Jan Jonas' blog

ISA Server 2006 as Reverse Web Caching Proxy: Analyze “Object source” and “Cache Info” fields

Jan Jonas — Thu, 10 May 2012 21:05:53 +0000

When publishing content from a web server that is behind a ISA 2006 firewall, the ISA can be configured as Reverse Web Caching proxy, i.e. the ISA caches responses from the web server(s) and sends the cached content to the clients instead of requesting the content from the web servers again (see “Reverse proxy” on Wikipedia). On the one hand this could reduce the load of the web severs and on the other hand enables the clients to view content even if the web servers are offline.

In this Reverse Web Caching scenario it could be interesting to analyze if a specific request was forwarded to the web server(s) and, in this case, if the web server’s response is cached by the proxy for future client requests or if the proxy delivers content from its cache.

Fortunately, the ISA server provides detailed ISA Server Web proxy log entries that include the “Object Source” and “Cache Info” fields (see http://msdn.microsoft.com/en-us/library/aa503433.aspx):

Object Source
The type of source that was used to retrieve the current object. A table of some possible values is provided in Object Source Values.
Cache Info
A number reflecting the cache status of the object, which indicates the reasons why the object was or was not cached. The number logged is the sum of the values for all the conditions that are met. A table of the possible values is provided in Cache Information Values.

The “Object source” provides information about whether the object was retrieved in the cache or not (see possible value). The “Cache Info” field is a bit mask that includes various cache information (see http://msdn.microsoft.com/en-us/library/aa503433.aspx#_isa_cache_info_values); the most interesting values are:

0×00000001: Request should not be served from the cache.
0×40000000: Response should not be cached.

Crashplan 3.8.2010: Fix “CrashPlan PRO Server failed to initialize – error 4001″ on Linux

Jan Jonas — Mon, 30 Apr 2012 20:46:28 +0000

For any reason my Linux CrashPlan PRO Server (version 3.8.2010) crashed. The CrashPlan manager web user interface shows the error

CrashPlan PRO Server failed to initialize – error 4001

After restarting the CrashPlan PRO Server the log (/var/log/proserver/com_backup42_app.log.0) shows the following error:

[...]
org.h2.jdbc.JdbcSQLException: File corrupted while reading record: "index not found 286". Possible solution: use the recovery tool [90030-131]
        at org.h2.message.DbException.getJdbcSQLException(DbException.java:316)
        at org.h2.message.DbException.get(DbException.java:167)
        at org.h2.message.DbException.get(DbException.java:144)
        at org.h2.store.PageStore.getPage(PageStore.java:502)
        at org.h2.index.PageDataIndex.getPage(PageDataIndex.java:212)
        at org.h2.index.PageDataNode.getLastKey(PageDataNode.java:221)
        at org.h2.index.PageDataNode.getLastKey(PageDataNode.java:221)
        at org.h2.index.PageDataIndex.(PageDataIndex.java:74)
        at org.h2.table.TableData.(TableData.java:78)
        at org.h2.store.PageStore.addMeta(PageStore.java:1313)
        at org.h2.store.PageStore.readMetaData(PageStore.java:1240)
        at org.h2.store.PageStore.recover(PageStore.java:1046)
        at org.h2.store.PageStore.openExisting(PageStore.java:301)
        at org.h2.store.PageStore.open(PageStore.java:259)
        at org.h2.engine.Database.getPageStore(Database.java:2014)
        at org.h2.engine.Database.open(Database.java:527)
        at org.h2.engine.Database.openDatabase(Database.java:203)
        at org.h2.engine.Database.(Database.java:198)
        at org.h2.engine.Engine.openSession(Engine.java:56)
        at org.h2.engine.Engine.openSession(Engine.java:141)
        at org.h2.engine.Engine.getSession(Engine.java:120)
        at org.h2.engine.Session.createSession(Session.java:119)
        at org.h2.engine.SessionRemote.connectEmbeddedOrServer(SessionRemote.java:235)
        at org.h2.engine.SessionRemote.createSession(SessionRemote.java:213)
        at org.h2.jdbc.JdbcConnection.(JdbcConnection.java:110)
        at org.h2.jdbc.JdbcConnection.(JdbcConnection.java:94)
        at org.h2.Driver.connect(Driver.java:58)
        at java.sql.DriverManager.getConnection(DriverManager.java:582)
        at java.sql.DriverManager.getConnection(DriverManager.java:185)
        at com.backup42.server.DbVerifierForH2.getConnection(DbVerifierForH2.java:92)
        at com.backup42.server.DbVerifierForH2.verifyDb(DbVerifierForH2.java:33)
        at com.backup42.main.CPServer.main(CPServer.java:361)
[...]

