adammatusiak

Disconnecting users from MSSQL database

noreply@blogger.com (Adam Matusiak) — Mon, 21 Feb 2011 16:31:00 +0000

When trying to restore database from backup often error saying about connected users is displayed.
Script below kills all connections to given database.
Just replace XYZ with required database name.
Remember that all connections will be dropped without notice.

USE master

GO

DECLARE @userId varchar(10) 

DECLARE cur CURSOR READ_ONLY      
FOR 
SELECT request_session_id
FROM master.sys.dm_tran_locks     
WHERE resource_type = 'DATABASE'     
AND resource_database_id = db_id('XYZ')    
GROUP BY request_session_id

OPEN cur     

FETCH NEXT FROM cur INTO @userId      
WHILE (@@fetch_status <> -1)      
BEGIN      
 IF (@@fetch_status <> -2)      
 BEGIN      
  PRINT 'Killing connection ' + @userId      
  EXEC ('KILL ' + @userId)      
 END      
 FETCH NEXT FROM cur INTO @userId      
END      
      
CLOSE cur      
DEALLOCATE cur

Deploying Sitecore CMS to Windows Azure using VM Role

noreply@blogger.com (Adam Matusiak) — Thu, 10 Feb 2011 10:47:00 +0000

Sitecore CMS can be deployed to Microsoft Azure using Sitecore CMS Azure Edition, but as a proof of concept I gave it a try to upload it using Virtual Machine (VM) Role.

Using exactly same procedure as with EPiServer you can deploy Sitecore CMS. Only change required was updating aspnet_Membership_GetNumberOfUsersOnline procedure as in original it's not compatible with Azure SQL. So after database migration additional query with code placed below is required:

SET ANSI_NULLS ON
SET QUOTED_IDENTIFIER OFF
IF NOT EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[aspnet_Membership_GetNumberOfUsersOnline]') AND type in (N'P', N'PC'))
BEGIN
EXEC dbo.sp_executesql @statement = N'CREATE PROCEDURE dbo.aspnet_Membership_GetNumberOfUsersOnline
    @ApplicationName            nvarchar(256),
    @MinutesSinceLastInActive   int,
    @CurrentTimeUtc             datetime
AS
BEGIN
    DECLARE @DateActive datetime
    SELECT  @DateActive = DATEADD(minute,  -(@MinutesSinceLastInActive), @CurrentTimeUtc)

    DECLARE @NumOnline int
    SELECT  @NumOnline = COUNT(*)
    FROM    dbo.aspnet_Users u WITH (NOLOCK),
            dbo.aspnet_Applications a WITH (NOLOCK),
            dbo.aspnet_Membership m WITH (NOLOCK)
    WHERE   u.ApplicationId = a.ApplicationId                  AND
            LastActivityDate > @DateActive                     AND
            a.LoweredApplicationName = LOWER(@ApplicationName) AND
            u.UserId = m.UserId
    RETURN(@NumOnline)
END' 
END

Deploying EPiServer CMS to Windows Azure using VM Role

noreply@blogger.com (Adam Matusiak) — Wed, 09 Feb 2011 12:41:00 +0000

Intro

This post describes running EPiServer CMS version 6 on the Windows Azure using Virtual Machine (VM) Role Beta Program. The main goal was to take developer Hyper-V machine and deploy it to the cloud as staging environment.

Prerequisites

Windows Azure account
VM Role Beta Program enabled (it took over a month to complete)
basic EPiServer knowledge
SQL Server 2008 R2
Microsoft SQL Server 2008 Feature Pack is required by database migration tool (SQL Azure Migration Wizard)
Visual Studio 2010 or Microsoft Visual Web Developer 2010 Express
Windows Azure SDK and VS Cloud Service installed

Creating VHD

First Windows Server 2008 R2 is required - I've used Windows Server 2008 R2 Enterprise (Full Installation). Additionally it should be installed on a single partition, which is not the default behaviour. There are few posts on how to remove this additional partition but sysprep (described later) fails after those operations - it can be done only during system installation.

select Custom (Advanced) instead of Upgrade
on partition selection screen press Shift + F10 to open console
run diskpart

list disk
select disk 0
create partition primary
exit

close console
refresh
select created partition

After installation go to Server Manager:

add Role: Web Server (IIS)
add Feature: .NET Framework 3.5.1 (inside .NET Framework 3.5.1 Features node)

You may need to switch off Windows Update, but it should be inactive by default.
You may backup Virtual Machine now to reuse it later.

Installing Application

Now it's time to install EPiServer or any other application you'd like to deploy to the could. I've used empty EPiServer version 6 with Public Templates. Only web.config change was to switch to SQL Role and Membership providers.

Setting up database

Local database needs to be migrated to Microsoft SQL Azure Database - there's quite simple way to do it using SQL Azure Migration Wizard

Because this tool doesn't clean tables previously filled with data this script can be used to remove all tables if another run is required for any reason.

You should modify connectionStrings.config to connect to Azure SQL instance and test if application works.

Finalizing VHD

It's highly recommended to backup Virtual Machine now.
Further steps caused it hang during start up until in cloud.

