Wednesday, May 25, 2016

How to avoid top navigation menu item wraps

To support mobile view, SharePoint 2013 wraps menu item automatically. That's not user friendly to PC users.  As the screenshot below show, 7 items are displayed in 13 rows!

 Most of solutions recommend to overwrite CSS for element "ul.dynamic", such as here and here. But that requires uploading a css file and then add the CSS reference in master page. Is there any other place uses "ul.dynamic"? If yes, then all of them get affected.

Below is a tip so we can avoid modifying master page.

Just a dummy menu item.

Friday, May 20, 2016

"Invalid feature definition" error during SharePoint 2016 configuration

The error message is confusing, but easy to fix the problem: give the SQL instance more RAM.

It needs at least 3.5GB RAM.

Below is what I got when only gave SQL instance 2GB RAM.

05/20/2016 11:45:26  9  ERR                                    An exception of type System.Xml.Schema.XmlSchemaException was thrown.  Additional exception information: Feature definition with Id ca7bd552-10b1-4563-85b9-5ed1d39c962a failed validation, file 'fieldswss4.xml', line 68, character 9:
The 'ListInternal' attribute is not allowed.
System.Xml.Schema.XmlSchemaException: Feature definition with Id ca7bd552-10b1-4563-85b9-5ed1d39c962a failed validation, file 'fieldswss4.xml', line 68, character 9:
The 'ListInternal' attribute is not allowed. ---> System.Xml.Schema.XmlSchemaValidationException: The 'ListInternal' attribute is not allowed.
   --- End of inner exception stack trace ---
   at Microsoft.SharePoint.Administration.SPElementManifest.ElementXmlValidationCallBack(Object sender, ValidationEventArgs evtargs)
   at System.Xml.Schema.XmlSchemaValidator.SendValidationEvent(String code, String arg)
   at System.Xml.Schema.XmlSchemaValidator.ValidateAttribute(String lName, String ns, XmlValueGetter attributeValueGetter, String attributeStringValue, XmlSchemaInfo schemaInfo)
   at System.Xml.Schema.XmlSchemaValidator.ValidateAttribute(String localName, String namespaceUri, XmlValueGetter attributeValue, XmlSchemaInfo schemaInfo)
   at System.Xml.XsdValidatingReader.ValidateAttributes()
   at System.Xml.XsdValidatingReader.ProcessElementEvent()
   at System.Xml.XsdValidatingReader.Read()
   at System.Xml.XmlReader.MoveToContent()
   at System.Xml.XmlReader.IsStartElement()
   at Microsoft.SharePoint.Utilities.SPUtility.XsdValidateXml(XmlTextReader xmlStreamReader, String friendlyName, String pathXsdFile, String tagExpectedRootNode, Int32 desiredPathVersion, ValidationEventHandler xsdValEventHandler)
   at Microsoft.SharePoint.Administration.SPElementManifest.ValidateDefinition(String pathToFeatureAndElementManifestXsdFile)
   at Microsoft.SharePoint.Administration.SPFeatureDefinition.ValidateElementManifestXml(String pathToFeatureAndElementManifestXsdFile)
   at Microsoft.SharePoint.Administration.SPFeatureDefinition.ValidateDefinition(String pathToFeatureAndElementManifestXsdFile)
   at Microsoft.SharePoint.Administration.SPFeatureDefinitionCollection.AddCore(SPFeatureDefinition featdef, SPSite site, String solutionHash, Boolean fForce, Boolean fDoValidation, String pathToFeatureAndElementManifestXsdFile)
   at Microsoft.SharePoint.Administration.SPFarmFactory.EnsureOutOfBoxFeaturesInstalled(SPFarm farm, String[] rgsOutOfBoxFeatures, Int32 compatibilityLevel)
   at Microsoft.SharePoint.Administration.SPFarmFactory.CreateBasicServices(SPFarm farm)
   at Microsoft.SharePoint.Administration.SPFarmFactory.Create()
   at Microsoft.SharePoint.Administration.SPFarm.Create(SqlConnectionStringBuilder configurationDatabase, SqlConnectionStringBuilder administrationContentDatabase, SqlConnectionStringBuilder siteMapDatabase, IdentityType identityType, String farmUser, SecureString farmPassword, SecureString masterPassphrase)
   at Microsoft.SharePoint.PostSetupConfiguration.ConfigurationDatabaseTask.CreateOrConnectConfigDb()
   at Microsoft.SharePoint.PostSetupConfiguration.ConfigurationDatabaseTask.Run()
   at Microsoft.SharePoint.PostSetupConfiguration.TaskThread.ExecuteTask()