Fortunately the CrashPlan PRO server frequently creates dumps of its internal database and I was able to restore the server’s database from such a dump with CrashPlan’s restore_database.sh script (see CrashPlan Pro support).

The location of the database dumps could be found in /var/log/proserver/app.log:

[...]
DBDumpDir   = /var/opt/proserver/dumps
[...]

To restore the database execute the restore_database.sh and pass a dump file as argument:

/opt/proserver/bin/restore_database.sh /var/opt/proserver/dumps/proserver-db_[...].sql.gz

This script shuts down the CrashPlan PRO server before restoring the database and restarts the server after the restore is completed.

After the CrashPlan server is up and running again the CrashPlan manager web user interface shows the message

Your PRO Server is running from a newly imported database. The old database has been preserved within: /opt/proserver/db.bak
This message will be displayed until that backup directory is manually moved to a safe location or deleted.

WordPress & jQuery: Fix JavaScript error “$ is not a function”

Jan Jonas — Sun, 15 Apr 2012 17:06:14 +0000

When using jQuery in WordPress you might be surprised by a

$ is not a function

JavaScript error.

For example, in a default WordPress 3.3.1 installation, the following JavaScript code leads to this error:

$(document).ready(function() {
  console.log($("body"));
});

The reason is, that WordPress by default uses jQuery’s “no conflict mode” to avoid compatibility problems with other JavaScript libraries that might could also use the $ variable. In detail, the jQuery script source included in WordPress activates this mode by calling jQuery.noConflict() (for WordPress 3.3.1 see line 5 in http://core.trac.wordpress.org/browser/tags/3.3.1/wp-includes/js/jquery/jquery.js#L5).

To fix the error replace jQuery’s $ shortcut by jQuery. Optionally, you can add a $ parameter your handler functions to define the $ shortcut inside the functions:

jQuery(document).ready(function($) {
  console.log($("body"));
});

ASP.NET How-To: Create User Control library by compiling User Controls into .dll files

Jan Jonas — Fri, 06 Apr 2012 21:49:44 +0000

This blog post shows how to create a User Control library (a .dll file containing User Controls). To reuse the contained User Controls in other ASP.NET web projects you just need to add a reference to this .dll file and register a tag prefix (as shown at the end of this post).

First, we need the following base class for the User Controls that should be included in the library:

public abstract class UserControlBase : System.Web.UI.UserControl
{
  protected override void FrameworkInitialize()
  {
    base.FrameworkInitialize();
    string content;
 
    var resourceName = GetType().FullName + ".ascx";
    var stream = Assembly.GetExecutingAssembly().GetManifestResourceStream(resourceName);
    if (stream == null)
    {
      throw new InvalidOperationException(
        string.Format("Loading resource '{0}' failed", resourceName)
      );
    }
 
    using (var reader = new StreamReader(stream))
    content = reader.ReadToEnd();
 
    var userControl = Page.ParseControl(content);
    if (userControl == null)
    {
      throw new InvalidOperationException(
        string.Format("Parsing user control in resource '{0}' failed", resourceName)
      );
    }
    Controls.Add(userControl);
 
    WireControls(userControl);
  }
 
  protected abstract void WireControls(System.Web.UI.Control userControl);
}

Next we remove the User Control’s designer file (the .ascx.designer.cs file) and remove all attributes except the Language attribute from the tag in the User Control’s .ascx file.