Mount C:\Program Files\Windows Azure SDK\v1.3\iso\wavmroleic.iso as DVD drive and install WaIntegrationComponents-x64.msi. Provide it with valid Administrator password and let it install all the drivers it needs. After reboot run %windir%\system32\sysprep\sysprep.exe enter options as on image below and let Virtual Machine shutdown - now VHD image is ready for deployment.

If you ever start this Virtual Machine again (which is not obvious to succeed) you need to launch sysprep once more.

Creating Management Certificate

Start Visual Studio
create project of type C# / Cloud / Windows Azure Project.
click Publish
Select 'Deploy your Windows Azure project to Windows Azure'
Credentials - select Add... and name it
Click 'Copy the full path' to get Public key path and upload it to Azure
Copy subscription ID back to Management Authentication dialog and click OK - it will be verified.
You can now forget about this project :).

Uploading VHD

Open a Windows Azure SDK Command Prompt as an administrator from Start | All Programs | Windows Azure SDK v1.x.
Run command below replacing XXXXX and YYYYY with your Subscription Id and Certificate Thumbprint respectively. You should of course provide path to VHD and destination location to reflect your needs.

csupload Add-VMImage -Connection "SubscriptionId=XXXXX; CertificateThumbprint=YYYYY" -Description "Base image Windows Server 2008 R2" -LiteralPath "D:\VM\AzureTest\Virtual Hard Disks\Clean win 2008 r2.vhd" -Name baseimage.vhd -Location "North Europe"

Anywhere locations are not supported (i.e. Anywhere Europe and Anywhere US).

Statuses visible in Management Portal while uploading:

pending means uploading
committed means ready

Creating the Service Model (Visual Studio part)

Start VS2010
create Windows Azure Project without any roles
right-click Roles > Add > New Virtual Machine Role
in properties of new role > Virtual Hard Disk tab > select account and uploaded VHD
in Endpoints tab add:

name: HttpIn
type: Input
protocol: Http
public port: 80
private port: 80

go to Publish
Configure Remote Desktop (save certificate to file WITH the private key)
select Create Service Package Only
click OK

Creating the Service Model (Management Portal part)

New Hosted Service - name it, give it URL prefix, choose region and 'Do not deploy'
Add Certificate (select one created in previous step)
New Production Deployment - name it and select two files created by VS (.cspkg and .cscfg)
wait...
wait more...
wait even more...
you can connect to IIS using URL on the right
Connect on the ribbon downloads RDP file

Further works

All file system modification on Virtual Machine will be lost when it's reimaged from uploaded VHD. It may happen because of machine failure or restart. At the moment only data stored in Azure SQL is safe - every change to VPP will be lost. There are two options - to store VPP data in database or in Cloud Storage. There's a post on this by Daniel Berg but I haven't tried yet.

There's also licensing issue as EPiServer license is linked either to IP or MAC address. Unfortunately MAC address changes as well during deployment. At the moment I've generated new demo license using MAC address obtained from Virtual Machine but it will be lost. Not sure what can be done about it.

WCF service consumed by jOuery on IIS7

noreply@blogger.com (Adam Matusiak) — Mon, 31 May 2010 15:22:00 +0000

For this example assume you need to return Info structure (that consists of two strings: Title and Body) that can be queried using it's Id.

First you have to define data contract:

namespace Example
{
    [DataContract]
    public class Info
    {
        [DataMember]
        public string Title;

        [DataMember]
        public string Body;
    }
}

Then interface of the service - it also defines request and response format as JSON (which can be easily consumed in JavaScript).

namespace Example
{
    [ServiceContract(Namespace = "")]
    public interface IGetInfoService
    {
        [OperationContract]
        [WebGet(ResponseFormat = WebMessageFormat.Json, RequestFormat = WebMessageFormat.Json)]
        Info GetInfo(int id);
    }
}

Service logic just to provide full example :)

namespace Example
{
    [AspNetCompatibilityRequirements(RequirementsMode = AspNetCompatibilityRequirementsMode.Allowed)]    
    public class GetInfoService: IGetInfoService
    {       
        public Info GetInfo(int id)
        {
            return
                new Info 
                    {
                        Title = id.ToString(),
                        Body = "Lorem ipsum"
                    });
        }
    }
}

WCF service requires also web.config section:

<system.serviceModel>
        <services>                        
                <service behaviorConfiguration="GetInfoServiceBehavior" name="GetInfoService">
                        <endpoint address="" behaviorConfiguration="GetInfoEndpointBehavior" binding="webHttpBinding" contract="Example.IGetInfoService">                                       
                                </endpoint>
                                <endpoint address="mex" binding="mexHttpBinding" contract="IMetadataExchange" />
                </service>
        </services>
               
        <behaviors>
                <serviceBehaviors>
                        <behavior name="GetInfoServiceBehavior">
                                <serviceMetadata httpGetEnabled="true" />
                                <serviceDebug includeExceptionDetailInFaults="false" />                                                                    
                        </behavior>
                </serviceBehaviors>
                <endpointBehaviors>
                        <behavior name="GetInfoEndpointBehavior">
                                <enableWebScript />
                        </behavior>
                </endpointBehaviors>
        </behaviors>
        <serviceHostingEnvironment aspNetCompatibilityEnabled="true" />
</system.serviceModel>

