A service I have in WCF occasionally goes down due a problem with a COM component. While I am troubleshooting I would like to setup another host to make regular calls to this service to monitor availability.
It is slightly more complicated that a simple HTTP call though as the service is secured by SSL and WCF authentication (username / password). I'd also like to be able to parse successful calls to see if they return warning / fail states from my code.
Would you recommend any monitoring providers for this or is it beyond the simple monitoring they normally provide?
Regards
Ryan
You could enable WCF logging and auditing facilities either on the server or the client to produce a log of all traffic. Then you can analyze the results using the WCF Service Trace Viewer Tool provided in .NET Framework 3.0 and 3.5 SDK.
In your situation I would probably enable logging only at the message level. This will reduce the amount of information that ends up in the log file and will help you focus on analyzing the data that's actually being sent back and forth from the services.
Related
What are the pros and cons of hosting a WCF service in IIS versus using a Windows service?
FYI - I have googled but it's surprisingly hard to find relevant answers.
We've just implemented a big WCF service, and did it as a self-hosted windows service. The reason we did it that way was our architects wanted the extra control you get from hosting your own and taking IIS out of the equation. Basically, when you go the self-hosted route,
you process each request
you configure your own endpoints
you configure your certs
you control the exception handling
etc.
Our WCF service is industrial scale with rev proxies, load balancing and about 50 methods attached to the endpoints. And we use multiple encryption protocols depending on the types of devices connecting.
However, if I was doing a smaller WCF web service with just a single server, a single endpoint and a few method calls, I'd probably use IIS to manage the endpoint and implement the SSL letting the UI do the configuration work that would otherwise have to be done in code. It's just easier from what I've seen.
Long story short, if you host it, you control everything in code. If you're interested in a quicker delivery, I'd start with IIS.
Is there any third party tools available to Monitoring WCF Service?
I have created WCF Service with 5 Methods (OperationContracts) and Hosted in IIS.
Does there any tool available to specify how many times each method was called and what is Request and Response of the each call ? (Similar to SQL server profiler).
Use standart tracing and diagnostic tool for WCF, and work with .svclogs using SvcTraceViewer.exe application.
If you host under IIS, use AppFabric. It a great tool, when you can easily manage the settings of your services, including tracing and diagnostic. There is a nice article about using it http://www.dotnetcurry.com/ShowArticle.aspx?ID=771
The AppFabric Dashboard provides a centralized gateway to view and manage the health of WF and WCF services deployed locally or to a server farm. It exposes real-time data from the Persistence database and historic data from the Monitoring database, and it allows you to “drill down” to examine instance and WCF Call metrics in greater detail.
http://msdn.microsoft.com/en-us/library/ee677236(v=azure.10).aspx
AppFabric helps to monitor WCF service like How many time the operation is invoked, Errors/ Faults and also supports historical data.
I need to create a web service to collect data from my customer’s applications.
Those applications are programmed with different technologies and they all have one thing in common: they can consume plain SOAP Web Service.
I already have a WCF Service that could be exposed but as it was built for internal purpose only, I never had to secure it.
I did read a lot of articles on how to secure WCF service and how to consume it from a Microsoft client application. However, I'm really concern about the customer’s non-microsoft applications abilities to implement a standard WCF Service security. I must keep in mind that some of them might be stateless and unable to hold on to a session or anything that might be required by a secure WCF Service.
So here are the options I have right now.
1) Add username/password parameters to each WCF function and perform a credential check on every call. (I do have an SSL certificate... is it enough to consider this option as secured?)
2) Drop my WCF Service and create a plain SOAP Web Service with username/password parameters as mentioned in option #1 to be closer to my customer’s applications capabilities.
3) Implement standard WCF security and let the customers find a way to deal with it on their own. (The real question here: is WCF security simple enough to be implemented by any SOAP client?)
4) Change my name and move to Jamaica with my customer’s money before they find out that I’m a Web Service security noob.
5) Something else…
So what is the my best option here?
Yes, I can offer the option we use. It sounds like you want basicHttpBinding.
We have a WCF web service using basicHttpBinding and set IIS to use basic http authentication.
Therefore non-.NET clients can consume it easily (basicHttpBinding) and we can give them an Active Directory domain account that allows them access via IIS. No usernames / passwords to constantly send back and forth through the web service and it runs over HTTPS for security.
It's currently being consumed by PHP, Java and .NET clients. Yes, .NET clients can still import this as a service reference which makes thing like trapping FaultExceptions easier.
No solution is perfect for everyone but works great for our needs.
Yes, but certain configurations favour certain vendors. See the WCF Express Interop Bindings project on CodePlex:
http://wcf.codeplex.com/wikipage?title=WCF%20Express%20Interop%20Bindings
They offer settings for interop with:
Oracle WebLogic
Oracle Metro
IBM WebSphere
Apache Axis2
The Oracle Metro (previously known as SUN WSIT) stack is by far the most advanced as regards the WS-*/Oasis standards.