The exemplary User Control in this blog post contains a Literal control that shows the current time; the .ascx file looks as follows:

<%@ Control Language="C#" %>
 
<asp:Literal ID="Output" runat="server">asp:Literal>

In the User Control’s code behind file we use the above defined UserControlBase class as parent class:

public partial class MyUserControl : UserControlBase
{
  private Literal output;
 
  protected void Page_Load(object sender, EventArgs e)
  {
    output.Text = DateTime.Now.ToString();
  }
 
  protected override void WireControls(System.Web.UI.Control userControl)
  {
    output = (Literal)userControl.FindControl("output");
  }
}

Please note: Since there is no designer file we need to create private fields for the controls and assign them manually by overriding the WireControls method of our UserControlBase parent class.

Before building the project to create the .dll library file we need to set the the build action for the .ascx to “Embedded Resource“.

You can download the Visual Studio 2010 project containing all the source code of the User Control library project here.

To include the User Control in any other ASP.NET project just add a reference to the (library) .dll file and add the User Control as follows on any ASP.NET page:

[...]
<%@ Register Assembly="UserControl-Dll-File" TagPrefix="uc" Namespace="UserControl_Dll_File" %>
<uc:MyUserControl ID="WebUserControl1" runat="server">uc:MyUserControl>
[...]

IIS 7.5 Application Pools: Understanding default identity “ApplicationPoolIdentity”

Jan Jonas — Fri, 06 Apr 2012 13:27:03 +0000

In former versions of IIS the default identity for application pools was “NetworkService“. Using this option, the application pool runs under the built-in and low-privileged Windows identity “NETWORKSERVICE“. The disadvantage of this (old default) option is that multiple services using the Windows identity “NETWORKSERVICE” are not isolated from each other, i.e. services could theoretically read, write or execute files of the other services.

In IIS 7 the concept of application pool identities (“ApplicationPoolIdentity“) was introduced and became the default identity for application pools in IIS 7.5. Using this identity Windows runs each application pool under a automatically created virtual account. Please note, that this virtual account is not listed in Windows User Management Console.

To grant these virtual accounts permissions to files or directories, you need to add the (virtual) identities to a file’s or directories Access Control List (ACL). Therefore use the permission tab of the file or folder and select the account as follows:

Set the location to you local machine
Set the object’s name to “IIS AppPool\“

Summarized application pool identities allow system administrator to easily isolate multiple application pools from each other without manually managing accounts and passwords.

ASP.NET MVC 3 & MongoDB: CRUD operations in NoSQL database using FluentMongo

Jan Jonas — Tue, 03 Apr 2012 19:55:12 +0000

This blog post shows how to use the open-source NoSQL database MongoDB in an ASP.NET MVC 3 web application. The showcase web application uses FluentMongo that adds LINQ-support for the 10gen MongoDB C# driver. Summarized the application demonstrates how to perform the basic CRUD (create, read, update and delete) database operations.

Before running the example web application, please make sure that a MongoDB server is up and running on your machine (see MongoDB Quickstart for Windows).

We start with an empty ASP.NET MVC 3 web application and use the NuGet package manager to install the following packages:

FluentMongo (installs also the dependent MongoDB C# driver)
MvcContrib (used in this example to easily create HTML grids)

After including the required NuGet packages we add our User model (Models/User.cs):

public class User
{
  [BsonId]
  public ObjectId ObjectId;
 
  [ScaffoldColumn(false)]
  public String Id {
    get { return ObjectId.ToString(); }
    set { ObjectId = new ObjectId(value); }
  }
 
  [DisplayName("First Name")]
  [Required]
  public String FirstName { get; set; }
 
  [DisplayName("Last Name")]
  [Required]
  public String LastName { get; set; }
}

Since ASP.NET MVC 3 does not support binding MongoDB’s ObjectId (used by the database to identify the records) the User model contains two id properties:

ObjectId: The (BSON) id used by MongoDB to identify the records in the database
Id: The (string) id used by the ASP.NET MVC 3 application in GET and POST request data to identify the records

The getter and setter in the Id attribute “synchronizes” the string and the BSON value. In a real world project one should solve this problem in a more generic way, e.g. by implementing a custom ModelBinder or by separating database entities from MVC models.