Consume service using jQuery $.ajax call.
Params is object that handles all parameters and msg in success function

var params = new Object;
params.id = 10;

$.ajax({
 type: "GET",
 url: "/Services/GetTheFactsService.svc/GetNextFact",
 data: params,
 dataType: "json",
 success: function (msg) {
  $(".info h3").html(msg.d.Title);
  $(".info p").html(msg.d.Body);
 },
 error: function (msg) {
  alert(msg);
 }
});

If ISS reports 404 or 500 error results you probably need to register WCF:

c:
cd "C:\Windows\Microsoft.NET\Framework\v3.0\Windows Communication Foundation"
servicemodelreg -i

Command above registers *.svc handlers for host so if you have other defined for site you have to remove them or go to Handler Mappings section and use Revert To Inherited action.

Lucene large queries efficiency

noreply@blogger.com (Adam Matusiak) — Fri, 23 Apr 2010 16:21:00 +0000

Reading large data sets from Lucene even after adding document cache can still be slow due to algorithm that loads only 100 results and queries again each time loaded amount is smaller than required. Documentation advises to use HitCollector to speed up but it's hard to find an example. Big disadvantage of this approach is no sorting support when HitCollector is used.
Below you can find example of LuceneHitCollector storing items in LuceneHitCollectorItems and using LuceneSorter.
LuceneSorter is based on internal Lucene sort algorithm and therefore uses it's dictionaries built while directory is created. It implements only subset of Lucene capabilities:

sorting by rank (when empty string is passed to Sort method)
sorting by string field
reverse order sorting

Last code example shows how to use HitCollector. It's as easy as using Lucene itself and again much faster.

public class LuceneHitCollectorItem
{
    public float Score;
    public int DocId;
    public Document Doc;
}

public class LuceneHitCollector : HitCollector
{        
    private readonly IndexReader reader;
    private List<LuceneHitCollectorItem> docs = new List<LuceneHitCollectorItem>();

    public LuceneHitCollector(IndexReader reader)
    {
        this.reader = reader;
    }

    public override void Collect(int doc, float score)
    {
        Document document = reader.Document(doc);
        docs.Add(
            new LuceneHitCollectorItem
                {
                    DocId = doc,
                    Doc = document,
                    Score = score
                }
            );
    }

    public void Sort(string field, bool reverse)
    {
        LuceneSorter sorter = new LuceneSorter(reader, field, reverse);
        sorter.Sort(docs);
    }

    public int Count
    {
        get
        {
            return docs.Count;
        }
    }

    public LuceneHitCollectorItem Doc(int n)
    {
        return docs[n];
    }
}

public class LuceneSorter
{
    private string field;
    private bool reverse;
    private StringIndex index;

    public LuceneSorter(IndexReader reader, string field, bool reverse)
    {
        if (field != "")
        {
            String fieldName = String.Intern(field);
            index = FieldCache_Fields.DEFAULT.GetStringIndex(reader, fieldName);
        }
        this.field = field;
        this.reverse = reverse;
    }

    public void Sort(List<LuceneHitCollectorItem> docs)
    {
        if (String.IsNullOrEmpty(field))
        {
            docs.Sort(CompareScore);
        }
        else
        {
            docs.Sort(CompareField);
        }
    }

    private int CompareField(LuceneHitCollectorItem a, LuceneHitCollectorItem b)
    {
        int result = index.order[a.DocId] - index.order[b.DocId];
        if (result == 0)
        {
            result = a.DocId - b.DocId;
        }
        return (reverse) ? -result : result;
    }

    private int CompareScore(LuceneHitCollectorItem a, LuceneHitCollectorItem b)
    {
        int result = a.DocId - b.DocId;
        if (a.Score != b.Score)
        {
            result = (a.Score > b.Score) ? 1 : -1;
        }
        return (reverse) ? -result : result;
    }
}

LuceneHitCollector collector = new LuceneHitCollector(reader);
searcher.Search(query, collector);
collector.Sort(field, reverse);
for (int index = 0; index < hits.Count; index++)
{
    LuceneHitCollectorItem item = hits.Doc(index);
    Document doc = item.Doc;
    double rank = item.Score;
    //process document
}

Speeding up Lucene Index

noreply@blogger.com (Adam Matusiak) — Fri, 23 Apr 2010 16:04:00 +0000

Lucene is indexing and search technology, as well as spellchecking, hit highlighting and advanced analysis/tokenization capabilities. .NET version is automatic Java Lucene port.
Solution works perfect in most cases but recently I've found it too slow. Directory can be stored in file system or in memory. While search indexes are cached in memory document contents are loaded each time from directory. I haven't checked Java implementation but .NET seems to parse file very slow. File system version requires additionally lots of disc access, which is probably well cached by operating system but not efficent for large indexes (in my case file size is over 500MB). Memory directory implementation tends to leak which is unacceptable.
Solution below demonstrates how to enable document caching in Lucene.
Cache needs to be added to FieldsReader class. Code below shows only changes that needs to be made:

Cache directory needs to be added.
Cache needs to be clear at the end of Close() method.
Document is returned from cache if already loaded at the beginning of Doc() method.
Document is added to cache after read in Doc() method.

