You can follow the below steps to configure query suggestions

1. Verify that the user account that is performing this procedure is a site owner on the Search Center site.

2. On the Search Center site home page, click Site Actions, and then click Edit page.

3. On the Search Box Web Part menu, click Edit Web Part.

4. Expand the Query Suggestions node.

5. Make sure that the Show query suggestions box is selected. If you want to turn off query suggestions, clear the Show query suggestions check box.

Query suggestions are automatically created over a period of time based on how often the users search for a specific query within a certain timeframe.

You can also manually add the query suggestion by using the PowerShell command. When you use Windows PowerShell to manually add query suggestions, you immediately provide the benefits of query suggestions to users of a newly installed search system.

If query suggestion contains special characters, then you can use the “Windows PowerShell ISE” to manage it as given in the below example.

Go to SharePoint Server --> Add the feature “Windows PowerShell ISE” in windows Server by using Server Manager.

Open “Windows PowerShell ISE” and Add snapin:  Add-PSSnapin Microsoft.Sharepoint.PowerShell

Use the below command to manually add the query suggestion “Parent�S Resource Guide”

$searchapp = Get-SPEnterpriseSearchServiceApplication "<QuerySearchServiceApplicationName>"

Get-SPEnterpriseSearchQuerySuggestionCandidates -SearchApplication $searchapp

New-SPEnterpriseSearchLanguageResourcePhrase -SearchApplication $searchapp -Language EN-US -Type QuerySuggestionAlwaysSuggest -Name "Parent�S Resource Guide"

Start-SPTimerJob -Identity "prepare query suggestions"

You can see the query suggestion immediately after running the timer job in the Search Center page as given below.

To remove the manually added query suggestion use the below command.

Remove-SPEnterpriseSearchLanguageResourcePhrase -SearchApplication $searchapp -Language En-Us -Type QuerySuggestionAlwaysSuggest -Identity "Parent�S Resource Guide"

Start-SPTimerJob -Identity "prepare query suggestions"

Then check the behavior. This should remove the query suggestion "Parent�S Resource Guide".

If query suggestions are automatically created, then you have to block it and you cannot remove it.

If let’s say query suggestion "Parent�S Resource Guide" is automatically created, then you have to block it by using below command

$searchapp = Get-SPEnterpriseSearchServiceApplication "<QuerySearchServiceApplicationName>"

Get-SPEnterpriseSearchQuerySuggestionCandidates -SearchApplication $searchapp

New-SPEnterpriseSearchLanguageResourcePhrase -SearchApplication $SearchApp -Language en-us –Type QuerySuggestionBlockList -Name "Parent�S Resource Guide"

Start-SPTimerJob -Identity “Prepare query suggestions”

Please note that query suggestion keyword is “case sensitive”.

Published by - Prasad Joshi

Index profile will check for the compatibility before the update and if the profile is incompatible then there can be chance of navigator fields failed to update.

You can see the below events in the bliss logs.

INFO      : bliss@yourservername: systemmsg: Compatibility check: Attribute "display" for string-navigator with name="your old navigator name" in result-specification has different value "New navigator name".QR-server must be restarted.

INFO      : bliss@yourservername: systemmsg: Compatibility check: Index-profiles are incompatible

INFO      : bliss@yourservername: systemmsg: Running index-profile update without doing compatibility check

INFO      : bliss@yourservername: systemmsg: New index-profile update finished successfully

If you check the output in FAST ESP QR-Server by accessing https://yourESPserver:15100 then you will still see the old navigator name. You can also verify the output for different views exist in your FAST ESP environment.

To get the updated navigator name try to clear the QR-server cache and to refresh and redeploy views. This can be done with the following steps:

1. Stop the qrserver processes on all qrserver nodes.

2. On all qrserver nodes, rename or move the $FASTSEARCH/var/qrserver/ directory to remove the 

              cache files.

1. Restart the qrserver processes on all qrserver nodes.

2. Run the following command to refresh the views:

              view-admin –m refresh

1. Run the following command to deploy the views:

              view-admin –m deploy –a

If that doesn’t resolve the issue then follow the below steps to manually update the navigator name.

