Support Engineer Days Online

W3wp High CPU Usage Scenario

omniao — Sun, 31 May 2015 02:26:17 GMT

I was working on one of my customers high CPU dumps for w3wp and the reason was synchronizing access to static collections Generic.Dictionary

Here is you start checking Dumps

!threadpool

CPU utilization: 100%

Worker Thread: Total: 32 Running: 32 Idle: 0 MaxLimit: 800 MinLimit: 8

Work Request in Queue: 886

!cpuid

CP F/M/S Manufacturer MHz

0 6,7,6 2490

1 6,7,6 2490

2 6,7,6 2490

3 6,7,6 2490

4 6,7,6 2490

5 6,7,6 2490

6 6,7,6 2490

7 6,7,6 2490

.time

Debug session time: Tue Jan 14 10:35:53.000 2014 (UTC + 2:00)

System Uptime: 26 days 19:17:15.503

Process Uptime: 0 days 0:08:32.000

Kernel time: 0 days 0:00:01.000

User time: 0 days 1:05:16.000

This is whole system cpu usage is 100% and 886 Request waiting in Queue

Total cpu time is number of CPUs* Process Uptime = 8 cpu x 0:08:32.000 = 1:8:26 ~68 min

We have 8 CPUs in total, so the process have 8 x 0:08:32.000 CPU total time which is 1:08:26. It means that, during the lifetime of the process, most of the CPU time is spent on user mode, which is 1:05:16.000/ 1:08:26 = ~95%.

# The cpu time of same threads are increasing over 3 dumps.

The CPU time taken for each thread:

Dump1:

!runaway

User Mode Time

Thread Time

39:319c 0 days 0:04:02.987

32:30b4 0 days 0:03:24.657

37:2ffc 0 days 0:03:21.397

31:32f8 0 days 0:03:19.322

36:3120 0 days 0:03:13.238

33:2330 0 days 0:03:11.460

41:900 0 days 0:03:03.987

Thread CPU Time= 0:19:37.48

Dump2:

Thread Time

38:319c 0 days 0:04:17.292

31:30b4 0 days 0:03:34.984

36:2ffc 0 days 0:03:33.752

30:32f8 0 days 0:03:29.821

35:3120 0 days 0:03:23.019

32:2330 0 days 0:03:21.038

40:900 0 days 0:03:14.814

Thread CPU Time= 0:21:4.75

Dump3:

Thread Time

38:319c 0 days 0:04:26.855

31:30b4 0 days 0:03:44.844

36:2ffc 0 days 0:03:43.331

30:32f8 0 days 0:03:37.714

35:3120 0 days 0:03:32.629

32:2330 0 days 0:03:30.055

40:900 0 days 0:03:24.142

Thread CPU Time= 0:22:4.5

Long Running Threads Details

So these are the threads increasing CPU Time. From Dump1, the threads are 39,32,37,31,36,33,41 they are all ThreadpoolWorker thredas.

OSID Special thread type

15 4b54 Gate

16 1f40 DbgHelper

17 4da8 GC

18 4dc GC

19 2170 GC

20 211c GC

21 4624 GC

22 1640 GC

23 3bc8 GC

24 42c4 GC

25 31c0 Finalizer

26 477c IOCompletion

27 1c30 ProfilingAPIAttach

28 3a9c IOCompletion

29 277c Timer

30 4c38 ADUnloadHelper

31 32f8 ThreadpoolWorker

32 30b4 ThreadpoolWorker

33 2330 ThreadpoolWorker

34 4c84 ThreadpoolWorker

35 1de8 ThreadpoolWorker

36 3120 ThreadpoolWorker

37 2ffc ThreadpoolWorker

38 3e0c ThreadpoolWorker

39 319c ThreadpoolWorker

41 900 ThreadpoolWorker

They are all running the web service AccessServer/AccessService.asmx each thread takes from 5 to 10 min and never complete

!aspxpages

Going to dump the HttpContexts found in the heap.