This change has big memory cost but speeds up large queries even up to 10 times.
Further optimization can be found here.

public sealed class FieldsReader
{
    private SortedList<int,Document> cache=new SortedList<int, Document>();
//---------
    public void Close()
    {
//---------
        cache.Clear();
    }
//---------
    public Document Doc(int n)
    {
        if (cache.ContainsKey(n))
        {
            return cache[n];
        }
//---------
        if (!cache.ContainsKey(n))
        {
            cache.Add(n, doc);
        }
        return doc;
    }
//---------

Listing your Twitter messages by RSS

noreply@blogger.com (Adam Matusiak) — Fri, 26 Mar 2010 10:24:00 +0000

Frequently it's useful to list your Twitter messages on web other website or process them in some other way. There are billions of plugins that will allow you to do it but if you need to download them manually it's best to use RSS. You don't need to use API so you don't need to register. Only disadvantage is that you need user ID not his name. You can obtain it from user page - there is a RSS link in right hand column - number in that link is user ID that needs to be passed to the code.

Code below consists of three classes that will get twitts for you. There is also caching so remember that new messages will be fetched only every UPDATE_EVERY_MINUTES minutes.

Single message:

public class TwitterCacheItem
    {
        public string Title {get;set;}
        public string Link {get;set;}
        public string Description {get;set;}
    }

All of them:

public class TwitterCacheChannel : TwitterCacheItem
    {
        public TwitterCacheChannel()
        {
            Items = new List();
            Loaded = false;
        }

        public DateTime NextUpdateTime { get; set; }
        public bool Loaded { get; set; }
        public List Items { get; private set; }
    }

And finally the code:

public class TwitterCache
    {
        private const int UPDATE_EVERY_MINUTES = 60;

        private static SortedList cache = new SortedList();
        private static object cacheLock = new object();


        public static TwitterCacheChannel GetTwitts(int id)
        {
            TwitterCacheChannel result = null;
            lock (cacheLock)
            {
                if (cache.ContainsKey(id))
                {
                    result = cache[id];
                }
            }
            if ((result==null)||
                (result.NextUpdateTime<DateTime.Now))
            {
                result = FetchTwitts(id);
            }
            return result;
        }

        private static string FetchTwittsUrl(int id)
        {
            return String.Format("http://twitter.com/statuses/user_timeline/{0}.rss", id);
        }

        private static TwitterCacheChannel FetchTwitts(int id)
        {
            TwitterCacheChannel result = new TwitterCacheChannel();
            if (id > 0)
            {
                try
                {
                    string url = FetchTwittsUrl(id);
                    string rssText = DownloadWebPage(url);
                    XmlDocument rss = new XmlDocument();
                    rss.LoadXml(rssText);
                    XmlElement channel = rss["rss"]["channel"];
                    result.Title = channel["title"].InnerText;
                    result.Link = channel["link"].InnerText;
                    result.Description = channel["description"].InnerText;
                    foreach (XmlElement node in channel.SelectNodes("item"))
                    {
                        TwitterCacheItem item = new TwitterCacheItem();
                        item.Title = node["title"].InnerText;
                        item.Link = node["link"].InnerText;
                        item.Description = node["description"].InnerText;
                        result.Items.Add(item);
                    }
                    result.Loaded = true;
                }
                catch
                {
                }
            }
            result.NextUpdateTime = DateTime.Now.AddMinutes(UPDATE_EVERY_MINUTES);
            lock (cacheLock)
            {
                if (cache.ContainsKey(id))
                {
                    cache[id] = result;
                }
                else
                {
                    cache.Add(id, result);
                }
            }
            return result;
        }

        private static string DownloadWebPage(string url)
        {
            HttpWebRequest request = (HttpWebRequest)HttpWebRequest.Create(url);
            request.UserAgent = ".NET Framework/2.0";
            request.Referer = "http://www.example.com/";
            WebResponse response = request.GetResponse();
            Stream webStream = response.GetResponseStream();
            StreamReader reader = new StreamReader(webStream);
            string result = reader.ReadToEnd();
            reader.Close();
            webStream.Close();
            response.Close();
            return result;
        }
    }

Debugging Deadlocks in ASP .NET

noreply@blogger.com (Adam Matusiak) — Thu, 18 Mar 2010 12:37:00 +0000

Deadlock detection

Deadlock is reported in Event Viewer

Section: Application
Source: W3SVC-WP
ID: 2262
Desc: ISAPI 'C:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\aspnet_isapi.dll' reported itself as unhealthy for the following reason: 'Deadlock detected'.

and

Section: System
Source: W3SVC
ID: 1013
Desc: A process serving application pool 'xxx' exceeded time limits during shut down. The process id was 'xxx'.

Server configuration

After detecting deadlock IIS kills old and creates new process.
To prevent it enable the Orphan Worker Process feature.

