Відео

it depends on the usecase i think.. some usecases will have a higher direct cost in logic apps but ive also seen customers with the reverse scenario where someone wrote thousands of lines of code to do what could be done with 4 actions in a logic app. Right tool for the right job

i am not sure i still have this demo but at the time i was just doing an http call to the chatgpt API with a formatted message but i think they added some preview out of the box connectors so there is probably a simpler way to do it now a days

you can still use the cloud connectors.. just remember thou the cloud connectors dont live in the vnet with an ase thou

​@@mikestephenson8525 It seems that LA inside the ASE can use the cloud connector(using public IP) only when ASE was deployed externally. Is this correct? I'm evaluating the restrictions for LA development when ASE was created internally. It seems to have too many restrictions so far. 1. Not being able to check LA history from portal since not within the VNET, thus, have to create a jumpbox VM to just check the LA's run history? 2. Not being able to use cloud connector since those connector's are using public IP and therefore outside of VNET. Now, I'm wondering if hosting LA standard inside ASE internal may be too impractical when we have LA developers all over the world. If I add a security group to the ASE external, can still restrict access from internet and whitelist only those allowed specifically and allow public IPs being used by Azure Services. This seems to be way practical development environment. Can you please share your thoughts on this? ASE external vs internal for LA Standard development. Thanks!

thanks mate

its just a standard feature on an app service which gives you a stream of log info being published by the runtime. its often a handy place to look for any error info you might not otherwise be seeing if you get stuck and arent sure what the problem is. For logic app its sometimes handy too if you have deployed an invalid workflow that the runtime cant load up

thanks mate, good shout on the suggestion

Hi Gary, try this video it covers creating connections in terraform. You will need to workout the exact properties for each connector which is a bit of a pain to do and I found the easiest way is to just use the browser trace when creating or updating a connection then you would know the right settings for your terraform - Video link = ua-cam.com/video/IMLB3h1T6l4/v-deo.html

thank you, its great to hear your using it. would you be interested in a chat about what your doing? we have been thinking about doing more with this but didnt think there was that much interest

@@mikestephenson8525 Very happy to chat on this further, Mike.

@@mikestephenson8525 Of course! Happy to chat about how I customized and used it to manage interfaces.

@@W12ardOfOz drop me a message on linked in and we can connect - www.linkedin.com/in/michaelstephensonuk1/

thank you mate, what kind of things do you want to see us talk about?

thanks for the feedback, will look into that

thank you

hi, i have a viewsonic whiteboard TV screen that is common for teachers to use and there is some software that comes with that called my whiteboard. its pretty cool actually

Busy building an application hosted on Azure itself and this might be something that I need to think about.

@@Rolo732 thanks mate, what app are you building and hows it going?

@@mikestephenson8525 Appologies for the late reply. It is a booking system for a sport that I play. Currently running in azure as well as a production service.

thanks mate 🙂

hi, its slightly out dated but there is a sample here: github.com/michaelstephensonuk/IntegrationPlaybook-LogicApp-Standard-Testing-Example. Ill put on my backlog to do a video

Thanks for the sample Mike. Very helpful. Thank you

hi its part of my logic app ASE demo here: github.com/michaelstephensonuk/CodeSnippets/tree/main/LogicApp-ASE/Modules/AzureRm.Resource.PrivateEndpoint

thanks for your kind words ❤

That's no longer correct, recent release means you can now do c# functions

I do hope they branch to python. Even if it was lower level like what Azure Automation supports

thank you mate, really appreciate your kind words

there can be a few causes of that issue. my first thoughts would be either your connection doesnt have line of sight to the sftp. Maybe you have a firewall configured on the storage account? Next up would be checking the credentials and server name etc. Ill double check on this but I think the user name needs to include the servername. Can you connect to it in winscp or file zilla too?

@@mikestephenson8525 iam getting the same error "Failed to create connection: BadGateway" , i can use the same credentials to connect via filezilla

thanks mate, good comments.... i like your suggestion on revisiting ideas. I one of the examples in the video with 1 RG was based on a decision that was made 5 years ago and we have made it work within the wider constraints we had. Be interesting to see how we feel about that in the future

@@mikestephenson8525 Yeah. I mean that can work too, I have been in projects where I was forced to use two RGs (prod, non-prod) because the company had a policy of one subsription per department... It is good to be flexible in the choice, but then also stick to that choice. Great video series. Integration architecture resources is really har to find!

Hi Rob, thanks for your comment. The aim of the video is definitely to drive a conversation about what people are doing and what people see working and not working so we can all learn from each other. I think one thing I didnt get across too well in the video is that in the scenario I was thinking about the subscription had loads of other resource groups that have nothing to do with the integration team. That was our starting point. I totally get what your saying about people knowing where stuff is. I think if your breaking stuff out into lots of resource groups its much easier if you have dedicated subscriptions for the integration team. In the case I was thinking through we would have ended up breaking stuff out into logical groups we could have ended up with 100+ resource groups per environment for 5 environments. I think you end up with the same problem in reverse. I am hoping the take away from the video is the separation of the deployment view and application view in our thinking. In the future Id like to see the new integration environment providing the abstraction between the 2 views which helps address this question

thanks mate

Glad you enjoy it!

Glad you enjoyed it!

when you say trace, do you mean your checking the app insights trace for APIM? you should see a dependency from apim to logic apps if its not a cached response

the logic app resource itself is the same, you are just pointing to the ID of a different app service plan that it lives on. The main difference would be that for ASE you would be using a different resource to create the ase and the plan has a couple of different properties on it. My sample is here github.com/michaelstephensonuk/CodeSnippets/tree/main/LogicApp-ASE

@@mikestephenson8525 Thank you for the response.

@@mikestephenson8525 How can we differentiate the Logic App -> Networking -> Inbound Traffic Configuration -> Unmatched Rule action for Main Site Vs Advanced tool site. What are the fields we need to use under terraform to allow traffic only for a list of IPs under them(Need to Enable "Allow Direct Access" field and have to Deny "Unmatched rule action" by providing my listed IPs with priority codes only for Main Site).

Hi Gareth, you configure alert channels which have emails at the app level which would send an alert to the notification channel owner at the moment. Thats a great feature idea thou and ill mention it to Arun and see if we can do something like this. I think one prob might be a lot of times i see app reg without an owner defined

@@mikestephenson8525 Yeah this is a feature I am building into my app reg monitoring tool. We have over 400 app registrations and are tidying them up to make sure they have owners. And putting a policy in place to make sure any app registrations created going forward will have at minimum 2 owners. Its still annoying that you can't add Entry ID Groups as owners though. But yeah, then when its 30 days prior to a credential expiration it will send an email to the owners. Then again at 7 days and finally when it expires. Since app registrations are used by so many processes and departments, its too hard to try to manage completely centrally so we are putting ownership on others to make sure they keep on top of them.

We had SAP exposing OData and Soap web services. On out integration platform we use API Management to act as a proxy between our Logic Apps and the web services on SAP. This gives a centralized point for all calls from our AIS platform to SAP for HTTP which allows centralized security, management and monitoring. Note we do also use the SAP connector for some use cases but we tend to prefer HTTP by default as its a bit less complex and the teams are really comfortable with it

thanks mate, your probably right. I think the management api's are a little different but should be able to achieve the same outcome

thanks