How to make your server-side javascript code private in IIS

You learned that you can use Server-Side Javascript mixed in our ASP Classic VBScript application.

Next, you want to create tons of javascript files, and mix them with your old ASP. You can name them “.asp”, or “.inc”, like you did with your old VBScript files.  But, what about using intellisense in a standard editor from these extensions?

The best way to use intellisense in your favorite editor, like Visual Studio Code or anything else, is to name these files with the .js extension.

But, there’s a security issue.  If you name these files like that, they should be returned to your client in plain text if called from a web browser.  You don’t want your server-side code to be exposed through IIS.

What you need to do, is just put all these files in a specific folder, in which you will disable all IIS handlers, by using the following steps.

  1. In your application root, create a folder called private.
  2. In that folder, create a web.config file, and add this content:
<system.webServer>
  <handlers>
    <clear />
  </handlers>
</system.webServer>

You can now add all your server-side JS in that folder, and they will private to your ASP code.

If you call that js file in the browser, you will get a 404:

From your ASP page in the root, include them like this:

<script runat="server" language="javascript" src="private/code.js"></script>

That’s all! Happy Coding.

How to mix Classic ASP and Server-Side Javascript

Sometimes I need to fix some old Classic ASP applications or add small new features to them.

Because I do Javascript or C# every day for new apps, it hurts to get back to VBScript for these old apps!

That’s why I try to do these using Server-Side Javascript instead.  But wait, how can I mix NodeJS and VBScript?  The answer is, you can’t.  But, there was Server-Side JS before NodeJS exists!

Now, look at this.

You have a ASP Classic site like this:

<body>
<% dim product : product = "apple"
response.write product %>
...

And you need to create a new function to get the product name.

But, you don’t want to do it in VBScript, because you prefer JavaScript.

All you need to do is to create an “include” file, in pure javascript, like this:

includeJS.inc:

function getProduct(){
	var product = "apple";
	return product;
}

And, add it to your existing ASP file:

<script runat="server" language="javascript" src="includeJS.inc" ></script>

Finally, you can fix your existing code like this:

<% dim product : product = "apple"
response.write getProduct %>

Yes, the VB code can call JS function directly.
Also, your JS code can also use already existing VBScript functions and variables.

IIS can run Server-Side Javascript since 1997

When I read articles about NodeJS, they almost all consider NodeJS like if Ryan Dahl invented Server-Side Javascript.

But, did you know that, in IIS 3.0 on Windows NT 4.0, it was possible to run server-side javascript?

On May 15, 1997, Microsoft released the Service Pack 3 of Windows NT4, introducing Active Server Page technology (ASP).

Classic ASP was built to run Active Script.  VBScript AND JScript are the first Active Script languages developped, and were included in all versions of Windows since 95.  And, that “Jscript”, can be used to run server-side javascript.

Wait, JScript is not JavaScript?  Yes it is. Just another name.  FireFox 1.0 run something named JavaScript 1.5, and IE6 run JScript 5.0  And, both were, in fact, ECMAScript 3.0. (more info)

That said, it means that if you still need to work with Classic ASP application, to maintain old software that can’t be rewritten, you can write some JavaScript to do that.  And, it is totally compatible with existing VBScript app!

Now, look at this sample of a 100% server-side javascript in an classic ASP file.

I’ll show you, in another article, how you can add some javascript in your existing Classic ASP / VBscript application.

Useful url rewrite rules for IIS

These are some rules I use in IIS to make all my web sites works as expected.

What I need to automate in all site:

  • Remove prefix “www.”
  • Redirect “http” to “https”, as http will soon disappear, thanks to LetsEncrypt.

Also, for very simple static sites, I want to make all pages respond to the same fixed static page, and get no 404.

You can do that without any code, when using the IIS Rewrite extension.  You first need to install the extension to your existing IIS configuration. (link)

This is an example of my IIS configuration:

  • Site 1: “https://foxontherock.com”
  • Site 2: “https://fredericmalenfant.com”
  • Site 3: default (ip binding)

Bindings strategy

The first strategy can be to add these rules to all your sites, one by one.
Exemple for site 1, these are the 4 bindings configuration:

  • http://foxontherock.com
  • http://www.foxontherock.com
  • https://foxontherock.com
  • https://www.foxontherock.com

If you set your IIS bindings like that, you need to add the rewrite rules to all your sites, one by one.