WARNING: Orphaned Worker Processes must be killed manually to free resources and memory.

cd \Inetpub\adminscripts
adsutil.vbs SET W3SVC/AppPools/DefaultAppPool/OrphanWorkerProcess TRUE

where DefaultAppPool should be replaced with proper Application Pool name
List of Application Pools together with Process Ids can be obtained by IISAPP tool (type iisapp in cmd).

You can use automatic dump collector but it seems that better get it by hand.

Dump collecting

Best way to get dump is using DebugDiag tool.
Go to Processes tab right-click on selected process and select Create Full Userdump.

Quick dump analysis

Using DebugDiag you can quickly get info from created dump.
Right-click on .dmp file and select Analyze Crash/Hang Issue it will show you

Stack trace of each Thread
All current requests
But it won't load symbols for your application.
Better way is to use WINDBG described below.

Detailed dump analysis

Procedure is described here.

1. Install Debugging Tools.
2. Run C:\Program Files\Debugging Tools for Windows\windbg.exe.
3. Open Crash Dump (form File menu)
4. Important commands:

.symfix binpath
.reload
.loadby sos mscorwks
.cordll -ve -u -l
!threads
~* e !clrstack

Last line should display all threads with call stacks and symbols resolved.

System.Threading.Monitor.Enter means that thread waits on lock (...) {...} statement.

What can possibly go wrong?

Wrong CLR version

!threads command says:

CLRDLL: C:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\mscordacwks.dll:2.0.50727.3053 f:0
doesn't match desired version 2.0.50727.3082 f:0
CLRDLL: Unable to find mscordacwks_x86_x86_2.0.50727.3082.dll by mscorwks search
CLRDLL: Unable to find 'mscordacwks_x86_x86_2.0.50727.3082.dll' on the path
CLRDLL: Unable to find mscorwks.dll by search
CLRDLL: ERROR: Unable to load DLL mscordacwks_x86_x86_2.0.50727.3082.dll, Win32 error 0n2
Failed to load data access DLL, 0x80004005

Solution

Install all Service Packs from here.

.NET Framework 2.0 Service Pack 2
.NET Framework 3.0 Service Pack 2
.NET Framework 3.5 Service Pack 1

EpiServer - Itera.MultiProperty examples

noreply@blogger.com (Adam Matusiak) — Wed, 17 Mar 2010 14:49:00 +0000

MultiProperty by Itera is very useful EpiServer extension.

Two basic ways of using MultiProperty are described on a blog by Anders Hattestad.

Basics

Most common way to use MultiProperty is to create set of fields and then create second property that allows user to create list of those items.

Let's assume user needs to provide list of ExampleItem.

public class ExampleItem
    {
        public string Field { get; set; }
        public string Regex { get; set; }        
        public string Default { get; set; }
    }

First you have to create property that lets user enter single item.

public class PropertyExampleField : PropertySingleBase
    {
        protected const string PROPERTY_NAME_FIELD = "Field";
        protected const string PROPERTY_NAME_REGEX = "Regex";
        protected const string PROPERTY_NAME_DEFAULT = "Default";

        PropertyDataCollection _innerPropertyCollection;
        protected override PropertyDataCollection InnerPropertyCollection
        {
            get
            {
                if (_innerPropertyCollection == null)
                {
                     _innerPropertyCollection =
                        new PropertyDataCollection
                            {
                                {PROPERTY_NAME_FIELD, new PropertyString()},
                                {PROPERTY_NAME_REGEX, new PropertyString()},
                                {PROPERTY_NAME_DEFAULT, new PropertyString()},
                            };
                }
                return _innerPropertyCollection;
            }
            set { _innerPropertyCollection = value; }
        }

        public string Field
        {
            get { return (PropertyCollection[PROPERTY_NAME_FIELD].Value as string) ?? ""; }
            set { PropertyCollection[PROPERTY_NAME_FIELD].Value = value; }
        }

        public string Regex
        {
            get { return PropertyCollection[PROPERTY_NAME_REGEX].Value as string; }
            set { PropertyCollection[PROPERTY_NAME_REGEX].Value = value; }
        }    

        public string Default
        {
            get { return PropertyCollection[PROPERTY_NAME_DEFAULT].Value as string; }
            set { PropertyCollection[PROPERTY_NAME_DEFAULT].Value = value; }
        }

        public ExampleItem ExampleItem
        {
            get
            {
                return new ExampleItem
                           {
                               Field = Field,
                               Regex = Regex,
                               Default = Default
                           };
            }
        }
    }

Then you have to create property that lets user create list of items.

