Avanade xRM Blog

Updated CrmService.js

Back in 2008, while at Ascentium, we posted our internal CrmService.js helper.  This JavaScript wrapper makes it easier to make CRM SDK calls from with in JavaScript.  Today, I am posting a minor update.  It's not a major update, but I wanted to post a more recent version post-Avanade merger.  This version has a few minor issues fixed and adds a wrapper for Assign.  Here is the new wrapper:

Usage - Assign(new owner id, record id, target (same as the WSDL), optional callback)

Example - service.Assign(strNewOwnerId, strPriceQuoteId, "TargetOwnedAC_pricequote");

You can download it here:

Download AvanadeCrmService

Cheers,

This posting is provided "AS IS" with no warranties, and confers no rights.

Import Organization Issue - Would you please upgrade my reports?

Recently we had a situation, where we needed to import a CRM organization from one environment to another.  To help with this, we were sure to make use of the import speed fix in RU8. The challenge though was that the source environment used SQL Reporting Services 2005, while the target environmented used SQL Reporting Services 2008.

During the import we got the following error:

Error| Import Organization (Name=CRMTEST, Id=79a213fe-7f89-df11-95e8-00505692406b) failed with Exception:
Microsoft.Crm.CrmReportingException: The report definition is not valid.  Details: The report definition has an invalid target namespace '_http://schemas.microsoft.com/sqlserver/reporting/2008/01/reportdefinition' which cannot be upgraded. ---> System.Web.Services.Protocols.SoapException: The report definition is not valid.  Details: The report definition has an invalid target namespace '_http://schemas.microsoft.com/sqlserver/reporting/2008/01/reportdefinition' which cannot be upgraded. ---> The report definition is not valid.  Details: The report definition has an invalid target namespace '_http://schemas.microsoft.com/sqlserver/reporting/2008/01/reportdefinition' which cannot be upgraded.
   at System.Web.Services.Protocols.SoapHttpClientProtocol.ReadResponse(SoapClientMessage message, WebResponse response, Stream responseStream, Boolean asyncCall)
   at System.Web.Services.Protocols.SoapHttpClientProtocol.Invoke(String methodName, Object[] parameters)
   at Microsoft.Crm.ReportingServices2005.ReportingService2005.CreateReport(String Report, String Parent, Boolean Overwrite, Byte[] Definition, Property[] Properties)
   at Microsoft.Crm.Reporting.ReportServer.UploadReport(String path, Byte[] reportDefinition, String name, String description)
   --- End of inner exception stack trace ---
   at Microsoft.Crm.Reporting.ReportServer.UploadReport(String path, Byte[] reportDefinition, String name, String description)
   at Microsoft.Crm.Reporting.ReportServer.UploadReport(SRSReport report, String reportId, String name, String description)
   at Microsoft.Crm.Setup.Server.Utility.ReportsUtility.OrganizationPublishReportsScaleGroup(IDbCommand command, Uri reportingUrl, String orgUniqueName)
   at Microsoft.Crm.Tools.Admin.DBImportHelper.RePublishReports(IDbCommand command, String organizationUniqueName, Uri reportUrl)
   at Microsoft.Crm.Tools.Admin.ImportOrganizationInstaller.Import(Guid organizationId, String organizationUniqueName, String organizationFriendlyName, String sqlServerName, String databaseName, Uri reportServerUrl, String privilegedUserGroupName, String sqlAccessGroupName, String userGroupName, String reportingGroupName, String privilegedReportingGroupName, ICollection`1 users, MultipleTenancy multipleTenancy)

 

The fix was simple, although not neccessarily obvious.  If we read the error, the issue seems clearish:  The namespace "http://schemas.microsoft.com/sqlserver/reporting/2008/01/reportdefinition" which is defined in the RDL of a Report was not found or is invalid.

 

But why was it not valid or not found?  The problem is that CRM relies on the client-side reporting components provided by the Microsoft Report Viewer Control to interact with the SSRS server.  In this case, the server only had the 2005 version of the control installed and did not have the 2008 version.  Installing the 2008 version and re-running the import fixed the issue.  You can download the latest version of this control here:

 

http://www.microsoft.com/downloads/details.aspx?familyid=BB196D5D-76C2-4A0E-9458-267D22B6AAC6&displaylang=en

 

 

Enjoy,

 

This posting is provided "AS IS" with no warranties, and confers no rights.