But there’s another simple strategy that I prefer, by using a “default” site on IIS.
On the first and second sites, I only set the “final” binding, e.g. foxontherock.com and fredericmalenfant.com, both on port 443 without www.

On the “default” site, I configure the bindings to respond to all other requests on my public IP address.

That way, I only need to add the redirect rules on the default site, as the redirection will be handled by the redirected site after.

IIS Rewrite Rules

First rule: Remove www prefix.

<rule name="Remove www" stopProcessing="true">
 <match url="(.*)" ignoreCase="true" />
 <conditions logicalGrouping="MatchAll">
  <add input="{HTTP_HOST}" pattern="^www\.(.+)$" />
 </conditions>
 <action type="Redirect" url="https://{C:1}/{R:0}" redirectType="Permanent" />
</rule>

Second rule: redirect http to https

<rule name="Redirect to https" stopProcessing="true">
 <match url="(.*)" />
 <conditions>
  <add input="{HTTPS}" pattern="off" ignoreCase="true" />
 </conditions>
 <action type="Redirect" url="https://{HTTP_HOST}/{R:0}" redirectType="Permanent" />
</rule>

The third rule can be set to all sites, and is used to get a fixed page always responding, without 404.

<rule name="Default page" stopProcessing="true">
 <match url=".*" />
 <conditions>
  <add input="{REQUEST_FILENAME}" matchType="IsFile" negate="true" />
  <add input="{REQUEST_FILENAME}" matchType="IsDirectory" negate="true" />
 </conditions>
 <action type="Rewrite" url="index.html" />
</rule>

Finally, you can add the “canonical” meta on that page to make robots index only 1 page, not all possible url.

(index.html)
<head>
...
<link rel="canonical" href="https://foxontherock.com/" />
</head>

Disable ssl3 for more security

SSL3 is over.  Some servers keep it active because they need to serve pages to IE6/XP users.  But, these days are over, as most of the updated to, at least IE8, and they have all the last updates allowing to connect using TLS 1.1 or 1.2.

SSL3 can also be enabled by default on older Windows Servers, like 2008 R2, even if you installed all Windows Updates.

Now, it’s time to disable SSL3 completely.

If your SSL3 is active, you may get one of these warnings from Qualys SSL Test:

  • This server is vulnerable to the POODLE attack. If possible, disable SSL 3 to mitigate. Grade capped to C.
  • This server uses SSL 3, which is obsolete and insecure. Grade capped to B.

It’s easy to disable it from a Windows registry, as described here.

Solution

Create a file, called DisableSSL3.reg, and copy that content on it:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\SecurityProviders\SCHANNEL\Protocols\SSL 3.0\Server]
"Enabled"=dword:00000000

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\SecurityProviders\SCHANNEL\Protocols\SSL 2.0\Server]
"Enabled"=dword:00000000

Save it, double-click on it, and accept.

Finally, reboot, and all SSL3 server services are disabled on your Windows server.

How to get “A” Rating on Qualys SSL Labs Test

That blog is hosted on an Amazon EC2 Instance, running Windows 2012 R2 Server.  And our SSL certificated is provided by Let’s Encrypt.

Starting from that default configuration, we ran the SSL test, and we got a B note.  We wanted to get the “A” Rating, and these are the 2 major warning we had to solve.

  • “This server supports weak Diffie-Hellman (DH) key exchange parameters”
  • “This server accepts RC4 cipher, but only with older protocols”

Solution

We were able to fix these issues with some simple registry tweaks that we describe in these articles

Then, after we ran these steps, we now have our A Grade!

Now, if you want an even better grade, you can continue to solve these little warnings that the SSLLabs test can give you.

How to solve RC4 warning on Qualys SSLLabs Test

In a previous article, I talked about how you can solve the Diffie-Hellman warning on Qualys SSLLabs test, by applying a registry configuration.

Now, we’ll talk about another common warning that most AWS EC2 customer can get.  By default, we got that security issue from SSLLabs:

This server accepts RC4 cipher, but only with older protocols. Grade capped to B.

Solution

Microsoft proposes a solution for disabling the 3 weak RC4 cipher suites in that article.
You need to create 1 new registry entry.  Create an empty text file called rc4fix.reg, and paste that content to it:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\SecurityProviders\SCHANNEL\Ciphers\RC4 128/128]
"Enabled"=dword:00000000

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\SecurityProviders\SCHANNEL\Ciphers\RC4 40/128]
"Enabled"=dword:00000000

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\SecurityProviders\SCHANNEL\Ciphers\RC4 56/128]
"Enabled"=dword:00000000

Then, double-click on it to apply these settings, an reboot.