[Serializable]
    [PageDefinitionTypePlugIn(DisplayName = "List of ExampleField")]
    public class ExampleFieldList : PropertyMultiBase
    {
        PropertyDataCollection baseProperties;
        public override PropertyDataCollection BaseProperties
        {
            get
            {
                if (baseProperties == null)
                {
                    baseProperties = new PropertyDataCollection { { "ExampleField", new PropertyExampleField() } };
                }
                return baseProperties;
            }
        }

        public List ExampleItems
        {
            get
            {
                List result = new List();
                foreach (PropertyExampleField item in PropertyCollection)
                {
                    result.Add(item.ExampleItem);
                }
                return result;
            }
        }

    }

Advanced

You can sort fields on save - it will help users to deal with long lists. You can remove empty fields - but you have to remember to let user have one empty so he can add new items. To achieve it override SerializeToXmlNode method in ExampleFieldList class.

public override System.Xml.XmlNode SerializeToXmlNode(PropertyDataCollection propCollection, PropertyDataCollection basePropertys, Dictionary newGuids)
        {
            List fields = new List();
            bool firstEmpty = true;
            foreach (PropertyData data in propCollection)
            {
                PropertyExampleField field = data as PropertyExampleField;
                if (field == null)
                {
                    continue;
                }
                if (field.Field != "")
                {
                    fields.Add(field);
                }
                else
                {
                    if (firstEmpty)
                    {
                        fields.Add(field);
                        firstEmpty = false;
                    }
                }
            }
            fields.Sort((a, b) => a.Field.CompareTo(b.Field));
            PropertyDataCollection newPropCollection = new PropertyDataCollection();
            foreach (PropertyExampleField data in fields)
            {
                newPropCollection.Add(data);
            }
            return base.SerializeToXmlNode(newPropCollection, basePropertys, newGuids);
        }

You an also force some mandatory items to be placed on the list each time user enters edit mode. To achieve it override ParseToSelf method in ExampleFieldList class.

Example below adds 3 fields named aaa, bbb and ccc.
Even if user removes any of them they will be recreated next time anyone enters edit mode.

public override void ParseToSelf(string xml, PropertyDataCollection propCollection, PropertyDataCollection basePropertys)
        {
            base.ParseToSelf(xml, propCollection, basePropertys);            
            string fieldList = "aaa,bbb,ccc";
            foreach (string part in fieldList.Split(",".ToCharArray()))
            {
                string fieldName = part.Trim();
                if (fieldName == "")
                {
                    continue;
                }
                bool found = false;
                foreach (PropertyData data in propCollection)
                {
                    PropertyExampleField field = data as PropertyExampleField;
                    if (field == null)
                    {
                        continue;
                    }
                    if (field.ExampleItem.Field == fieldName)
                    {
                        found = true;
                        break;
                    }
                }
                if (!found)
                {
                    propCollection.Add(
                        new PropertyExampleField 
                                           {
                                               Name = String.Format("ExampleField_{0}", Guid.NewGuid()),
                                               Field = fieldName,
                                               Regex = "",
                                               Default = ""
                                           });
                }
            }
        }

Binding Fluent NHibernate session to WCF call

noreply@blogger.com (Adam Matusiak) — Tue, 16 Mar 2010 11:12:00 +0000

First you should read previous post about WCF Session.

You have to modify the example provided in mentioned post. While binding the Endpoint Extension to the Call Context you need to open and close session.

public class ExampleEndpointContextInitializer : ICallContextInitializer
    {
        public object BeforeInvoke(InstanceContext instanceContext, IClientChannel channel, Message message)
        {
            bool found = false;
            foreach (IExtension extension in OperationContext.Current.Extensions)
            {
                if (extension is ExampleEndpointExtension)
                {
                    found = true;
                    break;
                }
            }
            if (!found)
            {
                OperationContext.Current.Extensions.Add(new ExampleEndpointExtension());
            }
            return ExampleSessionProvider.Instance.OpenSession();
        }

        public void AfterInvoke(object correlationState)
        {
            ExampleSessionProvider.Instance.CloseSession(((ISession)correlationState));
            ((IDisposable)correlationState).Dispose();
        }
    }

The class below will bind Fluent NHibernate session to HttpContext if exists otherwise it will try to bind to WCF context.

public class ExampleSessionProvider
    {
        private const string SESION_KEY = "SESSION_KEY";

        private ExampleSessionProvider()
        {

        }

        private static ExampleSessionProvider instance;

        public static ExampleSessionProvider Instance
        {
            get
            {
                if (instance == null)
                {
                    instance = new ExampleSessionProvider();
                }
                return instance;
            }
        }

        private ISessionFactory sessionFactory;

        public void CreateSessionFactory()
        {
            sessionFactory = Fluently.Configure();                
                .ExposeConfiguration(BuildSchema)
                .BuildSessionFactory();
            // add more configuration options above !!!
        }

        public ISession GetSession()
        {
            if (HttpContext.Current != null)
            {
                return sessionFactory.GetCurrentSession();
            }
            if (OperationContext.Current != null)
            {
                foreach (IExtension extension in OperationContext.Current.Extensions)
                {
                    if (extension is ExampleEndpointExtension)
                    {
                        if (((ExampleEndpointExtension)extension).State.ContainsKey(SESION_KEY))
                        {
                            return (ISession)((ExampleEndpointExtension)extension).State[SESION_KEY];
                        }
                    }
                }
            }
            return null;
        }

        public ISession OpenSession()
        {
            ISession session = sessionFactory.OpenSession();

