Apologies if this is a silly question but for the last 2 hrs. I have been reading about the difference between built-in and managed connectors for Azure logic apps and it's driving me crazy. Can someone please explain in simpler language? Initially, I thought built-in connectors are for Azure services like Azure function or table storage whereas managed are for Microsoft services like O365 & Sharepoint, but after going through the list of connectors on Microsoft documentation, its not true.
Is this tenant based? For example, built-in connector will only allow connecting to azure services in that tenant but if we need to connect to an Azure service in a different tenant, we need managed connector. Even if this is true, how can something like SMTP be an inbuilt connector?
Is the difference on the basis of authentication? The mechanism to authenticate is different for built-in / managed
I think there are some differences between the two under the hood regarding how they are hosted, but I don't know how much difference that makes for you as a user. You can read a bit more here.
However, the main difference that I have noticed and that makes me prefer the built-in connectors (when applicable) is how you can set up the authentication. Especially when trying to set up CI-CD for Logic Apps (and devloping locally in VS Code) this makes a big difference. For the managed connectors, a managed api connection is created in the portal and it's a nightmare trying to parameterize that when developing in VS Code and incorporating that to Devops-pipelines.
Built-in connectors run in the same platform where your logic apps is hosted whereas Managed connectors are hosted in public cloud. When you use built-in connectors the information configured for the connector will not be sent to public cloud for processing it will run natively in your Logic Apps. This makes it more secured. There are many differences for example authentication method, api connection file, and many more.
Related
Can anyone please point out any benefits of using Azure Mobile services vs using a plain Azure app service / clean web api? For a starter / project type for a backend mobile solution.
I have somewhat mixed feelings on why I would want to use Azure Mobile Services.
As far as I see on Azure Mobile services you have an easier way of authenticating, you can use the notifcations hub more easily
and you have the different "built-in" ways of handling data (table storage etc).
Usually you would want some custom logics, user registration and handling when users register to your backend and you would like a more solid way of handling
and storing the data not privided by the OOTB datastorage.
You might also have another preference than using the /Table/ odata-endpoint you get with it or end up doing lots of logics to make your DAO's return data in properly for the OData endpoints.
All these things; IMO makes it more difficult to make the API/backend clean when using Azure Mobile services rather than a simple Web API with OData endpoints and swagger documentet API that can be used in a mobile-app just as easy.
Implementing / handling authentication and notifications ++ in Web Api ain't that diffucult nor time consuming.
So my problem Azure Mobile services is that it tends to fine for dev / prototyping and testing, but it might get really messy really fast when developing a proper backend.
Any thoughts and reasons why one should choose one instead of the other?
Think of Azure Mobile Services as V1 and App Service/Mobile App as V2. While Microsoft hasn't announced that Mobile Services will be phased out in the near future, if you start a new project, you should definitively look at App Service.
due to the fact that many people are confused about wether to take Web API or Web App or something different. They are going to put it all under one name. The underlying technology will be the same "i think".
But now you'll have in your portal the opportunity to add mobile push notifications, or add your swagger api definitions.
So when you're goint to stick with App Services you're not going to limit yourself.
Even when you're going to take Web Api you'll get all the functions as if you would take an App Service (if i'm correct).
*Edit: I looked it up in the portal. As I said, my old Web App Projects have the same settings as Web Api projects. So you don't need to decide anymore which kind of project you're taking. You get all the benefits out of the App Service.
I need to connect to a middle tier (think Azure) between both my Windows store app and WP8 app. Windows Azure Mobile Services has been proffered as a/the solution.
But am I reading too much into the name "...MOBILES Services" (as one of the pieces is not a mobile app, but runs on desktops, laptops AND tablets)?
In my case, am I better off with Azure Cloud Services as opposed to Windows Azure Mobile Services?
Windows Azure Mobile Services is a fast, easy way to get a back-end in the cloud for your mobile apps and Windows 8 apps (it's fine if your Windows 8 apps aren't necessarily targeting mobile devices). It includes the things most commonly needed in a back-end, such as authentication, database storage, and push notifications. There's no server-side development needed here, you just request the back-end and moments later it's ready to use; however, if you do want server-side logic you can add it in the form of JavaScript scripts.
Alternatively, you could build your own back end in the cloud using a combination of Windows Azure Cloud Services such as the Compute, Storage, and SQL Database services. You have access to more features at this level, but you also are doing your own development. You can write server-side logic in C#, VB.NET, PHP, Java, Python, etc. as you prefer.
Which should you use?
If Windows Azure Mobile Services meets your needs, and you'd rather focus on your mobile app than learning anything cloud-specific, that's likely the best path for you.
If you're conversant with the Windows Azure platform, and need features different from what WAMS provides, that suggests creating your own back end with Cloud Services.
If you're not sure which way to go, I suggest experimenting first with WAMS since it is quick and painless to get started.
