lundi 6 juillet 2015

MVC3 App in Visual Studio 2010 Painfully Slow During and After Authentication

I am having issues with an MVC3 application. Some details:

This application has two "modes" basically. For users who are not logged in, there are a couple dropdowns they can select, and then post the form and get back results. For logged in users, the controls are the same, but the dropdowns are populated with extra data. It's not a significant amount of data to cause this type of slowdown in page rendering.

When I fire up the application in Visual Studio (2010), the initial page loads very quickly, like all our apps, and the form elements on the page post and work just fine. The issues begin as soon as I attempt to login. Once I click on the Login page submit button, the login process takes upwards of 15 minutes. In addition, after the user is logged in, every page submit after that (and AJAX calls) take about 15 minutes as well.

This is the only application that has this issue. All our other apps use the same authentication code.

I've tried deleting all breakpoints, no change in the problem.

I've tried starting the application without debugging; no change in the problem.

I've deleted the .SUO file. No change.

I've downloaded the code onto to separate machines from our HG source, and both machines experience the same problem.

what's Weird: A college can download the code from our HG source and has NO ISSUE running it. He logs in fine and every page thereafter runs quick.

I've installed GLIMPSE to get an idea of the execution time.

This is for the initial home page load without logging in:

enter image description here

And this is for the Login page:

enter image description here

So you can see this is a big problem.

I've tried adding breakpoints to determine what might be causing the issue, but it hasn't revealed anything. The authentication code itself runs very fast. It's the page time that is taking forever. Just adding a breakpoint into the main controller for the page and the authentication method show that it takes forever to reach that breakpoint.

I'm at my wits end. It's clear from our testing this is only happening to me. I don't have any idea what it could be. It's not a breakpoint because the code EVENTUALLY does complete.

Has anyone ever experienced anything like this before?

Aucun commentaire:

Enregistrer un commentaire