I have a Windows Service which performs a certain function, and then needs to send that information off to a webservice for processing. The webservice is hosted by a remote web application. I am trying to ascertain the best way to call the webservice(s) as each web application might be only 2.0, or 3.5 etc. In my windows service, I am defining each "client" in the app.config, e.g.
<Client WebServiceUrl="http://location.com/webservice.svc" Username="" Password="">
</Client>
The web application must implement two web services that are required for my windows service to run, however not sure the best way to implement the "rules" for the web application.
EDIT:
I'll try and rephrase..
The Windows Service runs every 30 seconds and obtains a list of information. The service supports multiple "clients" as shown above. When each client process is run, the data is collected and is then needed to be sent to the supporting web application.
The windows service does not know what to do with the data, it is just sending it. Each web application for a client would be in different locations, and could possibly be built in 2.0, 3.5, PHP, etc. All the windows service cares about, is that when it performs its processing for a client, it is able to send the data to the webservice location defined in the app.config of the windows service.
What I'm trying to determine is how to connect to the webservice (which I'm leaning towards WCF, however Basic or WS not sure), and what rules need to be defined for the web application in how to build the response.
If the Windows service is to support php applications etc, WSHttpBinding would not be an option, which would mean BasicHttpBinding would then work. The other thing to decide is whether or not to utilise a RESTful service or SOAP service.
Hope this makes more sense.
I'm not really clear on what you are doing.
It seems like you have 3 things: A Windows Service, and then a web service, hosted in a web app.
I think your question is, what to use, REST, ASMX, WSE or WCF, when interconnecting the Windows Service app with the remote web service.
ASMX, WSE and WCF are alternative programming models for the web service. REST is not a programming model. It is not like the other three.
ASMX and WSE will require that you use Web services and SOAP.
WCF can allow you to use Web services and SOAP, REST (XML or JSON) over HTTP, or a binary format over TCP, among other options.
Because it is flexible and current technology, I'd recommend WCF. ASMX is now termed "legacy technology" by Microsoft. Doesn't mean it won't work, but it will not get updates. (Much like WinForms versus WPF). WSE is no longer in mainstream support, as far as I know. For these reasons, I wouldn't recommend starting a new project on WSE, nor on ASMX.
WCF is more general than ASMX and can seem more complicated, for that reason. But once you make some choices and zero in on what you want (for example choose HTTP and REST, or choose binary and TCP), it's more powerful. WCF can be used as the programming model on both the client or sender (in your case, the Windows Service, I guess) and/or on the server (the web service hosted in the web app).
Using WCF on the client side does not imply you must use it on the server side, and vice versa. On the other hand, if you control the source code on both ends, I would recommend using WCF on both sides.
As for "how to implement the rules for the web app" - I don't understand what you are asking there. Maybe if you are more specific on the question there, someone will be able to help out.
Update: Based on your additional explanation, I'm going to suggest you look at the REST stuff in WCF for .NET 3.5. In PHP it's very easy to implement a REST-style service, and with WCF, the same is true for .NET. Now in your case the Windows service is the client and it is sending out a request, an update request, to various servers that reside on your customers' networks. According to REST principles, I'd make those outbound requests PUTs or POSTs, depending on the semantics of the call.
Then you could ship some example service code to your (uppercase) Clients, to get them started on building what they need to receive your outbound PUT/POST messages.
Security is a concern though. You didn't mention it at all, which is surprising. Security is not one of those things best deferred, so that you "add it on later". You should think about it early - it may affect the protocol choices you make. For example, if you need to mutually authenticate the clients and servers (the latter at your "uppercase" Clients' networks), then you may want to go with SOAP, which gives you good options on the protocol side for security. Secure Web services extensions (WS-Sec, etc) are well supported in WCF, but not sure about the status of this capability in PHP.
I'm trying to build a WCF self hosted service (eventually in a windows service) that will receive binary and text base messages from remote thick clients that have no accounts on my hosted machine. I'm trying to figure out both my binding options and security options, and in reading the patterns and practices guides, my head has completely spun around at least once.
The clients would be authenticated against a custom SQL based method, so I'd like to be able to pass that info in the initial login request and then set an authorization token of some kind. (This part of the problem is probably outside the scope of the question, but I included it in case it might make a difference.)
Any thoughts at all would be very helpfull.
Ryan
The choice of binding and security option depends on the usage of your WCF service. Is it just for your rich client or are you planning to expose it to the world as API? If it's just for your rich app, does it run on LAN or over untrusted, unreliable Internet?
With WCF you can configure the service to expose multiple endpoints with different bindings, for example both SOAP and REST. In general, I'd start with something stateless and lightweight like basicHttpBinding and webHttpBinding, passing user and password on every request. Once you have that up and running you can optimize cache authentication, provide binary endpoint etc.. only if it actually helps.
There's no need to have just one binding. Having said that if it's self hosted you're "on your own" here. I've never looked at what's involved.