HttpContext Timeout Completed Running ThreadId ReturnCode Verb RequestPath+QueryString

0x00000000ffa5bb68 3600 Sec no 511 Sec 31 200 POST /AccessServer/AccessService.asmx

0x00000000ffaf5e98 3600 Sec no 508 Sec 39 200 POST /AccessServer/AccessService.asmx

0x00000000ffdd6328 3600 Sec no 287 Sec 55 200 POST /AccessServer/AccessService.asmx

0x00000000ffe0c3c8 3600 Sec no 64 Sec 63 200 POST /AccessServer/AccessService.asmx

0x000000011fb11358 3600 Sec no 511 Sec 33 200 POST /AccessServer/AccessService.asmx

…………………………………

All threads are stuck on the same method .SessionFactoryImpl.Close() And Dictionary. Remove()

0:039> KL

Child-SP RetAddr Call Site

00000000`06fece30 000007ff`0072bae1 mscorlib_ni!System.Collections.Generic.Dictionary`2[[System.__Canon, mscorlib],[System.__Canon, mscorlib]].Remove(System.__Canon)+0xf4

00000000`06feceb0 000007ff`0072c038 .SessionFactoryImpl.Close()+0x411

00000000`06fecf60 000007ff`001e70d2 AMWebService!AMWebService.BaseController.Dispose()+0x18

00000000`06fecf90 000007fe`f9fa10b4 AMWebService!AMWebService.AccessService.synchronize2(System.String)+0x102

00000000`06fed010 000007fe`f9fa11c9 clr!CallDescrWorker+0x84

00000000`06fed060 000007fe`f9fa1245 clr!CallDescrWorkerWithHandler+0xa9

00000000`06fed0e0 000007fe`fa167ef9 clr!MethodDesc::CallDescr+0x2a1

00000000`06fed320 000007fe`fa167dbf clr!CallWithValueTypes_RetArgSlotWrapper+0x91

00000000`06fed5b0 000007fe`fa167a74 clr!InvokeImpl+0x2e6

00000000`06feda40 000007fe`f8f7587f clr!RuntimeMethodHandle::InvokeMethodFast+0x1d8

Conclusion

=========

The Dictionary operation is being performed on a Generic Dictionary- Because Dictionary (and generic Dictionary) objects are not thread safe and if any update to a generic dictionary is not performed in a thread safe way, it will cause high CPU. Dictionary.Remove is an update operation.

so please refer that back to dev team how they access the generic dictionary; this must be done in a thread safe way.

Thread-Safe Collections http://msdn.microsoft.com/en-us/library/dd997305(v=vs.100).aspx

High CPU in .NET app using a static Generic.Dictionary http://blogs.msdn.com/b/tess/archive/2009/12/21/high-cpu-in-net-app-using-a-static-generic-dictionary.aspx

The MSDN documentation about Generic.Dictionary has the following information about the thread safety of Dictionary objects

A Dictionary can support multiple readers concurrently, as long as the collection is not modified. Even so, enumerating through a collection is intrinsically not a thread-safe procedure. In the rare case where an enumeration contends with write accesses, the collection must be locked during the entire enumeration. To allow the collection to be accessed by multiple threads for reading and writing, you must implement your own synchronization.

What is happening here, and causing the high CPU is that the FindEntry method walks through the dictionary, trying to find the key. If multiple threads are doing this at the same time, especially if the dictionary is modified in the meantime you may end up in an infinite loop in FindEntry causing the high CPU behavior and the process may hang.

Resolution:

Access Dictionary in Thread safe way. For example Use the "lock" statement inside your codes and anywhere else that uses non-thread-safe objects.

Hashtable and Dictionary thread safety considerations http://blogs.msdn.com/kimhamil/archive/2008/03/08/hashtable-and-dictionary-thread-safety-considerations.aspx

Azure Managed Cache Service Sample Solution.

omniao — Sun, 24 May 2015 11:06:00 GMT

Azure is a growing collection of integrated services – compute, storage, data, networking and app – that help you move faster, do more and save money. But that’s just scratching the surface.

Azure supports any operating system, language, tool and framework – from Windows to Linux, SQL Server to Oracle, C# to Java. It puts the best of Windows and Linux ecosystems at your fingertips, so you can build great applications and services that work with every device.

In this Article today; you will learn how to Azure Web APPs scenarios using easy techniques which will allow you to dive quickly into the Azure web APP world.

But first things first you need to know some Azure Facts:

Azure Web Apps enable you to easily deploy and scale web applications written in a variety of languages (including .NET, Java, PHP, Node.js and Python). The Web App service automatically handles patch management and configuration of your infrastructure – allowing you to avoid having to worry about VM management. Built-in AutoScale capability enables you to scale up/down your web application based on actual customer traffic – enabling you to serve any load, and to save money when your site is not at peak. Built-in continuous deployment support enables you to automatically deploy code updates from Visual Studio Online, GitHub and BitBucket.
You can integrate Web Apps with a backend Relational or NoSQL database. SQL Database is a relational database-as-a-service offering that you can use to host and scale relational databases. DocumentDB is a NoSQL document service that you can use to store, index, and query JSON based data.

