How To Configure A WCF Service That Calls Other Services Behind A Load Balancer
Sep 3, 2010
I haven't deployed behind a load balancer before. My customer has a WCF service built and tested on servers using a service model configuration that is relatively straightforward. It provides a service to return an image of a map for another application. To get the map, it calls other services.
The service was built in Visual Studio 2010 targeting the 3.5 framework. The customer is using IIS 7.5 and an F5 load balancer. When moving to the production server, the Web.config was changed to add the load balancer behavior and specify the endpoint to show the physical and logical address of the service:
[Code].....
There was a problem when the service was deployed to a server behind the load balancer. When I try to call the service from WCFStorm or WebServiceStudio I get the message "The provided URI scheme 'https' is invalid; expected 'http'.
The endpoints for the service itself look right to me. However in the development and testing versions of the config, the client section uses http instead of http, while on the production servers it uses the load balancer's https address. This seems like it should be obvious, but we're missing it.
I have a couple of servers and a load balancer. I want to show a server name which is currently serving the page.
I am using HttpContext.Current.Request.ServerVariables["SERVER_NAME"] and HttpContext.Current.Request.ServerVariables["LOCAL_ADDR"] but is shows the same data for all servers (load balancer information is shown but not the information about exact server name).
I'm creating web application behind load balancer. To this moment I configured it to store session in database but I'm not sure how should I handle session expiration. The problem is not sessions are not removed from database but Session_End event because I have to call some web service method in it.
Assuming Session_End is called when expires the thing I'm afraid of is situation when session is created on one server but finished on another. In this case I'm afraid Session_End on first server will be executed prematurely and I will call web service too early. What would you suggest in this situation?
Edit:
I remember some time ago reading about Sql Agent reacting to session end event and then performing custom code. Can anybody confirm that this solution is possible?
We are using Two servers for our application managed by a load balancer. One of which server contains Windows server 2003 & other one contains Windows server 2008. When a request is served by windows 2003 server and its subsequent goes to the other one (2008 server). Users gets log out. Are we need same Operating systems on both systems because( when request served by 2008, and subsequent on 2003 is served without any issue).
I have published a website and installed it on my local machine and it works fine, but when i install it on the both servers who controlled from the load balancer i got an error like the WebResource.axd not found. how can i resolve this problem? i have checked the IIS for axd mapping, the "Check that file exists" is unchecked for both servers.
We host a SaaS application on 4 windows 2008 servers loadbalancer via a LVS. We use infragistics and Ajax Control Toolkit on several pages, one page represents a list of people using a Web Datagrid, each time a person is clicked in the list, an Ultra Web Tab component is updated at the botom of the page with detailed information on that person. After a certain amount of clics on different people (number of clics can be between 1 and 5) Scriptresource.axd craches with the following error :
[Code]....
I don't get this error when I run my application on one server using the server's private IP adress. I tried to synchronize the dates of the dlls System.Web and System.Web.Extensions on each server, I still get the same error.
Our server operations team has asked the web development team (ASP.NET) to provide a URL in our application, which the load balancer can ping to perform health checks.
What should be executed on this page? I think we should attempt a database connection to ensure connectivity between the web and database.
One issue with ASP.NET apps is that they periodically reload themselves, causing a long delay and possibly timeout for users who hit the site during that time.
This may not be a problem for small websites, but it can represent significant downtime for high-traffic sites, if users happen to get routed to node in the web farm that is restarting.
Can load balancers somehow "know" if an ASP.NET application domain on a specific server is restarting? Then, they can route traffic around this server until the application has completed restarting.
Currently, I have my load balancer ping a simple .aspx page on the site. If there is a delay or the page fails to load, the host is taken out of rotation. Is it possible to do a more targeted health check, perhaps at the IIS level rather than ASP.NET level?
we have an web site access on PRODTEST Environment . We are facing an issue site works fine when we try to access the site using the Individual Server names (with Ip address).
I have four servers 2 App Server(behind firewall) and 2 webservers and each server is accessing through Load Balancer.
App Server is a modules for Data Access layer and webserver is for Presentation layer.
