Comments by Christiaan Rakowski (Top 25 by date)

The code looks to be ok, could it be a problem with the SQL connection string?
Try putting a breakpoint at the ExecuteNonQuery and see if that returns a 1 or throws an exception.

As far as I know it is always possible to hijack a session somehow, you just need to figure out how the system works and (ab)use it in the right way.
The easiest way to make it a lot harder would be to use some kind of endpoint checking, for example IP based, but that could cause problems if the user's ISP decides to change it right when the user is busy. Though it is only a very small chance that this will ever happen, it is something to keep in mind.

You could use a system like this, but you will need to properly secure the cookie to make tampering as hard as possible. There are many articles around that explain how to go about doing that.

If you go about building this system, I think it will prove an interesting project. I wish you good luck, and look forward to your results.

Ah, so you are trying to rotate the Session State servers around, not just the web servers. The question just got even more interesting :-)
I did some Googling just now and came across some interesting articles.

Microsoft Azure uses a distributed system, which they talk about here: http://msdn.microsoft.com/en-us/library/windowsazure/gg278339.aspx
I was unable to find any more details, but it might help you, or others, search for more information.

You could try implementing your own Session State Provider, this does require some work but could be a valid option. You can read more on that here: http://msdn.microsoft.com/en-us/library/ms178587(v=vs.100).aspx
It might be possible, and profitable, to build this upon a distributed Velocity cache with userid based keys. Or using a mirrored SQLServer setup.

There were also some that talked about load balancing the SQLServer, for example by DNS.

I have not tried any of these things myself yet, but they all sound quite interesting and might be worth taking a look at.

MapPath only returns a string with the full path to the folder you gave it, so starting at the drive letter all the way down. In the method Bitmap.FromFile you will need to use the imagePath variable instead of the string. The final code should be:

string imagePath = Server.MapPath("/croppedimages/");
System.Drawing.Image imgToCrop = Bitmap.FromFile(imagePath + FileUpload1.FileName);

I'm not 100% sure, but I think the IIS URL rewrite module supports removing trailing slashes. Take a look into that.

When you are in ASP the using relative directory doesn't work (most of the time). You have to use string folder = HttpContext.Current.Server.MapPath("/croppedimages/"); to get the folder, then open the image with Bitmap.FromFile(folder + filename, true);

At the end of this method you set ImagePreview.ImageUrl = "~/Handler.ashx";
I assume this handler will return the image, if so you will have to do the same in the PageLoad method of the page, just copy that 1 line over to there and it should be good.

Small extra advice: jpeg (with the e) is also a valid jpg image, adding that to your check is probably a good idea.

I use this way of working quite often for bulk background processing, but never with a real 'busy waiting' scenario, like you have right now. So I can't be a 100% certain it will work for you, and if your code is quite complex then completely rewriting/organizing can be very time consuming. However, this is this the only system I know of that can help in this scenario (other than the request timeout variable in the web.config). I see above that you already tried this, and I find it strange that it will still timeout so quickly. Other than my solution here, another thing might be to just optimize the bulk processing part.

An HTTPHandler is kind of like a stripped down ASPX page. To call this code you need to make an AJAX call to the handler. So if the handler is called WorkerHandler.ashx, you can use the button's onclientclick for this: onclientclick="$.get(' WorkerHandler.ashx' ); return false;"

Add a new Simple HTTPHandler to your project and fill it with code something like this.

using System.Web;
using System.Threading;

public class WorkHandler : IHttpHandler
{
public bool IsReusable
{
get {return true;}
}

public void ProcessRequest(HttpContext context)
{
Thread thread = new Thread(new ThreadStart(WorkThreadFunction));
thread.Start();
context.Response.Write("Thread started");
}

public void WorkThreadFunction()
{
try
{
// Call methods to do background work
}
catch (Exception ex)
{
// log errors
}
}
}

In this case you need to log into your router and set it to forward traffic on port 8001 to the local server.

It depends on the specific router brand what the option is called, I have seen the menus being called: Port Forwarding, (Special) Applications, DMZ, (Advanced) Firewall.

Check here for your specific router: http://portforward.com/english/routers/port_forwarding/routerindex.htm

In update eventhandler you don't need to write any code into the DataAdapter anymore (You already did this during the loading of the data). All you need to do is get the DataTable and use da.Update(dt), .NET will take of the rest.
See the documentation of the Update method here: http://msdn.microsoft.com/en-us/library/z1z2bkx2(v=vs.100).aspx

What I meant with all buttons attached to the correct event does not apply to your situation, you only have one button and not 3.

There is nothing wrong with the rest of the code as far as I can see.
Maybe it is because of some double used variables for the SQL code, so the query becomes invalid and does not give the correct result?
The simplest way to check this is by cleaning up your code and putting all the code in helper methods so that the code in button1_Click will be like this:

if (tabControl1.SelectedTab == tabPage1)
{
getDataForTab1();
}
else if (tabControl1.SelectedTab == tabPage2)
{
getDataForTab2();
}
else if (tabControl1.SelectedTab == tabPage3)
{
getDataForTab3();
}

If that didn't help, maybe you could upload the whole project somewhere where I can download it, and I will take a closer look at it.

Anything that high is probably unused, and thus a very good choice.

The code you posted looks fine, but it has not logic that controls it.
Please post the code that shows/hides the panel, since that is where the problem probably is.

I assume the syntax errors in there are because of copy-pasting it into the formatter, but there appear to be some strings where they shouldn't be.

Is button1 inside, or outside of the tabs. (aka is there only 1 button, or are there 3, 1 per page) If it is the second option, are those buttons attached to correct event?

I'm not sure why my answer got downvoted, but I would like to know so I can improve it. It's true it's not a solid answer, but there is no solid question to answer either.

Assuming you are using Chrome or Firefox in your development, what does the console say? Is it doing the postback, does it give a javascript error, or is it completely empty?

If it is doing the postback, to what page/handler is it going? Is this also the correct page (the one where the TextBox23_TextChanged code is?

If it does postback to the correct page, try renaming the method and see if it gives a compile error. If it does not you have duplicate names (which can happen when you call your items TextBox23)

Good solution, but it would be clearer if you used code tags and indenting in your examples.

Easy solutions are typically the best ones, +5 from me.