Next, we implement the HomeController (Controllers/HomeController.cs) as follows:

public class HomeController : Controller
{
  private MongoCollection<User> _users;
 
  public HomeController()
  {
    MongoDatabase database = MongoDatabase.Create("mongodb://localhost:27017/TestMvc3MongoDB");
    _users = database.GetCollection<User>("Users");
  }
 
  [HttpGet]
  public ActionResult Delete(String id)
  {
    var user = _users.AsQueryable().FirstOrDefault(e => e.Id == id);
    if (user != null) _users.Remove(Query.EQ("_id", new ObjectId(id)));
    return RedirectToAction("Index");
  }
 
  [HttpGet]
  public ActionResult Edit(String id)
  {
    var user = _users.AsQueryable().FirstOrDefault(e => e.Id == id) ?? new User();
    return View(user);
  }
 
  [HttpPost]
  public ActionResult Edit(User user)
  {
    if (ModelState.IsValid)
    {
      _users.Save(user);
      return RedirectToAction("Index");
    }
    return View(user);
  }
 
  public ActionResult Index()
  {
    return View(_users.AsQueryable());
  }
}

In a last step we add the view that shows all records (Views/Home/Index.cshtml):

@Html.Grid(Model).Columns(column => {
    column.For(x => x.Id).Named("Person ID");
    column.For(x => x.FirstName).Named("First Name");
    column.For(x => x.LastName).Named("Last Name");
    column.For(x => Html.ActionLink("Edit", "Edit", new { id = x.Id }));
    column.For(x => Html.ActionLink("Delete", "Delete", new { id = x.Id }));
})
 
@Html.ActionLink("Create Record", "Edit")

and the view that shows the form to edit and create records (Views/Home/Edit.cshtml):

@using (Html.BeginForm())
{
    @Html.Hidden("Id", Model.Id);
    @Html.EditorForModel(Model)
    
}

You can download the Visual Studio 2010 project containing all the source code here.

ASP.NET MVC 3: Validate complex Types (Objects & Lists) in Ajax Form using jQuery and JSON on Client Side and Server Side

Jan Jonas — Fri, 16 Mar 2012 22:57:20 +0000

This blog post shows how to validate models containing complex types such as Objects and Lists in ASP.NET MVC 3.

We start with the source code described in one of my previous posts “ASP.NET MVC 3: Using JSON result for jQuery Ajax Forms validation” (you can download the Visual Studio 2010 project here). Summarized this project contains one Ajax form that supports (unobtrusive) client side and server side validation and uses jQuery and JSON to transfer the form data to the server.

In a first step we modify the properties of the model (Models/ValidationModel.cs) and add some complex types:

public class ValidUserNameAttribue : ValidationAttribute
{
  public override bool IsValid(object value)
  {
    return (value != null &amp;&amp; value.ToString() == "Bob");
  }
}
 
public class User
{
  [Required]
  [StringLength(8, MinimumLength = 3)]
  [ValidUserNameAttribue(ErrorMessage = "User name != 'Bob'")]
  [Display(Name = "User name")]
  public string UserName { get; set; }
 
  [Required]
  [StringLength(8, MinimumLength = 3)]
  [Display(Name = "Display name")]
  public string DisplayName { get; set; }
}
 
public class ValidationModel
{
  public User User { get; set; }
 
  public List Users { get; set; }
}

In a second step we modify the form (Views\Home\Partial\_Form.cshtml) to add input element for the new model properties:

@model MVC3_Ajax_Form_jQuery_Validation.Models.ValidationModel
 
@DateTime.Now: Partial/_Form.cshtml rendered
<hr/>
 
@using (Html.BeginForm("Form", "Home"))
{
  <h1><em>Userem> objecth1>
 
  <p>
    @Html.LabelFor(m => m.User.UserName):
    @Html.EditorFor(m => m.User.UserName)
    @Html.ValidationMessageFor(m => m.User.UserName)
  p>
 
  <p>
    @Html.LabelFor(m => m.User.DisplayName):
    @Html.EditorFor(m => m.User.DisplayName)
    @Html.ValidationMessageFor(m => m.User.DisplayName)
  p>
 
  <h1>List of <em>Userem> objectsh1>
 
  for (var i = 0; i <= 1; i++)
  {
    <h2>User @ih2>
 
    <p>
      @Html.LabelFor(m => m.Users[i].UserName):
      @Html.EditorFor(m => m.Users[i].UserName)
      @Html.ValidationMessageFor(m => m.Users[i].UserName)
    p>
 