My Issue is My Load Balancer has Secure certificate installed but certificate is not installed on servers and I want to make cookie as secure in site , as per my understanding " cookie should be set as secure only for SSL site other ways new session will get create every time" so should I install SSL certificate on webserver to make my cookie as secure or will it work properly even if only Load Balance has SSL.
I have the following web farm setup in production server. Browser --> HTTPS --> Load Balancer --> HTTP --> webserver node F5 Load balancer handles off box SSL termination. It implies that SSL resides on F5 load balancer. Problem Statement: Ajax calls do not go through unless "Access data sources across domains" option is enabled in IE security settings.
I have the similar setup in staging server except F5 Load balancer. The ASP.Net application makes perfect AJAX calls on both HTTP and HTTPS. However, the staging server web farm use windows NLB and SSL resides on individual web server nodes
I have 3 servers where 1 of them serves as a load balancing server. In my ASPX page, I want to add a HTML comment to show the IP address or even host name of the server selected by the load balancer. I tried looking through IIS Server variables and tried using SERVER_NAME but that just returns the domain URL.
I'm writing some merge functionality in C# asp.NET MVC2. I am also using using Linq2SQL.
I have a block of code which calls two services, MessageService and UserService. These both in term call their appropriate repositories and make the amendments to the db. Each repository declares it's own instance of the repository so I'm thinking this will escalate the following code to DTC . The code is called from the AccountService, is this going to work at this level? And also is it bad practise to declare the DataContext at the top of every repository or should I pass the object around somehow? //Run the merge try { using (TransactionScope scope = new TransactionScope()) [code]....
My website uses forms auth and I am trying to create a load balancer (hardware) rule which will keep track of sessions based on the aspxauth cookie. Is it safe assume that the value of the cookie is unique?
I have an AJAX web application.ASP pages are accessing various Web Services that are part of the application using standard Microsoft .NET AJAX extension APIs.Now, I have to call some of the web services from the Windows app.It works when I add web reference to a project on my local development machine.
I have been moving my application to be more AJAX-based. Currently I have a web service call where I use jQuery to gather some textbox (string) and checkbox (boolean) results and pass it to a web service. Everything works really well, the response time is super quick. The only downside that I can see to this is
1) you need to use jQuery, which will add to the user's download time - this doesn't really affect my application too much because I am using jQuery throughout the site
2) users can see the code and potentially try to hack the web service because they know where it is
I am most concerned with #2. I've been reading a lot about .NET web service calls, using IAsyncResult,
http://msdn.microsoft.com/en-us/library/2e08f6yc.aspx, and I am beginning to understand the 4 methods available in this article. However, is there extra bloat using this way compared to the jQuery web service call?If there isn't that much bloat or lag time using this method, I'd like to convert my application. Which one of these 4 methods is best in a scenario where I am just returning a small string value? The 4 methods in the article are to wait for EndInvoke itself, which from what I read is not recommended because this may block the first thread; and this should not be used on a service that affects the UI.Create a WaitHandle to wait for the 'OK' before EndInvoke is called Poll the IAsyncResult for .IsCompleted, using sleep() and a Do While until IAsyncResult.IsCompleted is true, then call EndInvoke Pass a delegate function to BeginInvoke to be called once the operation is complete.
my web site i have two databases, ASPNETDB.MDF and a self created one (database.mdf). (The don't contain a lot of data yet)But i need a relationship (foreign key) between a table in ASPNETDB.MDF and a table in database.mdf.So i guess i need to merge both databases first, would you just extend the "ASPNETDB.MDF" with the tables from "database.mdf" ?or better configure "database.mdf" for the asp.net Applicatoin Services and then delete "ASPNETDB.MDF"
I'm working with the MicrosoftAjax.js libraries to make my webservice calls via javascript, right now all my services are returning basic types, like string, int, bool. can i also have my services return things like DataSets, DataTables, or custom Classes that I have built to my javascript side?