Finally, run your SSL test again, that warning disappeared.

How to solve Diffie-Hellman warning on Qualys SSLLabs Test

In a previous article, I talked about how you can get a better note on Qualys SSLLabs test, by configuring CAA DNS entry.

Today, we’ll talk about another warning most of us must resolve to get the “A” Rating.

On our AWS EC2 Windows 2012 R2 server, by default, we got that security issue from SSLLabs:

This server supports weak Diffie-Hellman (DH) key exchange parameters. Grade capped to B.

Solution

That is caused by the Diffie-Hellman protocol accepted at 1024 bits.  The fix proposed by Microsoft (article) is to still accept that protocole, but only at 2048+ bits.

You need to create 1 new registry entry.  Create an empty file called df.reg, and paste that content to it:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\SecurityProviders\SCHANNEL\KeyExchangeAlgorithms\Diffie-Hellman]
"ServerMinKeyBitLength"=dword:00000800

Then, double-click on it to apply these settings, an reboot.

Run your SSL test again, that warning disappeared.

letsencrypt simple all sites

After using letsencrypt-win-simple (now win-acme) for my iis sites, I had some troubles, and I provide you the solutions I applied to fix them.

First, I had a lot of sites / domain names to register, and it was long to do that from the interactive app.  So I tried to do that from command prompt.  The documentation is not clear on how to do that, so I did several try-mistake.

And, finally, that’s what I found:

the first time you will add the “plugin” mode, you will get that error:
unable to find validation plugin
Because the “recommended” validation method from the ui is not the same in the command prompt, and you need to provide it, using the argument:
–validation selfhosting

These are 3 command line methods I tried.
The first allow you to get 1 single certificate for all your sites.  But, warning, if you request certificates info for any domain name registered, you will always see the first one registered.  So, I did not use that method.  But, you can try it.

letsencrypt.exe --plugin iissites --validation selfhosting --siteid 1,2,3,4,...

Warning, do not include sites id that are invalid or inexisting.

The second method allow you to create 1 certificate per site ID (that is not a domain name, you can have multiples domain name binding on 1 single site).

letsencrypt.exe --plugin iissite --validation selfhosting --siteid 1

And now, my preferred one, that I put on a batch file with all my existing sites:

letsencrypt.exe --plugin iisbinding --validation selfhosting --manualhost yoursite1.com
letsencrypt.exe --plugin iisbinding --validation selfhosting --manualhost www.yoursite1.com 
letsencrypt.exe --plugin iisbinding --validation selfhosting --manualhost yoursite2.com
letsencrypt.exe --plugin iisbinding --validation selfhosting --manualhost www.yoursite2.com 
...

You execute that only once.

After, you can run the renewal process once per month, or every day as you wish, but the renewal process will only renew certificates that were generated more than 55 days ago.

letsencrypt.exe --renew

If you experience some issues when updating, with locked certificates files, I suggest to run “iisreset” before renewing, as it always helps me to clear all these renewal errors.

How to track your server-server call in Fiddler

Telerik Fiddler is the web developer’s best friend.

It does a wonderful job of telling you why it worked, or not.  Its ability to edit and replay some requests is wonderful.  That, with Advanced REST Client, both help you accomplish your day to day job.

It tracks everything.
Almost.
Sometimes, you need to track server-to-server requests.  Example, if your app calls your back-end api, and then, your server needs to call Mandrill API to send an email to your customer.

By default, Fiddler is not tracking that request, that goes out from your server-side app. But, there’s a solution.

In fact, Fiddler track all web traffic generated by “you”, the user currently logged, and who started the fiddler app.

Now, knowing that, all you need to do, is make your IIS Pool run at your name.

By default, IIS uses “ApplicationPoolIdentity” user, it’s a kind of virtual user generated by the web engine, to run the app.  Each pool have its own username.

But, you can change it to something else.

Open IIS.  Go to the Application Pools section. Choose your app pool, then open Advanced Settings / Identity.  Click the dots, choose Custom account, enter your credentials, and you’re done!

Now, your outgoing server-side calls from your back-end can be tracked in Fiddler!