    <p>
      @Html.LabelFor(m => m.Users[i].DisplayName):
      @Html.EditorFor(m => m.Users[i].DisplayName)
      @Html.ValidationMessageFor(m => m.Users[i].DisplayName)
    p>
  }
 
  <input type="submit" value="Submit" />
}

In a last step we adapt the “success-view” (Views\Home\Partial\_Success.cshtml) that is shown after the data have been successfully validated on the server side:

@model MVC3_Ajax_Form_jQuery_Validation.Models.ValidationModel
 
<p><strong>Model is valid :)strong>p>
 
<p>
  Model.User.Username: '@Model.User.UserName'<br />
  Model.User.DisplayName: '@Model.User.DisplayName'<br />
  Model.Users[0].Username: '@Model.Users[0].UserName'<br />
  Model.Users[0].DisplayName: '@Model.Users[0].DisplayName'<br />
  Model.Users[1].Username: '@Model.Users[1].UserName'<br />
  Model.Users[1].DisplayName: '@Model.Users[1].DisplayName'
p>

As you can see in the source code above, there is no magic; model binding and validation of complex objects and lists work out of the box in ASP.NET MVC 3.

You can download the Visual Studio 2010 project containing all the source code here.

Java: Test if String matches wildcard expression

Jan Jonas — Tue, 06 Mar 2012 09:00:17 +0000

The following Java method tests if a string matches a wildcard expression (supporting ? for exactly one character or * for an arbitrary number of characters):

/**
 * @param text Text to test
 * @param pattern (Wildcard) pattern to test
 * @return True if the text matches the wildcard pattern
 */
public static boolean match(String text, String pattern)
{
  return text.matches(pattern.replace("?", ".?").replace("*", ".*?"));
}

Here are some tests for this method:

Assert.assertFalse(match("HELLO WORLD", "ELLO"));
Assert.assertFalse(match("HELLO WORLD", "HELLO"));
Assert.assertFalse(match("HELLO WORLD", "*HELLO"));
Assert.assertFalse(match("HELLO WORLD", "HELLO WORLD2"));
Assert.assertFalse(match("HELLO WORLD", "HELLO WORL"));
Assert.assertFalse(match("HELLO WORLD", "hello world"));
 
Assert.assertTrue(match("HELLO WORLD", "*ELLO*"));
Assert.assertTrue(match("HELLO WORLD", "HELLO*"));
Assert.assertTrue(match("HELLO WORLD", "*LLO*"));
Assert.assertTrue(match("HELLO WORLD", "HELLO WORLD"));

IIS 7.5 on Windows 7: How to fix “HTTP Error 401.3 – Unauthorized”

Jan Jonas — Thu, 02 Feb 2012 21:23:19 +0000

If you try to run a web application on IIS 7.5 and Windows 7 and you get the following error

HTTP Error 401.3 – Unauthorized
You do not have permission to view this directory or page because of the access control list (ACL) configuration or encryption settings for this resource on the Web server.

it is most likely because of missing or wrong NTFS permissions. In detail, the identity that is used by the IIS web-server for “Anonymous Authentication” needs read permission on the web application’s physical folder. By default this identity is set to “IUSR“. To solve the HTTP Error 401.3 you could either

Grant read permissions on the web application’s physical folder for the IUSR user or
change the identity that is used for “Anonymous Authentication” to a user that has the required read permissions. In order to change the identity use the IIS Manager, select the website, select “Authentication” (in section IIS), right click on “Anonymous Authentication”, run edit action and define a “Specific user”.

ASP.NET: Disable browser and proxy caching for aspx resources

Jan Jonas — Sat, 28 Jan 2012 15:45:50 +0000

The following lines of code can be put into a .aspx file to prevent the resource from being cached on clients or proxies:

<%
Response.Cache.SetCacheability(HttpCacheability.NoCache);
Response.Cache.SetNoStore();
%>

To prevent caching, these lines of code change the header of the server’s HTTP response by setting the following values:

Cache-Control: no-cache, no-store
Expires: -1
Pragma: no-cache

You can read more about other possible caching options that can be passed to SetCacheability() here.