So now let’s imagine the below scenario:

You are developing an e-commerce website targeting 5 different countries, each country got its own suppliers, so you are having the same portal and you have cached the application to enable faster reach from users, what if you want the server to have a super-fast access to data with a distributed cache, in-memory, scalable solutions to have, that’s simple Azure managed cache service which will give you the below benefits

Pre-built ASP.NET providers for session state and page output caching, enabling acceleration of web applications without having to modify application code.
Caches any serializable managed object - for example: CLR objects, rows, XML, binary data.
Consistent development model across both Azure and Windows Server AppFabric.

So in order to enable all these features, we need to do a very simple task

Creating the cache

Once a Managed Cache Service instance is created using the PowerShell cmdlets it can be viewed and configured in the Azure Management Portal, and we just need to run the below command

New-AzureManagedCache -Name Otlob -Location "Egypt" -Sku Basic -Memory 128MB

Basic 128MB cache is created with name Otlob, in the Egypt geographic region.

Configure the cache
The Configure tab for Cache in the Management Portal is where you configure the options for your cache. Each cache has a default named cache, and the Standard and Premium cache offerings

support up to nine additional named caches, for a total of ten. Each named cache has its own set of options which allow you to configure your cache in a highly flexible manner.



To create a named cache, type the name of the new cache into the Name box, specify the desired options, click Save, and click Yes to confirm. To cancel any changes, click Discard.

Configure the cache clients
A cache created using the Managed Cache Service is accessible from Azure applications running on Azure Websites, Web & Worker Roles and Virtual Machines. A NuGet package is provided that simplifies the configuration of cache client applications.
To configure a client application using the Cache NuGet package, right-click the project in Solution Explorer and choose Manage NuGet Packages.

Type WindowsAzure.Caching into the Search Online text box, and select Windows Azure Cache from the results. Click Install, and then click I Accept.

The NuGet package does several things: it adds the required configuration to the config file of the application, and it adds the required assembly references. For Cloud Services projects, it also adds a cache client diagnostic level setting to the ServiceConfiguration.cscfg file of the Cloud Service.

For ASP.NET web projects, the Cache NuGet package also adds two commented out sections to web.config. The first section enables session state to be stored in the cache, and the second section enables ASP.NET page output caching. For more information, see How To: Store ASP.NET Session State in the Cache and How To: Store ASP.NET Page Output Caching in the Cache.

The NuGet package adds the following configuration elements into your application's web.config or app.config. A dataCacheClients section and acacheDiagnostics section are added under the configSections element. If there is no configSections element present, one is created as a child of theconfiguration element.

    type="Microsoft.ApplicationServer.Caching.DataCacheClientsSection,

          Microsoft.ApplicationServer.Caching.Core"

    allowLocation="true"

    allowDefinition="Everywhere" />

    type="Microsoft.ApplicationServer.Caching.AzureCommon.DiagnosticsConfigurationSection,

          Microsoft.ApplicationServer.Caching.AzureCommon"

    allowLocation="true"

    allowDefinition="Everywhere" />

These new sections include references to a dataCacheClients element, which is also added to the configuration element.













After the configuration is added, replace the following two items in the newly added configuration.

Replace [Cache role name or Service Endpoint] with the endpoint, which is displayed on the Dashboard in the Management Portal.

2.Uncomment the securityProperties section, and replace [Authentication Key] with the authentication key, which can be found in the Management Portal by clicking Manage Keys from the cache dashboard.

These settings must be configured properly or clients will not be able to access the cache.

For Cloud Services projects, the NuGet package also adds a ClientDiagnosticLevel setting to the ConfigurationSettings of the cache client role in ServiceConfiguration.cscfg. The following example is the WebRole1 section from a ServiceConfiguration.cscfg file with a ClientDiagnosticLevel of 1, which is the default ClientDiagnosticLevel.




             value="1" />

The client diagnostic level configures the level of caching diagnostic information collected for cache clients. For more information, see Troubleshooting and Diagnostics

The NuGet package also adds references to the following assemblies:

Microsoft.ApplicationServer.Caching.Client.dll

Microsoft.ApplicationServer.Caching.Core.dll

Microsoft.WindowsFabric.Common.dll

Microsoft.WindowsFabric.Data.Common.dll

Microsoft.ApplicationServer.Caching.AzureCommon.dll

Microsoft.ApplicationServer.Caching.AzureClientHelper.dll

If your project is a web project, the following assembly reference is also added:

Microsoft.Web.DistributedCache.dll.