            if (HttpContext.Current != null)
            {
                ManagedWebSessionContext.Bind(
                    HttpContext.Current, session);
            }
            else if (OperationContext.Current != null)            
            {
                foreach (IExtension extension in OperationContext.Current.Extensions)
                {
                    if (extension is ExampleEndpointExtension)
                    {
                        if (((ExampleEndpointExtension)extension).State.ContainsKey(SESION_KEY))
                        {
                            ((ExampleEndpointExtension)extension).State.Remove(SESION_KEY);
                        }
                        ((ExampleEndpointExtension)extension).State.Add(SESION_KEY, session);
                    }
                }
            }
            return session;
        }

        public void CloseSession()
        {
            CloseSession(null);
        }

        public void CloseSession(ISession session)
        {
            if ((session != null) && (HttpContext.Current != null))
            {
                session = ManagedWebSessionContext.Unbind(HttpContext.Current, sessionFactory);
            }
            if (session != null)
            {
                if (session.Transaction != null &&
                    session.Transaction.IsActive)
                {
                    session.Transaction.Rollback();
                }
                else
                {
                    session.Flush();
                }
                session.Close();
            }
        }

        private static void BuildSchema(Configuration config)
        {
            config.SetProperty("current_session_context_class", "managed_web");
        }
    }

Before binding to HttpContext you need to create HttpModule described in post about Binding Fluent NHibernate session to HttpContext or you can remove parts related to HttpContext and bind always to WCF call.

Binding Fluent NHibernate session to HttpContext

noreply@blogger.com (Adam Matusiak) — Tue, 16 Mar 2010 11:10:00 +0000

It is very useful and common practice to bind Fluent NHibernate session to HttpContext. Easiest way to do it is to create HttpModule.

public class ExampleSessionModule : IHttpModule
    {
        public void Init(HttpApplication context)
        {
            context.BeginRequest += OpenSession;
            context.EndRequest += CloseSession; 
        }

        private void OpenSession(object sender, EventArgs e)
        {
            ExampleSessionProvider.Instance.OpenSession();            
        }
        
        private void CloseSession(object sender, EventArgs e)
        {
            ExampleSessionProvider.Instance.CloseSession();
        } 

        public void Dispose()
        {
            
        }
    }

The you just need a session provider.

public class ExampleSessionProvider
    {
        private const string SESION_KEY = "SESSION_KEY";

        private ExampleSessionProvider()
        {

        }

        private static ExampleSessionProvider instance;

        public static ExampleSessionProvider Instance
        {
            get
            {
                if (instance == null)
                {
                    instance = new ExampleSessionProvider();
                }
                return instance;
            }
        }

        private ISessionFactory sessionFactory;

        public void CreateSessionFactory()
        {
            sessionFactory = Fluently.Configure();                
                .ExposeConfiguration(BuildSchema)
                .BuildSessionFactory();
            // add more configuration options above !!!
        }

        public ISession GetSession()
        {
            if (HttpContext.Current != null)
            {
                return sessionFactory.GetCurrentSession();
            }           
            return null;
        }

        public ISession OpenSession()
        {
            ISession session = sessionFactory.OpenSession();
            ManagedWebSessionContext.Bind(HttpContext.Current, session);            
            return session;
        }

        public void CloseSession()
        {
            ISession session = ManagedWebSessionContext.Unbind(HttpContext.Current, sessionFactory);            
            if (session != null)
            {
                if (session.Transaction != null &&
                    session.Transaction.IsActive)
                {
                    session.Transaction.Rollback();
                }
                else
                {
                    session.Flush();
                }
                session.Close();
            }
        }

        private static void BuildSchema(Configuration config)
        {
            config.SetProperty("current_session_context_class", "managed_web");
        }
    }

And a bit of web.config configuration:

<system.web>
<httpModules>
<add name="ExampleSessionModule" type="Example.Session.ExampleSessionModule, Example" />
</httpModules>
</system.web>

OR (depending on IIS version)

<system.webServer>
<modules>
<add name="ExampleSessionModule" type="Example.Session.ExampleSessionModule, Example" />
</modules>
</system.webServer>

Last thing to do is to create session factory in Application Start event.

protected void Application_Start(Object sender, EventArgs e)
{
    ExampleSessionProvider.Instance.CreateSessionFactory();
}

WCF Session

noreply@blogger.com (Adam Matusiak) — Mon, 15 Mar 2010 17:40:00 +0000

You may need to define object that behaves like HttpSession and bind it to the WCF request. It's pretty useful when you want to keep Fluent NHibernate session opened during WCF call.