Tuesday, May 17, 2016

"DirectoryNotFoundException" from "Restore-SPSite"

There are a few possible causes of this exception, but the error message is confusing.  In my case, it's either about "cannot find directory" or "invalid data".  Below are the ones.

  • Error 1

This one is easy. As the error message shows, the SQL Server Instance needs more RAM.
During my test, it needs at least 3.5GB memory.

  • Error 2

The source and destination farm are with slightly different version.
We can get the version number through PowerShell script:


In the end, I use free tool "HxD Hex Editor" modified the version number in the site collection backup file, to make it identical to the target farm. Of course this should only be applied to non-production environment, and may cause unexpected issues.

Error 1 =========================================

Restore-SPSite : 0x80070003
At E:\EricFang\Scripts\Restore\Restore.intranet2013SIT.80.Main.ps1:30 char:1
+ Restore-SPSite -Identity $SiteUrl -Path $FilePath -DatabaseServer
$DatabaseServe ...
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    + CategoryInfo          : InvalidData: (Microsoft.Share...dletRestoreSite:
   SPCmdletRestoreSite) [Restore-SPSite], DirectoryNotFoundException
    + FullyQualifiedErrorId : Microsoft.SharePoint.PowerShell.SPCmdletRestoreS

.Net SqlClient Data Provider: System.Data.SqlClient.SqlException: There is insufficient system memory in resource pool 'default' to run this query.
   at System.Data.SqlClient.SqlConnection.OnError(SqlException exception, Boolean breakConnection)
   at System.Data.SqlClient.TdsParser.ThrowExceptionAndWarning(TdsParserStateObject stateObj)
   at System.Data.SqlClient.TdsParser.Run(RunBehavior runBehavior, SqlCommand cmdHandler, SqlDataReader dataStream, BulkCopySimpleResultSet bulkCopyHandler, TdsParserStateObject stateObj)
   at System.Data.SqlClient.SqlDataReader.HasMoreRows()
   at System.Data.SqlClient.SqlDataReader.ReadInternal(Boolean setTimeout)
   at System.Data.SqlClient.SqlDataReader.NextResult()
   at Microsoft.ResourceManagement.Data.DataAccess.RetrieveWorkflowDataForHostActivator(Int16 hostId, Int16 pingIntervalSecs, Int32 activeHostedWorkflowDefinitionsSequenceNumber, Int16 workflowControlMessagesMaxPerMinute, Int16 requestRecoveryMaxPerMinute, Int16 requestCleanupMaxPerMinute, Boolean runRequestRecoveryScan, Boolean& doPolicyApplicationDispatch, ReadOnlyCollection`1& activeHostedWorkflowDefinitions, ReadOnlyCollection`1& workflowControlMessages, List`1& requestsToRedispatch)
The Execute method of job definition Microsoft.SharePoint.Administration.SPConfigurationRefreshJobDefinition (ID 20715ec8-eff1-459b-a5a6-f1facd58609e) threw an exception. More information is included below.

There is insufficient system memory in resource pool 'internal' to run this query.
A severe error occurred on the current command.  The results, if any, should be discarded.