I have a SOAP web service and I'm trying to figure how to save/log the last 10 requests for each user. Each user is required to send their user/pass in each request, so it's easy to know who the request originated from. With these last 10 requests saved, my goal is to develop some sort of page that will allow them to log-in with their credentials and view the raw request, the actual SOAP message, http header information, and anything relevant that I can think of.
The point is to allow people to troubleshoot their own connection issues instead of having to contact me each time they can't connect, have trouble formatting their request, etc....
My first thought was to store all this information in memory in a hashtable or something, but that may have scalability issues when we have hundreds/thousands of users hitting the web service.
We could use our database to store these requests. Instead of hitting the database each time, I may need to create some "buffer" mechanism that will only update the database after the buffer gets to a certain number of requests. Is there an existing library or mechanism that will do this?
We can't store these requests on the file system on the machine hosting the web service. Since these requests can potentially contain sensitive information, it's a business decision that I'll need to work around. Or maybe there's a better way to achieve what I'm trying to do?
Are web service calls synchronous or asynchronous by default? How is synchronicity determined, by the service or by the client?
I have code similar to the following:
try { string result = MakeWebServiceCall_1(); // this is a third party webservice [code]....
In the above, SetStatus is writing to the same tables that the third party web services read from. If I change the status before both web service calls have completed, it's going to make a big mess and I'm going to get fired. How do I know/ensure that the webservice calls are synchronous?
I've got a web app running with purely web services like these:
Code: <WebMethod()> _ <ScriptMethod(ResponseFormat:=ResponseFormat.Json, UseHttpGet:=False)> _ Public Function EditService(ByVal toddtype As String _ , ByVal fromddtype As String _
[Code] ....
And I want to log each web service call. I don't want to bother my database with this log - I would prefer to have the web app log to a text file on the web server itself.
How would I open and access a log file like this from a bunch of web services? Does web app still have application startup like events???
I'm trying to make API calls to a service provider. I've never tackled JSON requests before but I want to give it a go - as they seem less verbose than XML. Could someone be kind enough to give me a simple example here?
My confusion stems from some libraries I've come across. Do I need to add any libraries to my project? Or is it just a matter of creating a string that complies with JSON syntax and send it via a WebRequest?
We have a page that makes a request to a 3'rd party web service. When under heavy load this response time extends significantly, however the 3'rd party reports back that there times for processing remains constant. There timings show that from the time they receive a request to the time they send it back is always around 1.5-2.0 seconds. Now we are experiencing wait times of over 20 seconds. My understanding of ASP.NET is that each request will run on a IIS thread pool thread and make the request to the 3'rd party service return and process. So I don't really understand what could be blocking on our end. Is there something I am missing?? Is there a threshold limit beyond IIS that is blocking?
If I am missing something a physical book recommendation that covers this subject would also be a very welcome addition to any answer.
I have an existing asp.net web application I am redesigning to use a service architecture. I have the beginnings of an WCF service which I am able to call and perform functions with no problems. As far as updating data, it all makes sense. For example, I have a button that says Submit Order, it sends the data to the service, which does the processing.
Here's my concern: If I have an ASP.NET page that shows me a list of orders (View Orders page), and at the top I have a bunch of drop down lists for order types, and other search criteria which is populated by querying different tables from the database (lookup tables, etc). I am hoping to eventually completely decouple the web application from the DB, and use data contracts to pass information between the BLL, the SOA, and the web app. With that said, how can I reduce the # of WCF calls needed to load my "View Orders" page? I would need to make 1 call get the list of orders, and 1 call for each drop down list, etc because those are populated by individual functions in my BLL.
Is it good architecture to create a web service method that returns back a specialized data contract that consists of everything you would need to display a View Orders page, in 1 shot? Something like this pseudocode:
public class ViewOrderPageDTO { public OrderDTO[] Orders { get; set; }[code].....
Or is it better practice in the page_load event to make 5 or 6 or even 15 individual calls to the SOA to get the data needed to load the page? Therefore, bypassing the need for specialized wcf methods or DTO's that conglomerate other DTO?
This is quite straight forward. I'm calling a web-service (.asmx, with session enabled) from a c# application. I want each call to be with the same session key as the previous one (as opposed to creating a new session key each time).