 

Scribe 7.0 Released

When it comes to getting data into Microsoft Dynamics CRM, users and partners have a lot of options.  Including:

• Microsoft Data Migration Manager
  
• Microsoft Dynamics CRM Bulk Import
  
• Custom written migration tools (like our beloved "Dancing Monkey" framework)
  
• ETL tools like SQL Server Integration Services
  
• 3rd Party tools like Scribe

What's cool about all of these, is that they all make use of the same underlying Microsoft Dynamics CRM web services and thus are all supported ways of getting data into the system.

Depending on a number of factors, it is common for us to use Scribe Insight, especially for data migrations.  This post just highlights some of the new features in Scribe 7.0, with a few that I am particularly excited about called out.

You can read more about Scribe 7.0 here:
https://openmind.scribesoftware.com/download/ReleaseNotes700ScribeInsight

You can download the bits here:
https://openmind.scribesoftware.com/html/insight_download

Scribe 7.0 Release Themes

Scribe Insight 7.0 adds a number of significant pieces of functionality focused along the following themes:

1. Improved Productivity: Scribe Insight has always been based on a very simple concept - Moving data from a SOURCE to a single TARGET. Scribe Insight 7.0’s new Multi-Target capability allows users to design integrations (Scribe DTS files) which include 'steps' that can operate on multiple target data sources or applications. By breaking the current source/target paradigm Scribe users are able to develop sophisticated integrations much more simply and quickly than before.
   
2. Greater Flexibility: With the new paradigm of multiple target connections in Scribe Insight 7.0 operations that were challenging, if not impossible, previously can now be accomplished simply. For example, with a Sales Order integration between CRM and ERP systems Sales Order data can be passed from CRM to ERP and then the ERP-assigned order number can be immediately updated back into the CRM system in a single operation.
   
3. Streamlined Maintainability: Complex integrations can be developed with fewer moving parts in Scribe Insight 7.0. This means less time is required for development, modification and maintenance. Processes that previously consisted of multiple linked integrations (Scribe DTS files) now can be neatly organized into a single integration.
   
4. Simplified Cloud Connectivity: Scribe Insight 7.0's new multi-target capabilities coupled with the Scribe Web Services adapter provide a simple and convenient way to take advantage of commercial or custom web services during an integration process. Operations such as using a commercial web service to cleanse an address during an integration process are now easily possible in a single operation.

Stuff I am particularly excited by:

• Multiple connections and multiple data sources - One DTS can now connect to multiple systems - So if you need to make ETL decisions based on data in another system, that is now possible with making a bunch of messy daisy-chained DTS files.  They also added calling web services in the DTS, as well as making DBLOOKUP allow you to hit the different data sources.
  
• Automatically create required MSMQ Queues - This just makes things that much easier.
  
• Bug #32308 Fixed - Having to do with only parsing part of an XML file - We had run into this and I am glad to see it fixed.

Cheers,

This posting is provided "AS IS" with no warranties, and confers no rights.

Tweedle Dee & Tweedle Dum - SharePoint 2010

As the XRM story unfolds, Microsoft Dynamics CRM and Microsoft SharePoint will hopefully continue to grow closer together and indeed better together.  CRM 5.0 already includes SharePoint document library integration out the box.  This of course makes sense; easily 80% of the CRM deployments we work on include SharePoint as part of the solution... so it is natural for Microsoft to make this easier in the next wave.

However, like CRM, SharePoint has its little quirks.  One of them has to do with search results not showing up for users.  This can be very frustrating and hard to debug, but if you have a cross-domain SharePoint 2010 environment you might find that users from the domain where SharePoint is not installed can not find any records when they do a search.  Read on to learn more.