Error 2 =========================================

Restore-SPSite : 0x80070003
At E:\EricFang\Scripts\Restore\Restore.intranet2013DEV.80.Main.ps1:30 char:1
+ Restore-SPSite -Identity $SiteUrl -Path $FilePath -DatabaseServer
$DatabaseServe ...
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    + CategoryInfo          : InvalidData: (Microsoft.Share...dletRestoreSite:
   SPCmdletRestoreSite) [Restore-SPSite], DirectoryNotFoundException
    + FullyQualifiedErrorId : Microsoft.SharePoint.PowerShell.SPCmdletRestoreS


Thursday, May 12, 2016

Workflow Manager 1.0 and SharePoint 2013 server object model

(This post is for SharePoint On Premise.)

With the release of Workflow Manager 1.0, developers are encouraged to choose "Client side object model" (CSOM or JSON).  The reason is simple: make SharePoint farm more stable, and easy to migrate to SharePoint Online.

But, what's the drawbacks?

1. Not easy to convert existing SharePoint 2010 workflow assemblies to the new version.

2. Client side object model is not as powerful as Server object model.

3. Learning curve for developers.

4. To improve performance & decrease network latency, may need to consider extra caching module.

All right, if, for some reason, we want to utilize the powerful Workflow Manager 1.0 and SharePoint server object model, what should we do?

One obvious options is to build web services which are hosted on SharePoint server. Then Workflow Manager can call them to get the functionalities.

That's awkard. We have to build two parts for a workflow.

Can we use Server object model in workflow CodeActivity directly, under CodeActivityContext?

The answer is YES. And, it's not too hard.

1. Install and configure Workflow Manager 1.0 on a SharePoint Server.

2. Build the activity.

Need to think about the parameters. I prefer to add 4 parameters to each activity. As the name suggested, the last one will pass in the SPListItem URL and ItemID, so we can get SPWeb and SPListItem easily.

I didn't find a way to put log into workflow history from activity, so all information need to be recorded in parameter "ReturnDetailedInfo".

"ReturnCode" tells the workflow instance whether the activity succeed or fail.
"ReturnValue" return the result.

<Parameter Name="ReturnCode"  Type="System.String, mscorlib" Direction="Out" DesignerType="ParameterNames" Description="Workflow variable output" />
<Parameter Name="ReturnValue"  Type="System.String, mscorlib" Direction="Out" DesignerType="ParameterNames" Description="Workflow variable output" />
<Parameter Name="ReturnDetailedInfo"  Type="System.String, mscorlib" Direction="Out" DesignerType="ParameterNames" Description="Workflow variable output" />
<Parameter Name="CurrentItemUrl" Type="System.String, mscorlib" Direction="In" DesignerType="StringBuilder" />

3. Deploy the workflow (activity) assemblies and solution files to the SharePoint farm and the  workflow server.

Don't forget to create/modify "AllowedTypes.xml" in "C:\Program Files\Workflow Manager\1.0\Workflow\WFWebRoot\bin" and "C:\Program Files\Workflow Manager\1.0\Workflow\Artifacts"

Need to reboot windows service "Workflow Manager Backend" during the deployment.

Reference link: 

4. Make sure workflow service account has "SPDataAccess" rights of SharePoint Configuration database and the relevant content databases

$PrincipleName = "domainname\_WFServices"
Get-SPWebApplication | %{$_.GrantAccessToProcessIdentity($PrincipleName)}

5. Make sure workflow service account is in "WSS_WPG" local windows user group
Reboot the server after adding the service account to "WSS_WPG".

6. Confirm workflow service account has rights on c2WTS

RDP to the workflow server as the service account. The test can be done through c2WTS Tester

7. Restore the site collection to the target SharePoint farm


8. Make sure workflow service account has rights to access target site collection

We can build a X64 windows console program to test it.
RDP to the workflow server as the service account to test it.

9. Clean up Distributed Cache

Reboot all caching servers, or run the script below on all caching servers.
The script below come from here.

