azure / integration-services-landing-zone-accelerator Goto Github PK

Integration scenarios are spread across multiple services. A failure or degradation within a service could potentially cause a ripple effect through the system. Show how to leverage Azure Load Test service to drive load against the integration scenario to ensure it is able to handle the expected load. At the same time, show how to build a Chaos Studio experiment which will simulate a degraded service or service outage during the load test so app resilience is understood.

Show how to incorporate load testing and chaos testing within AzDo and GitHub pipelines.

In the documentation, it is discussed that LogicApps can leverage SQL Server for storage and persistence instead of an Azure Storage account. How difficult would it be to also have the ability to leverage CosmosDB. My thoughts here are, if we can use CosmosDB, that could lead to a logicApp that runs in multiple regions. Outage in region A, logicApp keeps running without issue since it was running in regions A and B simultaneously.

I think this would be a great talking point during compete scenarios.

Would be great to see a section about typical costs associated with this landing zone accelerator and how to best manage cost within Azure Integration Services.

The Logic Apps deployment script (logicApp.json) references V3 of Functions. Now that Logic Apps Supports V4 this can be updated to reflect latest version of runtime

"appSettings": [
{
"name": "FUNCTIONS_EXTENSION_VERSION",
"value": "~3"
},

Keen to understand it the logic apps implements the pattern mentioned at https://learn.microsoft.com/en-us/azure/logic-apps/deploy-single-tenant-logic-apps-private-storage-account and associated code https://github.com/VeeraMS/LogicApp-deployment-with-Secure-Storage

looking at the code at https://github.com/Azure/Integration-Services-Landing-Zone-Accelerator/blob/main/src/infra/scenario1/la/logicApp.json it seems like is enabled
"publicNetworkAccess": "Enabled", // Need to disable after deployment of all artefacts.

Show a customer the best way to setup functional testing of a integration scenario which leverages multiple different azure services. Want to show how to deploy the solution via a pipeline, run some functional tests from the pipeline, and based on the results, make a deploy don't deploy decision.

Azure App Configuration should be a key part of this landing zone; it helps organise all the different environments.

Show how to receive n related messages for processing. Correlate based on message contents.

• independently deployable
• self-contained demo
• independent documentation
• intended to be built upon
• message tracking and reporting with promoted properties
• Show completion based on control message, number of messages, duration
• Show how to raise alerts if messages no subsequent messages received if expected

Deployable demo that illustrates large message handling with a claim check pattern.

• demonstrate how to read message contents in order to promote properties
• demonstrate how to store the large message
• self contained demo
• independently deployable demo
• independent documentation

How can we leverage HIS capabilities with logicApps? Mainframes are still around and integration needs persist. What are the best practices, what technologies do you need.

Deployable Demo that shows how to receive two different, related messages and continue processing only when both messages have been received.

• Self-contained demo
• independently deployable
• independent documentation
• simple, intended to be built upon
• include message tracking and property promotion

Would be great to see a reference implementation that uses Service Endpoints instead of Private Endpoints; as not all enterprise grade implementations mandate the need for using Private Endpoints.

Adding in examples and guidance for session management for functions and logic apps

if the selling point of scenario 1 is to make networking easier in AIS, then you're definitely missing AppGW, Firewall and NATGW for ingress and egress. those are the ones that are difficult to setup. Make them optional please.

I'd also suggest to include a sample hello world app in scenario 1

Show how to implement a message envelope where important message information is promoted and contained within a common envelope structure allowing for flexible message contents.

• find data within a message
• promote important data into envelope
• How to process across LogicApp, Functions, ServiceBus, etc.
• Envelope tracking and reporting

While we have EDI batching... so of, there are other scenarios where we need to batch and debatch messages. Need to have a pattern than makes this simple to implement, track, and manage for our customers.

Demonstrate an ESB solution using iPaaS similar to the BizTalk ESB toolkit.

• itinerary processing
• onramp implementation
• offramp implementation
• property promotion
• message tracking and reporting

Guidance, tips and tricks regarding LogicApp Standard Connector development.

• Simplified documentation and examples
• Visual Studio Scaffolding to get you started, much like asp.net scaffolding, ask some questions, generat3e the code
• Detailed guidance of how to setup a local development environment where you can debug and step through code when running the designer and when running an instance
• Discuss how to deploy the custom connector with the logicApp when it is being deployed through Azure Arc

Help a customer understand how to increase the resilience of an integration scenario by leveraging multiple azure regions. Need to focus on scenarios that utilize multiple azure service. Need to handle scenarios where messages have only been partially processed when an outage occurs.

The recent fix to the broken links ( PR #32 ) has indeed fixed the broken links on the landing page. However, now the image links are broken in those pages. Example would be the Network Topology page which has lots of embedded images.

Hope we can fix this before my INTEGRATE presentation in London on 7th June - as I was hoping to show this page! :-)

The recent reorganisation (commit # 8cdd84d) has broken all the links in the table on the main ReadMe page. The docs were moved from the root of the /docs folder into the /scenario1 folder, but the links were not updated.

Personally, I'm not sure why they were moved? The concepts discussed within these excellent pages would apply to all scenarios. Burying them under a specific scenario page makes them harder to find (especially when the links on the landing page don't take you there. ;-) )

Hi,
The deployment fails at times when resource name (like key vault) exceeds the allowed number of characters. For example, if ais is used as name and landing zone needs to be deployed in Australia East, it exceeds the limit i.e. between 3-24 characters. Would be good if character limit can be enforced at the time of creation.

Thanks

Deployable demo that shows a customer a couple of implementations of a request/response aggregator

• Show how messages are fanned out and processed in parallel
• Show how to implement in LogicApps
• Show how to implement in Durable Functions
• Discuss When to use what (logicapps vs functions)
• demonstrate message tracking with property promotion, i.e. show all instances where po xyz was processed
• Self-contained demo
• independently deployable
• independent documentation
• Design as a starting point for customers, simple, easy to build upon