The issue had to do with my environment and the fact that my system was stand-alone in one domain, but my users were from another USERACCOUNT domain.  First some context.  This is what happens when you perform a search (simplified):

1. User logs into the site and enters their query
2. The query is passed to the search service for processing:
   http://[SERVERNAME]:[PORT]/[GUID]/SearchService.svc
3. The user is redirected to the “results” page
4. The results page calls the search service again to get the results
   
   fd52ef63-dfa8-4518-be93-d92a270a27b9: Query completed 0 ms, detailed time:  Query stage execution ms times: 0 0 0 0 0 0 0 0   Query stage cpu ms times: 0 0 0 0 0 0 0 0   Query stage hit counts: 1 1 1 7 1 0 1 1   Cursor count: 206  Mapped page count: 266  Total index count: 14    [srequest.cxx:5526]
   
   
5. The results are then filtered based on the requesting users permissions
6. If all goes well, the results are provided to the results page which transforms them and presents them to the user.

In my case, the problem was with Step #5.  The log showed I was getting this error:

AuthzInitializeContextFromSid failed with ERROR_ACCESS_DENIED. This error indicates that the account under which this process is executing may not have read access to the tokenGroupsGlobalAndUniversal attribute on the querying user's Active Directory object. Query results which require non-Claims Windows authorization will not be returned to this querying user.

The root cause of the problem is that the Service Account that is being used by the Search Service AppPool did not have access to user requesting the search’s domain.  In this example, the AppPool was NETWORK SERVICE on the Customer1 machine; but the user was coming from USERACCOUNT domain.  In my search for the fix, the following blog posts were “helpful” but not entirely complete.

• http://mattstratton.com/tech-tips/configuring-sharepoint-2010-search-in-a-one-way-trust-scenario
• http://blogs.technet.com/b/wbaer/archive/2010/04/11/managed-accounts.aspx

To get around this issue, I had to do the following:

1. Add a new “Managed Account” from the USERACCOUNT domain
2. Configure the SharePoint Service AppPool account to use this account

You have to do this via PowerShell, because the “Register New Managed Account” won’t work for similar reasons as above.  The procedure is as follows:

1. Load the SharePoint 2010 PowerShell environment
2. Run:  new-spmanagedaccount -Credential USERACCOUNTDOMAIN\USERNAME
3. Load SharePoint Central Administrator
4. Go to:  Security | Configure Service Accounts
5. Select: Service Application Pool – SharePoint Web Services Default
6. Set the account to the one you just created in Step #2
   
   
   Note: This account must have appropriate elevated privileges within SharePoint and the server to run the service account under.
   
   
   
   
7. Click OK
   

Before you embark on this fix, you should make sure that everything is configured correctly and working.  You should also consider enabling Diagnostic Logging and find the error message I noted above to make sure that this is indeed why you are not getting results.

Some general tips for searches not getting results:

• Verify that the index is both crawling and indexing the expected items.
• This is done via Central Administrator, under General Application Settings | Search |  Farm Search Administration | Search Service Application
• From here you can use the Crawl Log to see if the items you want indexed are indexing.  You can also verify that your content is within the scope of Content Sources
• Verify that the Crawler Service has permission to read the items you want indexed
• Verify that the Site / Site Collection has the “Allow this site to appear in search results” enabled
• Verify that the List you want indexed has “Allow items to appear in search results” enabled.
• Verify that the SharePoint Server 14 Search Service is running

Cheers,

This posting is provided "AS IS" with no warranties, and confers no rights.

CRM Online: Security Features Whitepaper

This came out a week or two ago; but it is a good read if you are getting (or have) questions about the various ways data is protected within CRM Online's.

"Microsoft takes a holistic approach to providing a highly secure environment for Microsoft Dynamics CRM Online. After an overview of the inherent risks to three key areas of the service, the remaining sections of this paper describe how Trustworthy Computing, Microsoft’s core commitment to build software and services that better help protect customers and the industry, is reflected in the design and operation of Microsoft Dynamics CRM Online."

http://www.microsoft.com/downloads/details.aspx?displaylang=en&FamilyID=eb07464e-ef76-4785-af13-9c5617e14644

Cheers,

This posting is provided "AS IS" with no warranties, and confers no rights.

Quick Tip - Installing CRM 4.0 with SSRS 2008

Wow time does fly, it's been over a month since my last post!  However, now that the Ascentium / Avanade deal is closed and we are comfortably in our new home, I have the time to start posting again.