1. Export the view in xml format which requires new navigator name update using the below command.

   view-admin -m export -V myview

2. Manually edit the xml file, update the new navigator name and save it.

3. Import the view using the below command.

   view-admin -m import -V teamsd1.xml -o –d

Note: Make sure you have taken the backup of exported view.xml file before editing and importing it to the ESP server. Even a minor index-profile update should be viewed as a serious procedure. This is not a simple task that lends itself to being performed every day on a multi-node cluster.

Reference: https://support.microsoft.com/en-us/kb/2587443

Published by - Prasad Joshi

In the below example I am trying to crawl the below documents from my SharePoint site.

After the crawl, in the Search Center I can see the results with hit highlighted summary fetching the document Title property.

You might have noticed that the document name which we crawled is “ieeecopyrightform.doc” and in the search results highlighted summary is displaying the “Title” property.

If you want hit highlighted summary to fetch the “document name” instead of “Title” property, then you have to add the ows_linkfilename(Text) and FileExtension(Text) in the Title Managed property of the FAST Query SSA.

Go to Central Administration > fast query SSA > FAST Search Administration > Managed Properties and edit the “Title” property.

Add the Mapping ows_linkfilename(Text) and FileExtension(Text). Move these crawled properties to the Top. Click ok.

Initiate a full crawl and after the crawl in the search center you can see that hit highlighted summary fetches the “document name” as shown below.

Please note that this requires full crawl to be initiated on the content source.

Published by - Prasad Joshi  

Steps

• Now go to site setting and then Search settings -> Configure Search Navigation 

• Add link -> Select the new created page “SSA2” with title “SSA2”

• Edit the existing “Everything” link as “SSA1” 

• Click Ok to save and apply 

• Now consuming Search center site should look like below 

• Edit the search center site and edit the search result web-part query to display the results from result source created “SSA2” 

• Save the webpart and then the page. 

• Clicking the Search vertical links namely “SSA1” and “SSA2” will display results from corresponding search service application 

Note:- You can modify the result source or the search result web part query to display the results differently for both the Search service applications in the consuming search center site

Published by - Sujeet Sharma

You might have observed that sometime command "spreladmin showstatus" shows status as Running and Progress is stuck @ x.x %. Also on the FAST search index drive "FASTSearch\data\sprel\worker\build_proc" consumes more disk space.

One of the common causes of the SPRel data directory filling up is due to jobs failing to complete and thus never cleaning up properly. This can happen due to connection issues. Check the sprel_proc.log log forany potential connection issues.

To resolve the issue try below steps on the FAST search servers.

1. Run the following command to stop the sprel process:

    nctrl stop sprel.

2. Clean the data/sprel directory EXCEPT the config directory. 

3. Run the following command to start the sprel process again:

    nctrl start sprel. 

4. Confirm everything is ok by running.

    spreladmin ShowStatus 

5. Clean the directory C:\FASTSearch\components\resourcestore\generic\clickthrough (data pushed by the SP extraction job). 

6. SharePoint Server: On the SharePoint node, it is possible to trigger the ClickThrough extraction job through Windows PowerShell. This example assumes SSA name is “Search Service Application”.

    $job = Get-SPEnterpriseSearchExtendedClickThroughExtractorJobDefinition -SearchApplication "Search Service Application"  

    $job.ResetLastExtractionDates()

    $job.Update()

    $job.RunNow() 

7. In the FAST Server: 

1. Check the directory C:\FASTSearch\components\resourcestore\generic\clickthrough see if data are coming in. 

       2.  After Job’s completion, start the processing

            spreladmin StartProcessing 

       3.  Regularly check the status

            spreladmin ShowStatus 

       4.  Once completed, status shows ready doing nothing : FASTSearch\bin> .\spreladmin.exe showstatus

             Processing:   
                           
             Status           : ready 
                           
             Makefile         :  nothing

If the above steps didn't resolve the issue then follow the below steps.