These assemblies are located in the C:\Program Files\Microsoft SDKs\Windows Azure\.NET SDK\[sdk version]\ref\Caching\ folder.

Once your client project is configured for caching, you can use the techniques described in the following sections for working with your cache.

References

http://azure.microsoft.com/en-us/services/cache/

http://azure.microsoft.com/en-gb/overview/what-is-azure/

http://azure.microsoft.com/en-gb/documentation/scenarios/web-app/

http://www.techrepublic.com/blog/10-things/10-reasons-to-use-azure-for-your-cloud-apps/

http://azure.microsoft.com/en-us/documentation/articles/cache-dotnet-how-to-use-service/

http://azure.microsoft.com/en-us/documentation/articles/cache-dotnet-how-to-use-service/#create-cache

How to loose session data in Web Farm ?

omniao — Wed, 15 Oct 2014 07:16:00 GMT

There are many things can cause session data lose but I came across very interesting scenario where session data intermittently get lost while storing
session data in SqlSessionStateStore. And after long troubleshooting found it was the App ID assigned to the same web application is different in the 2 servers in the web farm. Thus when request complete processing on
the 2nd server it consider as new one and assign new AppID.

SqlSessionStateStore scopes data by application, so that multiple applications can store sessions in one session state database. To that end, the session state database contains an ASPStateTempApplications table that records application names and application IDs. Application names are not explicitly specified as they are for other providers; instead, SqlSessionStateStore uses the website's IIS metabase path as the application name. Application IDs
are hashes generated from application names by the stored procedure GetHashCode. (SqlSessionStateStore differs in this respect, too, from other SQL providers, which use randomly generated GUIDs as application IDs.)

Ref http://msdn.microsoft.com/en-gb/library/aa478952.aspx>

Here is how can you reproduce the same:

In Web Farm of 2 web servers for example But not in the same order so you may have different app id

Then when you check the ASPStateTempApplications table in ASPState DB you will find 2 different record for the same APP like:

Hence when session gets created it is formed as Session ID + App ID. So if the APP ID is different its hashing would be different so it will generate 2 different
session IDs for the same user as in the following example

o2f3t4crsssufg0ppenzvduv0e8cc1f9 = O2f3t4crsssufg0ppenzvduv (SessionID) + 0e8cc1f9 (APP ID)

o2f3t4crsssufg0ppenzvduv2476b033 =O2f3t4crsssufg0ppenzvduv (Session ID) + 2476b033 (APP ID)

So make sure to make every single server in the farm identical.

Log Parser And IIS Log Analysis.

omniao — Thu, 09 Jan 2014 16:42:00 GMT

First have Log Parser installed. Then Start it from C:\Program Files (x86)\Log Parser 2.2

Here are some Log Parser queries which can be used across IIS logs for analysis

Request Type Distribution

"SELECT EXTRACT_EXTENSION(TO_UPPERCASE(cs-uri-stem)) AS ExtType, COUNT(*) INTO FrontEndRequestType.csv FROM *.log GROUP BY ExtType ORDER BY COUNT(*) DESC " -i:W3C -o:CSV

Top 20 Hits

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, COUNT(*) AS Hits INTO Top20Hits.csv FROM *.log WHERE URI LIKE '%%.ASPX' GROUP BY URI ORDER BY Hits DESC " -i:W3C -o:CSV

Top 20 Hits Overall

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, COUNT(*) AS Hits INTO Top20Hits.csv FROM *.log GROUP BY URI ORDER BY Hits DESC " -i:W3C -o:CSV

Top 20 Hits Chart

"SELECT TOP 20 cs-uri-stem, COUNT(*) AS Hits INTO Top20Hits.jpg FROM *.log GROUP BY cs-uri-stem ORDER BY Hits DESC" -i:W3C -o:CHART -charttype:ColumnClustered -groupsize:640x480 -chartTitle: Top 20 Hits Chart

Top 20 ASPX Hits

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, COUNT(*) AS Hits INTO Top20ASPX.csv FROM *.log WHERE URI LIKE '%.ASPX' GROUP BY URI ORDER BY Hits DESC " -i:W3C -o:CSV

Top 20 Slowest ASPX Pages

"SELECT TOP 20 cs-uri-stem, max(time-taken) as MaxTime, avg(time-taken) as AvgTime INTO Top20SlowASPX.csv FROM *.log WHERE extract_extension(to_lowercase(cs-uri-stem)) = ‘aspx’ GROUP BY cs-uri-stem ORDER BY MaxTime DESC" -i:W3C -o:CSV