First define the Endpoint Extension that will hold the session state.

public class ExampleEndpointExtension : IExtension
    {
        public ExampleEndpointExtension()
        {
            State = new Dictionary();
        }

        public IDictionary State { get; private set; }

        public void Attach(OperationContext owner) { }
        public void Detach(OperationContext owner) { }
    }

Then you need to bind it to the Call Context.

public class ExampleEndpointContextInitializer : ICallContextInitializer
    {
        public object BeforeInvoke(InstanceContext instanceContext, IClientChannel channel, Message message)
        {
            bool found = false;
            foreach (IExtension extension in OperationContext.Current.Extensions)
            {
                if (extension is ExampleEndpointExtension)
                {
                    found = true;
                    break;
                }
            }
            if (!found)
            {
                OperationContext.Current.Extensions.Add(new ExampleEndpointExtension());
            }
            return new object();
        }

        public void AfterInvoke(object correlationState)
        {            
        }
    }

Last class to implement is Endpoint Behaviour Extension.

public class ExampleEndpointBehavior : BehaviorExtensionElement,IEndpointBehavior
    {

        public override Type BehaviorType
        {
            get { return typeof(ExampleEndpointBehavior); }
        }

        protected override object CreateBehavior()
        {
            return new ExampleEndpointBehavior();
        }

        public void Validate(ServiceEndpoint endpoint)
        {            
        }

        public void AddBindingParameters(ServiceEndpoint endpoint, BindingParameterCollection bindingParameters)
        {            
        }

        public void ApplyDispatchBehavior(ServiceEndpoint endpoint,
            EndpointDispatcher endpointDispatcher)
        {
            foreach (DispatchOperation operation in endpointDispatcher.DispatchRuntime.Operations)
            {
                operation.CallContextInitializers.Add(new ExampleEndpointContextInitializer());
            }
        }

        public void ApplyClientBehavior(ServiceEndpoint endpoint, ClientRuntime clientRuntime)
        {            
        }
    }

You should add your extension in web.config (or app.config).
Remember to define assembly together with version and culture. There's a bug in framework that will cause load fail if you omit it.

<behaviorextensions>
    <add name="ExampleBehaviorExtension" type="Example.Session.ExampleEndpointBehavior, Example, Version=1.0.0.0, Culture=neutral, PublicKeyToken=null" />
</behaviorExtensions>

Now you can use it:

foreach (IExtension extension in OperationContext.Current.Extensions)
{
    if (extension is ExampleEndpointExtension)
    {
        ((ExampleEndpointExtension)extension).State["SESSION_ITEM"] = "test";
    }
}

Example of usage in next post Binding Fluent NHibernate session to WCF call.

Fluent NHibernate - List of Int as String

noreply@blogger.com (Adam Matusiak) — Mon, 15 Mar 2010 15:35:00 +0000

Having field of type List it is sometimes worthy to save it in database as string rather than create separate table just to keep few values per record.
It seems to be completely wrong from design point of view but can be much cleaner solution when you have just few values per record.

To map this easily in Fluent create just two classes.
Then just define your field as IList.

User type convention:

public class IntListConvention : IUserTypeConvention
    {
        public bool Accept(IProperty target)
        {
            return target.PropertyType == typeof(IList);
        }

        public void Apply(IProperty target)
        {
            target.CustomTypeIs(typeof(IntListUserType));
        }

        public bool Accept(Type type)
        {
            return type == typeof(IList);
        }
    }

And new user type:

class IntListUserType : IUserType
    {
        public new bool Equals(object x, object y)
        {
            return x != null && x.Equals(y);
        }

        public int GetHashCode(object x)
        {
            return x.GetHashCode();
        }

        object IUserType.NullSafeGet(IDataReader rs, string[] names, object owner)
        {
            string stringValue = (string) NHibernateUtil.String.NullSafeGet(rs, names[0]) ?? String.Empty;
            List result = new List();
            foreach (string stringItem in stringValue.Split(",".ToCharArray(), StringSplitOptions.RemoveEmptyEntries))
            {
                int item;
                if (Int32.TryParse(stringItem, out item))
                {
                    result.Add(item);
                }
            }
            return result;
        }

        public void NullSafeSet(IDbCommand cmd, object value, int index)
        {
            StringBuilder stringValue = new StringBuilder();
            List result = value as List;
            if (value != null)
            {
                foreach (int item in result)
                {
                    if (stringValue.Length > 0)
                    {
                        stringValue.Append(",");
                    }
                    stringValue.Append(item);
                }
            }
            ((IDataParameter)cmd.Parameters[index]).Value = stringValue.ToString();
        }

        public object DeepCopy(object value)
        {
            return value;
        }

        public object Replace(object original, object target, object owner)
        {
            return original;
        }

        public object Assemble(object cached, object owner)
        {
            return cached;
        }

        public object Disassemble(object value)
        {
            return value;
        }

        public SqlType[] SqlTypes
        {
            get
            {
                return new[] { new SqlType(DbType.String, 1000) };
            }
        }

        public Type ReturnedType
        {
            get
            {
                return typeof(List);
            }
        }

        public bool IsMutable
        {
            get
            {
                return false;
            }
        }
    }

Shrinking MSSQL log files

noreply@blogger.com (Adam Matusiak) — Mon, 15 Mar 2010 15:09:00 +0000

Best solution I've found below.
You just have to replace XYZ with database name in few places.
You can also uncomment print to get filename.
Yes, I know it would be useful to have version that iterates through all the databases - I'm just too lazy.

alter database XYZ set recovery simple
GO
alter database XYZ set recovery full
GO
use XYZ
GO
declare @filename sysname
select @filename=Name from sys.database_files where Type=1
--print @filename
DBCC Shrinkfile(@filename,100) 
GO
alter database XYZ set recovery bulk_logged
GO