1)     Pause any running crawls and wait a few minutes to make sure there aren’t any documents left in the document processing pipeline. 

2)     Stop the sprel process on the FAST Search server in your FAST farm :               
         nctrl stop sprel       

3)     Rename the below folder       
        a)     Rename folder sprel to sprel.bak       
        b)     Create two new folder: sprel       
        c)     Copy  SPRelConfig.xml from sprel.bak\config to the new sprel\config folder     

4)     Start the processes stopped in step 2.

        nctrl start sprel  

5)     SPRel job runs every night at 04:00 AM, it’s best to let it be and check the status of it the next day. 

6)     Resume crawl. 

Check the status of it on the next day by executing the command spreladmin.exe showstatus.

Published by - Prasad Joshi

robots META tags can be embedded within any HTML web page, within the "head" section. For a META tag of name "robots", the content value will indicate the actions to take, or to avoid. While a page without such tags will be parsed to find new URIs before being indexed, the possible settings can prevent either or both of these actions by a crawler.

In the following example, the page is being effectively blocked from further processing by any crawler that downloads it.

Example:

 <html>

<head>

<meta name="robots" content="noindex,nofollow">

<meta name="description" content="This page ....">

<title>...</title>

</head>

.

.

.

<html>

Robots META Tags Settings 

In the page, if you have this setting : <meta name="robots" content="noindex,nofollow" /> the page will not be crawled. You can also cross verify the fetch logs to see if we are getting the noindex, nofollow in it.

You can bypass this settings in ESP by updating Check meta robots value to No. Cross verify that check_meta_robots option is successfully updated to “no” in the config file.

Restart the crawler service and crawl the website again.

Published by - Prasad Joshi

You might have observed that sometime crawling legacy office documents (.doc, .xls, .ppt etc..) will fail with below warning message in the crawl log. 

"The FAST Search backend reported warnings when processing the item. Document conversion failed: LoadIFilter() failed: Bad extension for file (0x800401e6)."

This is an issue with legacy 97-2003 office documents and the Microsoft iFilter 2.0 which fails to convert the files to html in preparation for the indexing process. 

Unfortunately fix is not available for the Microsoft iFilter 2.0 but a workaround is possible with a configuration change in the FAST environment which will ensure that the legacy file types such as “.doc”, “.ppt” and “.xls” are converted using the SearchExportConverter and not the iFilterConverter. The SearchExportConverter supports legacy formats and handles their conversion better than the iFilterConverter. The iFilterConverter will still be used to convert new versions of office documents such as “.docx”, “.pptx”, “.xlsx” etc..

To resolve the issue, try to execute below steps on each server in your FAST environment. I would recommend backing up the files before making any changes to it.

1)     Open E:\FASTSearch\etc\config_data\DocumentProcessor\optionalprocessing.xml, change  “active” to yes, for “SearchExportConverter”, save the file.

2)     Open E:\FASTSearch\etc\formatdetector\converter_rules.xml on the FAST server.

3)     In the converter_rules.xml, go to the node on line 673, and comment the first “doc” type. e.g  "<!--<ext>.doc</ext-->". Now we have allowed the “.doc” type of extension to be in another document processing stage “SearchExportConverter”. Repeat this for the .xls and .ppt document types.

4)     Open E:\FASTSearch\etc\pipelineconfig.xml, and locate the following pipelines:

              <pipeline name="Attachments (webcluster)" default="0">

              <pipeline name="Office14 (webcluster)" default="1">

5)     Within the two pipeline sections, swap the order of “IFilterConverter” and “SearchExportConverter”, to make “SearchExportConverter” in front of “IFilterConverter” in  both pipelines.

6)      In FAST PowerShell, run “nctrl reloadcfg”

7)      In FAST PowerShell, run “nctrl restart procserver_X”, X applies to all the procserver numbers you have in the system, and there could be several procservers.

Once you have followed all the steps on all the servers, you will need to run a full crawl.