Top 20 ASMX Hits

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, COUNT(*) AS Hits INTO Top20ASMX.csv FROM *.log WHERE URI LIKE '%.ASMX' GROUP BY URI ORDER BY Hits DESC " -i:W3C -o:CSV

Top 20 Client IP Addresses

"SELECT TOP 20 c-ip, COUNT(*) AS Hits INTO Top20ClientIP.csv FROM *.log GROUP BY c-ip ORDER BY Hits DESC " -i:W3C -o:CSV

Requests Per Hour

"SELECT TO_LOCALTIME(QUANTIZE(TO_TIMESTAMP(date, time),3600)) AS Hours, COUNT(*) AS Hits INTO RequestsPerHour.jpg FROM *.log GROUP BY Hours ORDER BY Hours " -i:W3C -o:CHART -charttype:ColumnClustered -groupsize:640x480 -chartTitle:"Requests Per Hour"

HTTP Status Counts

"SELECT DISTINCT sc-status AS Status, COUNT(*) AS Hits INTO HTTPStatusCount.csv FROM *.log GROUP BY Status ORDER BY Status ASC " -i:W3C -o:CSV

HTTP Status Distribution

"SELECT DISTINCT sc-status AS Status, COUNT(*) AS Percent INTO HTTPStatusDistribution.jpg FROM *.log GROUP BY Status ORDER BY Percent DESC " -i:W3C -o:CHART -chartType:PieExploded -chartTitle: "HTTP Status Distribution"

Top 20 HTTP 304 Errors

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, sc-substatus, sc-win32-status, COUNT(*) AS Hits INTO Top20Http304Errors.csv FROM *.log WHERE sc-status = 304 GROUP BY URI, sc-substatus, sc-win32-status ORDER BY Hits DESC" -i:W3C -o:CSV

Top 20 HTTP 404 Errors

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, sc-substatus, sc-win32-status, COUNT(*) AS Hits INTO Top20Http404Errors.csv FROM *.log WHERE sc-status = 404 GROUP BY URI, sc-substatus, sc-win32-status ORDER BY Hits DESC" -i:W3C -o:CSV

Top 20 HTTP 403 Errors

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, sc-substatus, sc-win32-status, COUNT(*) AS Hits INTO Top20Http403Errors.csv FROM *.log WHERE sc-status = 403 GROUP BY URI, sc-substatus, sc-win32-status ORDER BY Hits DESC" -i:W3C -o:CSV

Top 20 HTTP 500 Errors

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, sc-substatus, sc-win32-status, COUNT(*) AS Hits INTO Top20500Errors.csv FROM *.log WHERE sc-status = 500 GROUP BY URI, sc-substatus, sc-win32-status ORDER BY Hits DESC" -i:W3C -o:CSV

Top 20 HTTP 503 Errors

"SELECT TOP 20 TO_UPPERCASE(cs-uri-stem) AS URI, sc-substatus, sc-win32-status, COUNT(*) AS Hits INTOTop20503Errors.csv FROM *.log WHERE sc-status = 503 GROUP BY URI, sc-substatus, sc-win32-status ORDER BY Hits DESC" -i:W3C -o:CSV

Top 20 Longest Processing Requests

"SELECT TOP 20 cs-uri-stem AS URI, date AS Date, time AS Time, sc-status, time-taken as TimeTaken(ms) INTO

Top20LongRunningRequests.csv FROM *.log ORDER BY time-taken DESC" -i:W3C -o:CSV

To view results in Charts- this needs Office Web Components; there is a good post about this here http://blogs.msdn.com/b/carloc/archive/2008/08/07/charting-with-logparser.aspx

Can I Develop Windows 8.1 App using VS 2012?

omniao — Wed, 08 Jan 2014 08:56:30 GMT

The Short Answer is NO. I had a customer looking to use VS 2012 to develop Windows phone 8.0 and 8.1 Apps.

The current Windows Phone 8.0 SDK is compatible with Visual Studio 2012 (Update 2) and higher (see http://dev.windowsphone.com/en-us/downloadsdk) .

To develop Windows Store Apps for Windows 8.0 you generally need Visual Studio 2012 or higher. Different editions of Visual Studio may introduce additional restrictions.

To develop Windows Store Apps for Windows 8.1 you generally need Visual Studio 2013 or higher.