Get-SPServiceInstance | Where-Object { $_.TypeName -eq "Distributed Cache" } | Stop-SPServiceInstance -Confirm:$false | Out-Null
While (Get-SPServiceInstance | Where-Object { $_.TypeName -eq "Distributed Cache" -and $_.Status -ne "Disabled" }) {
    Start-Sleep -Seconds 15

Get-SPServiceInstance | Where-Object { $_.TypeName -eq "Distributed Cache" } | Start-SPServiceInstance | Out-Null
While (Get-SPServiceInstance | Where-Object { $_.TypeName -eq "Distributed Cache" -and $_.Status -ne "Online" }) {
    Start-Sleep -Seconds 15

10. Build and publish the workflows through SharePoint Designer;

11. IISReset on Web Front End servers

Below is the sample code.

        protected override void Execute(CodeActivityContext context)
            string strTmp = string.Empty;

            SPList objSPList = null;
            SPListItem objSPListItem = null;

                this._ArgCurrentItemUrl = this.CurrentItemUrl.Get(context);
                Uri UriCurrentItemUrl = new Uri(_ArgCurrentItemUrl);
                NameValueCollection queryStrings = HttpUtility.ParseQueryString(UriCurrentItemUrl.Query);
                if (queryStrings["ID"] == null)
                    strTmp = string.Format("list '{0}' doesn't exist!", _CurrentListName);
                this._ItemID = int.Parse(queryStrings.Get("ID"));

                for (int i = 0; i < UriCurrentItemUrl.Segments.Length; i++)
                    if (UriCurrentItemUrl.Segments[i].Equals(@"Lists/", StringComparison.InvariantCultureIgnoreCase))
                        this._CurrentListName = UriCurrentItemUrl.Segments[i + 1];
                        this._CurrentListName = this._CurrentListName.Substring(0, this._CurrentListName.Length - 1);
                using (SPSite objSPSite = new SPSite(this._ArgCurrentItemUrl))
                    using (SPWeb objSPWeb = objSPSite.OpenWeb())
                        this._CurrentWebUrl = objSPWeb.Url;
                        objSPList = objSPWeb.Lists.TryGetList(this._CurrentListName);
                        if (objSPList == null)
                            objSPList = objSPWeb.Lists[this._CurrentListName];
                            if (objSPList == null)
                                strTmp = string.Format("list '{0}' doesn't exist in web '{1}'!", this._CurrentListName, this._CurrentWebUrl);
                                throw new Exception(strTmp);
                            this._CurrentListName = objSPList.Title;
                        objSPListItem = objSPList.GetItemById(_ItemID);
                        if (objSPListItem == null)
                            strTmp = string.Format("list item '{0}' doesn't exist in list '{1}', in web '{2}'!",
                                this._ItemID, this._CurrentListName, this._CurrentWebUrl);

                this._ArgSQLConnectionString = this.SQLConnectionString.Get(context);
                this._ArgColumnMappingXml = this.ColumnMappingXml.Get(context);
                this._ArgSQLSelectCommand = this.SQLSelectCommand.Get(context);

            catch (Exception ex)
                strTmp = string.Format(@"Execute(), ex.Message={0}", ex.Message);
                strTmp = string.Format(@"Execute(), ex.StackTrace={0}", ex.StackTrace);
                this.ReturnCode.Set(context, this._ArgReturnCode);
                this.ReturnValue.Set(context, this._ArgReturnValue);
                this.ReturnDetailedInfo.Set(context, this._ArgReturnDetailedInfo);

Monday, May 9, 2016

The future of SharePoint development

SharePoint 2016 is finally released in GA. The astonishing thing is "SharePoint Framework".

It seems Microsoft stopped pushing "Addins (APPS)", and invented another development model. I completely understand why some SharePoint experts not happy about that.