Published by - Prasad Joshi

You might have observed that many times executing the command "logtransformer\bin\launch.bat downloader" in FAST ESP 5.3 to download the query logs will simply get stuck or throws an error "Premature end of file" without downloading any query logs.

Detailed error stack:  

E:\esp\components\logtransformer\bin>launch.bat downloader

[ jacorb.home unset! Will use '.' ]

[ File .\jacorb.properties for configuration jacorb not found ]

[ configuration jacorb loaded from classpath]

In CSInfoInt ctor

Log4J configured from: file:/E:/esp/etc/log4j.properties

Log file downloader 5.1.7 build 5.3.4.6

Started download.

[Fatal Error] :1:1: Premature end of file.

Finished downloading files.

java.lang.reflect.InvocationTargetException

        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)

        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.

java:39)

        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAcces

sorImpl.java:25)

        at java.lang.reflect.Method.invoke(Method.java:597)

        at org.apache.commons.launcher.ChildMain.run(ChildMain.java:228)

Caused by: com.thoughtworks.xstream.io.StreamException: Premature end of file.

        at com.thoughtworks.xstream.io.xml.DomDriver.createReader(Unknown Source

)

        at com.thoughtworks.xstream.XStream.fromXML(Unknown Source)

        at com.fastsearch.esp.logtransformer.datastore.meta.dao.DaoSupport.loadO

bjectFromXml(DaoSupport.java:114)

        at com.fastsearch.esp.logtransformer.datastore.meta.dao.MetaDataDAOImpl.

getLogFileProducingSource(MetaDataDAOImpl.java:127)

        at com.fastsearch.esp.logtransformer.datastore.FdsLogStoreManagerImpl.ge

tOrCreateLogfileProducingSource(FdsLogStoreManagerImpl.java:368)

        at com.fastsearch.esp.logtransformer.harvest.download.DownLoader.fetchDa

taSource(DownLoader.java:290)

        at com.fastsearch.esp.logtransformer.harvest.download.DownLoader.updateL

istOfActiveQueryServers(DownLoader.java:227)

        at com.fastsearch.esp.logtransformer.harvest.download.DownLoader.runBatc

h(DownLoader.java:104)

        at com.fastsearch.esp.logtransformer.harvest.download.DownLoader.main(Do

wnLoader.java:316)

        ... 5 more

Cause : 

This issue occurs if there are any corruption of the file 

$FASTSEARCH/components/logtransformer/data/datastore/datasources.xml 

Resolution :

 1. Stop the logtransformer

                a. nctrl stop logtransformer

                b. Please verify that the logtransformer process is no longer running in Task Manager.  If necessary please kill the process. 

2. Truncate the query statistics table:

                a. %FASTSEARCH%\rdbms\bin\psql -p16070 -dvespa -Ufast (password is fast)

                b. truncate queryStatistics; truncate periodProperty;

                c. \q 

3. Take a backup and then delete the below files 

                a)  Check that all META files in

                    $FASTSEARCH/components/logtransformer/data/download have a size greater than zero, and remove any that has 0 KB.

        b)  Delete $FASTSEARCH/components/logtransformer/data/datastore/datasources.xml

        c)  Delete $FASTSEARCH/components/logtransformer/data/datastore/logmerger-statistics.xml

        d)  Delete $FASTSEARCH/components/logtransformer/data/datastore/querylog/crateindex.xml 

4. Start the logtransformer

                a. nctrl start logtransformer

The downloader should start downloading the querylogs within 15 minutes. The query report generator should start running at 5 past the hour. If for some reason this doesn’t happen you can run the commands manually: 

cd %FASTSEARCH%\components\logtransformer\bin

launch.bat downloader

launch.bat logmerger 

Once those commands are finished, run the following: 

cd %FASTSEARCH%\esp4j\bin

launch.cmd queryreportgenerator -s "2014-09-............" -e "2014-10............." -D -L

Set the start date to the date you want the statistics to first be generated for and the end date to the current time.

Published by - Prasad Joshi