Does Dacpack supports sqlcmd variables?

omniao — Mon, 30 Dec 2013 23:40:27 GMT

I have a got a question from a customer want to use dbdacfx provider with Web Deploy 3 which is working fine and only failing with sqlcmd variables with below error?

And the Question was Does Dacpack supports sqlcmd variables?

Ref dbDacFx Provider for Incremental Database publishing http://www.iis.net/learn/publish/using-web-deploy/dbdacfx-provider-for-incremental-database-publishing

And the Answer is: The dbDacFx provider does not support sqlcmd variables at this time. This functionality might be added later, but for now you’ll have to rely on the dbsqlpackage provider for that sqlcmd variables support.

How to block certain client apps or browsers from accessing a web App in IIS?

omniao — Sun, 01 Dec 2013 03:48:19 GMT

I got a question from one of my colleagues asking how to block certain client application or browsers from accessing a web App in IIS ?

The Question was raised at first from a customer have already disabled Active Sync and still his users can access through 3rd parity Applications like Cloud Magic https://cloudmagic.com and still would like to prevent that ??

So to do so we can use common request-filter settings to improve the security of the IIS web server. Any Client access your site, the browser software identifies itself by sending a user-agent string, and that is already get logged in IIS log as below:

#Fields: date time s-ip cs-method cs-uri-stem cs-uri-query s-port cs-username c-ip cs(User-Agent) cs(Referer) sc-status sc-substatus sc-win32-status time-taken

2013-12-01 10:51:45 ::1 GET / - 80 - ::1 Mozilla/5.0+(Windows+NT+6.3;+WOW64)+AppleWebKit/537.36+(KHTML,+like+Gecko)+Chrome/31.0.1650.57+Safari/537.36 - 500 24 50 2206

User Agent (cs(UserAgent)) is the browser type that the client used.

So simply you can block Chrome by configuring a Deny Rule for user-agent= Chrome

Forexample



















Or For our main Customer who liked to block CloudMagic it would be:

Why can't place an ascx file in the App_Code folder?

omniao — Fri, 20 Sep 2013 01:11:48 GMT

When a developer creates a user controls (ascx files); it is not allowed to put it in the App_Code folder.

Before we answer this lets quickly remember the value of App_Code folder: it is just one of the web site project folders that can store source code BUT its content would be automatically compiled at run time. And this is reason because that would cause the user control to be compiled out of its required sequence and therefore is not allowed.

.....

IIS8: What is the problem that CCS (Centralized Certificate Store) resolved?

omniao — Tue, 10 Sep 2013 00:43:00 GMT

Securing sites is already a mandatory step in many business
as usually it is saving sensitive data and documents. Moreover, with cloud
hosting, server capacity must be added and removed dynamically. Before Windows
Server 2012, there were a couple of obstacles must be considered :

Scalability: We can only use one certificate for one IP:Port combination only.

Manageability: Storing certificates locally on the web server so for example when we
renew a certificate it must be repeated on every server.

And CCS is the Solution: The DNS name and CN name of the
certificate must match. This contract can be further extended to the file names
of the certificates. (For example, www.Server1.com would use the certificate
with a file name www. Server1.com.pfx.) This contract enables Windows Server
2012 to have just one SSL binding, regardless of the number of secure sites
that use CCS store for binding certificates.

Simply previously if there were 100 secure sites, there would 100 SSL
bindings “Explicit Binding”. But with CCS regardless the number of secure sites, there would be single SSL
bindings “Implicit Binding”.

How to configure it in IIS ?

How to configure website to use CCS?

When choose CCS then no need to select certificate … it just go and picked it up from the share but it must be certificate name like website name exactly

Where CCS is Configured ?

CCS configuration is stored in Registry and not in Application Host

IIS sslFlags is just to verify if website uses CCS.

sslFlags="3" />

Don't add special bindings for FTP site

omniao — Thu, 16 May 2013 06:34:51 GMT

Shortly if you add bindings for ftp site, it will start to show abnormal behavior. You may not be able to browse it by IP address or name.

You may see error 530

530-Valid hostname is expected.

Win32 error:   No such host is known.
Error details: Hostname didn't match any configured ftp site.
530 End

To resolve that simply remove any added bindings for the site and use the FTP name you added in the FTP creation wizard.

HTTP pass the host name in the headers that are passed between the client and server, but FTP does not support that option.

Ref http://blogs.msdn.com/b/robert_mcmurray/archive/2007/08/28/virtual-hosts-and-host-names-in-ftp7.aspx