the most feature-complete, with MQTT 5, LoRaWAN, and edge computing integration. Azure IoT Operations (2025) extends IoT Hub to industrial environments with OPC UA.

3.2 IoT Edge

IoT edge services deploy cloud workloads—analytics, ML inference, custom logic—to devices at the network edge, enabling local processing with or without cloud connectivity.

AWS — AWS IoT Greengrass AWS IoT Greengrass is an open-source edge runtime and cloud service for building and deploying edge device software. Greengrass Core runs on Linux-based devices and supports containerized and Lambda function components deployed from the cloud. Pre-built components accelerate common scenarios (stream manager, ML inference, MQTT bridge, Docker application manager). The Greengrass Nucleus Lite runtime (released 2024) operates on as little as 5 MB of memory, extending cloud connectivity to constrained devices. Greengrass v2 is the current generation; v1 reached end-of-life in June 2023.

Azure — Azure IoT Edge Azure IoT Edge is a device-focused runtime for deploying containerized Linux workloads to edge devices. Modules run as Docker-compatible containers and can include Azure services (Stream Analytics, Machine Learning, Functions), custom code, or third-party modules from Azure Marketplace. IoT Edge integrates with IoT Hub for remote deployment, monitoring, and management of edge modules. Azure IoT Edge supports both connected and offline (disconnected) operation.

OCI — No native IoT Edge service OCI does not offer a first-party IoT edge runtime comparable to Greengrass or IoT Edge. OCI Roving Edge Infrastructure (powered by OCI Compute inside a rugged appliance) provides cloud-compatible compute at the edge for more substantial workloads, but is not a device-level IoT edge runtime.

GCP — No dedicated IoT Edge service With the retirement of Cloud IoT Core, GCP has no dedicated IoT edge runtime. GCP Edge TPU and Coral dev board ecosystem provide ML inference at the edge, but no managed IoT edge deployment and management service is available as of 2025.

3.3 Digital Twins

Digital twin platforms create virtual representations of physical environments, equipment, or systems, linked to real-time sensor data and enabling simulation, monitoring, and predictive analytics.

AWS — AWS IoT TwinMaker AWS IoT TwinMaker creates digital visualizations using data from sensors, cameras, and enterprise applications to represent physical factories, buildings, and industrial plants. It builds a knowledge graph linking physical and logical components, and integrates with IoT SiteWise (industrial sensor data), S3, and third-party data sources. Grafana-based dashboards are the primary visualization layer.

Azure — Azure Digital Twins Azure Digital Twins is a PaaS service for creating complete digital models of entire environments (buildings, factories, cities, energy networks). Models are defined using Digital Twins Definition Language (DTDL), a JSON-LD based language. Azure Digital Twins integrates with IoT Hub (device data ingestion), Azure Maps (spatial context), Azure Synapse Analytics (historical analysis), and Azure Time Series Insights. The service supports live twin graph updates from IoT data and historical data querying.

OCI — No dedicated Digital Twins service OCI does not offer a first-party digital twins platform. Oracle's industrial IoT use cases are addressed through OCI IoT Platform's digital twin models and Oracle's industry-specific cloud applications (Oracle Utilities, Oracle Construction and Engineering).

GCP — No dedicated Digital Twins service GCP does not offer a first-party digital twins service. Google's Connected Devices and Smart Cities use cases are addressed through partners or third-party solutions (e.g., Bentley iTwin on GCP).

Key differentiator: Azure Digital Twins is the most mature and fully featured digital twins platform of the group. AWS IoT TwinMaker is primarily industrial-visualization-focused. OCI and GCP have no dedicated first-party offering.

Summary: Service Coverage Matrix

Notable gaps:

• GCP has no native email delivery service, no IoT platform, and no IoT edge runtime.
• OCI has no native IoT edge runtime, no digital twins service, and no mobile push notification service.
• AWS Snowball Edge and OCI Data Transfer are both closed to new customers as of 2025.

References

• Amazon SQS, SNS, or EventBridge? — AWS Decision Guide
• AWS Database Migration Service
• AWS Application Migration Service
• AWS IoT Overview
• AWS IoT Greengrass v2 Release Notes — Nucleus Lite
• AWS Snowball Edge Availability Change
• Azure Asynchronous Messaging Options — Architecture Center
• Azure Migrate: Overview
• Azure Notification Hubs Overview
• Azure Digital Twins Overview
• Azure IoT Services and Technologies
• OCI Data Transfer End of Life
• Oracle Cloud Migrations Overview
• OCI Database Migration
• OCI IoT Platform
• GCP Database Migration Service
• GCP Migration Documentation
• Firebase Cloud Messaging
• Google Cloud IoT Core Retirement (2023)
• Oracle Integration Cloud (OIC)