As part of our move to Avanade, I had to setup a staging CRM server for our incoming Ascentium team data.  I installed Microsoft Dynamics CRM 4.0 on an environment with SQL Reporting Services 2008 (SSRS) and then imported our old organization.  During the import we got the following error:

19:16:29|  Error| Import Organization (Name=Ascentium, Id=5223ab21-5a61-df11-ac17-00155d02a321) failed with Exception: System.Reflection.TargetInvocationException: Exception has been thrown by the target of an invocation. ---> System.Runtime.InteropServices.COMException (0x800700B7):

Cannot create a file when that file already exists. (Exception from HRESULT: 0x800700B7)

--- End of inner exception stack trace ---
at System.DirectoryServices.DirectoryEntry.Invoke(String methodName, Object[] args)
at Microsoft.Crm.Setup.Server.Utility.IIsUtility.StartWebSite(String metabasePath, String machineName)
at Microsoft.Crm.Setup.Server.Utility.ReportsUtility.ConfigureReports(Uri reportingUrl, String organizationName, Boolean grantNetworkService)
at Microsoft.Crm.Tools.Admin.DBImportHelper.RePublishReports(IDbCommand command, String organizationUniqueName, Uri reportUrl)
at Microsoft.Crm.Tools.Admin.ImportOrganizationInstaller.Import(Guid organizationId, String organizationUniqueName, String organizationFriendlyName, String sqlServerName, String databaseName, Uri reportServerUrl, String privilegedUserGroupName, String sqlAccessGroupName, String userGroupName, String reportingGroupName, String privilegedReportingGroupName, ICollection`1 users, MultipleTenancy multipleTenancy)

Unfortunately, this error is not very useful or accurate.  The problem really had nothing to do with creating a file, well maybe it did, but the real problem is that Internet Information Server (IIS) could not restart.  The error from the Windows Event Viewer was:

The World Wide Web Publishing Service (WWW Service) did not register the URL
prefix http://*:80/ for site 1. The necessary network binding may already be in use. The site has been disabled. The data field contains the error number.

As you can see, the real issue was that IIS could not start the default web site (port 80).  This was because another site (SSRS in fact) was already bound to port 80. Unfortunately, Microsoft Dynamics CRM setup is not aware that SSRS 2008 does not actually host its pages in IIS and instead hosts it from the SSRS service.  Because of this, setup (Deployment Manager) just default to assume it is the default web site.

The fix was to temporarily disable the other site (the SSRS 2008 service) on port 80, which allowed CRM to start the IIS set and get past the error.  After CRM was happy, I re-disabled the default web site.  CRM 5.0 is aware of SSRS 2008 and does not have this problem.

Cheers,

This posting is provided "AS IS" with no warranties, and confers no rights.

Microsoft CRM 4.0 SDK Update Released

Archive:  May 6, 2010

An updated version of the Microsoft Dynamics CRM 4.0 SDK has been released.  The new version (4.0.12) has some really cool new stuff in it.  You can download it here:

http://www.microsoft.com/downloads/details.aspx?FamilyID=82e632a7-faf9-41e0-8ec1-a2662aae9dfb&displaylang=en

Its important to note that the schemas that the SDK assemblies are built against changed in RU7.  So be mindful of what DLL you are using for compilation.

Here is a summary of the big new features:

• New Microsoft.xRM folder with Advanced Developer Extensions
• New authentication samples for CRM Online

Here is the full list:

New and updated topics	Description of changes
SDK\SDKReadme.htm	Updated the SDK readme.
SDK\Bin	Updated binaries to Microsoft Dynamics CRM Update Rollup 10 and Microsoft Dynamic CRM Online May Service Update. Note that the WSDL contracts for these releases are unchanged so there is no need to change or recompile source code.
SDK\Client\Images	Updated images provided for use in customizations and applications.
SDK\WSDL\CrmDiscoveryService.wsdl	Added the Discovery Service WSDL file. This is unchanged, but was missing in the previous SDK package.
Accelerators for Microsoft Dynamics CRM 4.0	Added information about the accelerator add-on solutions available for Microsoft Dynamics CRM 4.0.
Advanced Developer Extensions for Microsoft Dynamics CRM 4.0 (Microsoft xRM)	Added a new set of tools that simplifies and accelerates the development of Internet-enabled applications. The conceptual documentation is in this Help file. All other resources can be found in the SDK\Microsoft.xRM folder.
SDK\Microsoft.xRM\
AccessRights, DisplayMasks, EntityItems, MetadataItems, OwnershipTypes,SecurityTypes	Added a note describing the differences between enumeration values using the WSDL compared to the values using the assemblies.
AssignUserRolesRole	Added a reference to a KB article describing an issue using this message.
Comparison between Data Import and Data Migration	Updated the table to include comparison information about Data Import (on-premises), Data Import (online) and Data Migration (on-premises and online). The topic's title in previous releases was "Differences between Data Migration and Data Import".
Creating a Dynamic IFRAME	Removed a note that said that Scripts in Microsoft Dynamics CRM cannot interact with DHTML elements in the source document of an IFRAME. Added code sample about how to do this.
CrmService Sample Code	Added a note about fixing the namespace when you add the helper code file to your project.
Custom Workflow Activity: Date Checker	Added sample code that shows how to compare dates within a workflow, and then send an e-mail.
Customization Best Practices	Added to the Customization Best Practices: How Can I Tell What Customizations Have Been Made to a System?
Data Migration and Import	Updated information about where to find information about Data Migration Manager and Import Data Wizard.
Filtered Views Use Filtered Views Finding the ID Using Filtered Views	Added a note clarifying that filtered views are not accessible for Microsoft Dynamics CRM Online.
Recurrence Pattern in Asynchronous Job Execution	Corrected information about job execution frequency. Removed "MONTHLY" interval and added "YEARLY" interval.
Registering Plug-ins Programmatically	Added information about registering custom workflow activities.
Related Documentation	Added links to relevant white papers.
Server-to-Server Authentication	Added new authentication documentation and sample code for Microsoft Dynamics CRM Online that does not require using certificates. The use of certificates is deprecated.
Server-to-Server Authentication with Impersonation
Walkthrough: Server to Server Authentication with CRM Online
LiveIdTicketManager Classes
SDK\Walkthroughs\Authentication\CS\ ServerToServerNoCerts
SDK\Server\Helpers\CS\CrmOnlineAuth\WLIDTicket.cs
Server-to-Server Authentication using Certificates	Updated topics to indicate that the Windows Live ID team is phasing out the use of certificates.
Server-to-Server Authentication using Certificates with Impersonation
Supported Messages for Plug-ins	Added information about messages that are used by the Web application to retrieve entity data and populate a grid view.
Performance Best Practices
Unsupported Customizations	Added a note about unsupported customizations to the IIS Website.
Upgrade Issues for Workflow and Callouts	Added new information about possible upgrade issues and workarounds for assemblies upgraded from Microsoft Dynamics CRM 3.0.
Using the Web Services within a Custom Workflow Activity	Added a note explaining why it is recommended to use dynamic entity within custom workflow activity code.
Workflow Life Cycle	Added information about avoiding infinite loops in workflows.

Cheers,

This posting is provided "AS IS" with no warranties, and confers no rights.

posted at: 10:41 AM by Aaron Elder

Manual Labor: Best practices and CRM rollups

If things stay on schedule, Microsoft CRM Rollup Update #10 will be out in a little bit and I am reminded that each rollup includes many fixes that are only enabled if you elect to do so.  This is typically done via a registry key tweak, or executable.  There are approximately 40 fixes that need to be enabled manually.  Many of them change default CRM behavior / add features and as such may not be desired by all users.

Of the list, there are some that we enabled by default as part of our Best Practices.

Here is the list:

Improve the performance of an organization import:
http://support.microsoft.com/kb/977867/
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\MSCRM
DWORD: OptimizeOrgImport
Value: 1

Fix Email Templates not displaying formatted HTML correctly:
http://support.microsoft.com/kb/975770/
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\MSCRM
DWORD: EmailTemplateBodyHTMLNoEncode
Value: 1

Async Table gets large over time:
http://support.microsoft.com/kb/974896/
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\MSCRM
DWORD: AsyncRemoveCompletedJobs
Value: 1

Old workflows are not cleaned, resulting in poor performance:
http://support.microsoft.com/kb/968755/
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\MSCRM
DWORD: AsyncRemoveCompletedWorkflows
Value: 1

Line feeds in email templates are not respected:
http://support.microsoft.com/kb/955452
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\MSCRM
DWORD: SDKEmailTemplateHTMLLineBreak
Value: 1

There are many others, that included fixes for non-English installations, time zones, lookup behavior, permission behavior and more.  It is best to review the KB of the latest rollup and read through the list of fixes to see if there is something that might be of use to you.