I am OK about it, because I didn't spend much time on "Addins" model, and no plan to study "SharePoint Framework".  :-)

I don't know why Microsoft keep building these complex development model. This is just like how Microsoft Vista handle the security problem: pop up a prompt window, let users decide whether accept the potential threat. If the user choose the wrong one......Microsoft doesn't need to be responsible for that, right?

Same problem with full trust farm solution. There are a lot of poor quality fully trusted code, and Microsoft's solution is simple: ban all of them. If users want the functionality, they have to run it in other application servers. If the other application servers crashed because of the poor quality code......Microsoft doesn't need to be responsible for that, right?

Client-side application is very important. I have no doubt about that. But, use it to replace server-side application? In many cases, that make things very complicated. And complexity make poor quality code much worse! And, it decreases the productivity of development, A LOT! (Think about latency, caching, stability, network bandwidth, etc.)

Kicking out the problem, is not really a solution.

What's the real solution?  I have some ideas here.

Split the development into two categories: server-side and client-side.

For "server-side", put each fully trusted solution into a dedicated "container". If the code crashed, the container will crash and reset, and it doesn't affect the other parts of SharePoint. Many of the SharePoint OOTB service instances should also run in "container".

For "client-side", it's actually JavaScript running in something like Content Editor Web Part, with the help of OOTB toolbox/library. (Users can build TypeScript then turn it into JavaScript, of course)

Can we get that in SharePoint 2019? Let's wait and see.

Any comments are welcome.

PS: What is going to replace InfoPath?!

[update, 2016-05-20]

Don't get me wrong. SharePoint 2016 is good, just not so friendly to developers.  :-)

Friday, April 8, 2016

What is "DevOps" anyway?

When I first time saw the word "DevOps", I thought it's quite straightforward: a system administrator with software development skill. That's why it is called "DevOps", not "OpsDev", isn't it?

No. Based on wikipedia, it's the opposite.

It seems that the "DevOps" concept is created purely for software development improvements.

All right. Then what should we call it as "a system administrator with software development skill"? Don't know. And unfortunately, I am one of that kind of guys.


( This post is inspired by Managed services killed DevOps )

Wednesday, February 17, 2016

How to generate new Visual Studio Solution from existing solution?

Most of posts on Internet suggest to create a new solution and feature in Visual Studio, then copy/attach the files to the new project. (such as this one)

I find another opotion.

After copying the whole Visual Studio solution/project to new folder, there are 4 parts we need to replace.

1. DLL new key
2. Solution ID
3. Feature ID
4. Change the Solution name, Feature name, Assembly name (and webpart / workflow action name if needed).

The first two steps are easy. Open the project in Visual Studio, then do the change straight away.

But the third step is tricky: we cannot just go to "Feature Designer" and replace the GUID in "Feature Id" property.  This ID is used in 7 different places and we cannot change them all in Visual Studio.

I did that through the famous freeware "Notepad++", go to "Search", then "Find in Files"

Below is the search result after clicking "Find All" button.  We can click "Replace in Files" button to replace all old Feature Id with new GUID.

Step 4 is similar to step 3.

That's it.

Now we can compile and publish the new solution file (.wsp) in Visual Studio, and then deploy it SharePoint.


To use new solution and feature to replace old solution and feature in farm, don't forget to uninstall the old feature and solution first, if you don't want to change the solution name or feature name.

Below is the PowerShell script.

Get-SPSite -Limit ALL | ForEach-Object {Disable-SPFeature -identity "FeatureName" -URL $_.URL -confirm:$false -ErrorAction SilentlyContinue}

Get-SPFeature | ?{$_.DisplayName -eq "FeatureName"} | Uninstall-SPFeature -confirm:$false -force

Get-SPSolution | ?{$_.DisplayName -like "SolutionName"} | Remove-SPSolution -confirm:$false -force

You can go to SharePoint system folder "C:\Program Files\Common Files\Microsoft Shared\Web Server Extensions" on all SharePoint servers, then search for the feature name, to confirm the feature is removed from farm completely.