Azure mobile services is designed to get you up and going with storing data, push notifications and authentication whichever of those components you might need.
It abstracts the need of creating a data access layer and a web/wcf service to access it from your applications; it's simply there to boilerplate as much functionality off the bat.
This however does not mean that it is only for mobile applications, behind it all is a normal SQL Azure database and an API that you can use from any .NET based application.
I would reason that if you need more than basic CRUD operations and won't be using authentication and push notifications, I would roll my own set of APIs and DAL and use cloud services instead.
If you need a flexible schema, boilerplate data access and want to use some of the other mobile services, it would suit you quite well.
Here is a link to getting going with mobile services from a non Windows8 or WP - ASP application : link
I've recently been asked to redevelop an .Net 2.0 WinForms application with a back end SQL Server Express DB.
One of the requirements is to allow remote users access to the application, so I've been considering hosted options to avoid VPN setup. The data is not sensitive and does not fall under data protection act, so a basic security approach for the web will cover me.
I like the idea of using Azure for a few reasons, but I'm not sure if a good fit for a users base of 5 or 6 with no real scope to grow. I've never used Azure and I plan to develop using MVC and a SQL backend as this is my main skillset.
A few points in favour of Azure in my mind are:
Tight integration with the TFS preview that I'm using for this project
Easy to setup a sandpit and a live version
Easy maintenance as I expect other hosted options will require more knowledge of underlying OS
Sticking to a full Microsoft stack should hopefully make things simpler
From what I find on the Azure site the message is all about scalability, which is great if you need it.
My question is simply, do you need a large user base, or plans to grow quickly, to use azure or is it how we should be hosting apps now?
What you're asking here is the perfect case for Windows Azure Web Sites:
You get 10 web sites for free (no custom DNS, but this is perfect for your 'sandpit'/test version). The shared mode supports custom DNS and is very cheap.
Tight integration with TFS preview and GitHub
You don't need to worry about the underlying OS, you simply publish from Visual Studio or with TFS Preview.
Sticking to the Microsoft stack is the easiest solution, but other technologies work great aswell. Since you're talking about MVC I'm assuming you are considering ASP.NET MVC, which is a perfect match with Windows Azure. Take a look at the training kit for some good examples.
The day you'll need a solution which more scalable (meaning you'll have more users and more income) you can easily upgrade to a reserved instance or to a Cloud Service (Web/Worker Role).
About your question: "My question is simply, do you need a large user base, or plans to grow quickly, to use azure or is it how we should be hosting apps now?"
Windows Azure is a cloud service platform (includes PaaS as Cloud Services, IaaS as Windows Azure Virtual Machines and also Websites suggest by Sandrino above), and with cloud services you have ability to start very small and grow as much and as quickly as your user requirement is, so you can use Azure with both cases. On the other hand there are some advantages using certain offering depend on your which service you are going to use to run your application.
I think article (Section: "What Should I Use? Making a Choice") will explain the strategy about how you make a selection among various services.
This SO discussion does talks about the difference between cloud Services and Azure WebSites as well.
I'm looking into the possibility of integrating an Azure hosted .NET solution with Dynamics GP and I'm new to Dynamics. In general it seems like there are two approaches to connecting to GP: 1) web services and 2) eConnect. This article has some good background.
I would think as long as the web services are accessible, that option would work. I see that there are MSMQ and other requirements for eConnect which makes me think that would be a headache if it is even possible without something like Azure Connect. Has anyone has done this one way or the other?
Thanks
Yes you can connect to your Dynamics GP from a Windows Azure Worker role (why Worker, why not web role?) and it's all depend which route your would want to take.
Web Services method is comparatively neat & easier to access your Secure Web Services configured Dynamics GP.
On the other hand, eConnect Integration requires several other configurations so If you decided to use eConnect, I think you are better of using eConnect along with BizTalk Server/Adapter combination set into Service Bus (which is talking directly to eConnect) and your Azure application is talking to BizTalk services directly. This could be much easier to implement but you can not beat web services.
I need some help with creating a simple WebRole that uses federated authorzation/authentication with LiveId and the Access Control Service. I'm able to get it working with a local test ASP.NET application, but can't seem to find any information on the steps necessary to do this with a Web Role that can be deployed to Azure. The only information that I've found is to handle this scenario using a custom STS and the ACS or just LiveID, but nothing that demonstrates using both together.
Is there currently a limitation with Azure that prevents this? I've read some articles that seem to indicate it isn't currently possible due to the Geneva Framework not being fully implemented on Azure - can anyone confirm?
Thank you very much for any help!
You may find this resource useful - http://code.msdn.microsoft.com/wifwazpassive. It shows how to use ACS in an Azure Webrole. It does use a custom STS, not LiveID, but given that it's using Geneva framework components it should be possible to make it work with LiveID.