Why not enable them all by default?  The answer is risk.  Some of these fixes are deep and as such Microsoft only wanted them enabled if customers are experiencing problems.  Microsoft generally codes these fixes such that they only address the specific issue at hand, the benefit is a reduction in regressions.  The items I selected above, seem to be ones that will be a problem for most people, or will be a problem over time and as such, we preemptively “fix” them. 

Cheers,

This posting is provided "AS IS" with no warranties, and confers no rights.

posted at: 12:14 PM by Aaron Elder

We need some MaxLength Inflation: Going beyond the 100k limit

Archive:  April 19, 2010

Microsoft Dynamics CRM is a metadata driven application built on top of Microsoft SQL Server.  In the current version of Microsoft Dynamics CRM, the data types are abstracted from the underlying SQL types.  This was done for many good reasons, most of them now legacy.  One of those abstracted types is the NTEXT field. 

 
In addition to basic data types, Microsoft CRM is the ability to put both UI and platform enforced limits on fields, including NTEXTs.  Unfortunately, the limit for NTEXT fields is restricted to 100,000 characters. 

 

First some history:

The truth of the matter is I am probably the one to blame for this!  You see, back in the day, when the Application Team had to enforce its own metadata (it’s a long story) via Form Properties we added a feature for MaxLength.  At the time, we were in a world of Internet Explorer 6 and the TextArea control did not support a maximum length property.  We therefore had to enforce it via JavaScript.  The problem was that the JavaScript engine had a very inefficient internal routine for counting characters and such, the validation would get slower and slower, the more text the user put in the TextArea control.  In fact, with a few hundred thousand characters the browser would “freeze” for 5 or more seconds while JavaScript tried to count the characters and the form would become unusable.  The decision was made to just limit customizers to some “reasonable” number.  And so it was.  100k was deemed to be good enough and so it was coded and thus it was shipped.  That limit was later passed from the application layer into the platform layer in version 3.0, the same limit exists in version 4.0 and remains today in betas of CRM 5.0.

So what is an XRM developer to do?

Hack #1 (doesn’t work):

The first and most logical attempt I made to hack around this limit was to simply tell the metadata layer that the field could have a max length of more than 100,000 characters.  I did this by updating the MetadataSchema.Attribute table and setting the MaxLength.  Unfortunately, this did not work as the Metadata cache internally validates that the MaxLength for a NTEXT not be larger than the hard-coded “MemoMaxLength” CONST of 100,000!

 

Since that didn’t work I proceeded to look for another solution.

Hack #2 (works):

SQL Script – Run against your ORG_MSCRM database

BEGIN TRAN -- Convert a field to an email body and thus remove the -- limit check in the platform UPDATE MetadataSchema.Attribute SET MaxLength = 1073741823, AttributeLogicalTypeId = 'emailbody' WHERE [name] in ( 'ac_xmlblob', 'ac_htmlblob') COMMIT TRAN

After running the script, run an IISRESET to flush the MD cache.  Hack #2 converts the field into an “email body” (a feature Microsoft should add as a standard feature anyway) and thus the size restriction is ignored at both the Application and Platform levels.  Your mileage may of course vary.

If you are in the CRM 5.0 CTP program, please take a moment and vote up my request to have this limit raised:

https://connect.microsoft.com/site687/feedback/details/552141/ntext-limited-to-100-000-characters

 

Enjoy,

This posting is provided "AS IS" with no warranties, and confers no rights.

posted at: 9:12 PM by Aaron Elder

Convergence Schedule

Archive:  April 25, 2010

Aaron Elder and I will be working at Convergence this year manning the Microsoft CRM xRM booth.  Here is our schedule.  If you are in attendance make sure to stop by and say hi!

• Sunday: 2:30 – 5:00 Aaron and Ross
• Monday: 1:30 – 3:30 Aaron and Ross
• Tuesday: 10:30 – 11:30 Ross
• Tuesday: 11:30 – 1:30 Aaron

posted at: 3:39 PM by Ross Lotharius