Full-Text Indexing

Indexing is the process in West km that prepares documents for retrieval. Taken as a whole, it includes the indexing, synchronization, and full-text indexing processes. This topic explains full-text indexing.

Full-text indexes are used to retrieve your organization's documents when a West km member at your organization performs a term search on your organization's documents. The full-text indexing capability in West km is provided by the full-text searching service and the search component, and is based on the HTML renderings of indexed documents.

When full-text indexes are created, all documents are processed. When full-text indexes are updated, processing occurs for new, modified, and deleted documents since the last full-text indexing.

While the search component is full-text indexing, searches can simultaneously be performed in West km. Note, however, that search performance may be negatively affected. For this reason, it is recommended that you schedule or launch full-text indexing outside regular business hours. In addition, you should not simultaneously run full-text indexing and other West km indexing and synchronization processes.

You can schedule or manually launch full-text indexing. In particular, you can

• Schedule Full-Text Indexing
• Manually Launch Full-Text Indexing
• Manually Launch Full-Text Indexing by Storage Collection
• Modify and Administer Storage Collections

[top of page]

Scheduling Full-Text Indexing

Full-text indexing is scheduled by defining indexing time windows in the full-text searching service.

You will likely need to define multiple time windows to account for different storage collections (e.g., documents vs. clauses), different frequencies for the same storage collection (e.g., weekdays vs. weekends), and different full-text indexing processes (i.e., indexing vs. reclaiming processes).

To schedule full-text indexing, complete these steps:

1. Point to the Setup link, and then click HTML Storage.
2. Click Details and Indexing in the left frame.
3. Click Full Text Indexing Scheduling in the right frame.
4. Click Add New FTI Scheduling Window.
5. Under Add FTI Scheduling window, do the following and click Save.
   
   

   • Collection Name: Select the storage collection for which you want to schedule indexing. To schedule all storage collections, select All Collections.

   This storage collection:	Contains:
   WestKM	HTML renderings of files
   WestKM_Doc	HTML renderings of documents
   WestKM_Clause	HTML renderings of clauses
   WestKM_Def	HTML renderings of defined terms
   WestKM_Scan	a document representing the KeySearch hierarchy

   • Start time and End time: Type the start time and the end time for the window in HH:MM:SS format (e.g., type 23:00:00 for 10:00 p.m. and type 06:00:00 for 6:00 a.m.).

     A window's start time always occurs on the day specified by the frequency. The end time may occur the day afterward, depending on whether the start time is set before or after the end time.

     Note: The start time and end time should not be the same, or the window is considered to be open for zero time instead of 24 hours. For example, use a start time of 00:00:00 and an end time of 23:59:59 if you want a fully open window.

   • Frequency Type: Select the frequency for which you'd like the window to be in effect. Weekdays refer to Monday through Friday. Weekends refer to Saturday and Sunday.

   • Service Type: Select the full-text indexing process that the windows controls.

     FullTextIndexProcess creates and updates index files for searching. For example, for a storage collection you may want the full-text indexing process to run daily.

     ReclaimProcess optimizes index files for searching. For storage collections that are being updated, the reclaim process is recommended weekly.

     Note: In general, the ReclaimProcess window should be open about three hours. It can overlap the FullTextIndexProcess window, but the reclaim process takes priority if they are scheduled at the same time.

6. Repeat steps 4-5 for each full-text indexing window you want to schedule.

For example, if you want full-text indexing to run twice each day, define two windows with a frequency type of Daily, but with different start and end times.

If you want full-text indexing to run during one window on weekdays and another on weekends, define two windows and select a frequency type of Weekdays for one window and Weekends for the other, with different start and end times for each.

Full-text indexing will start automatically within 10 minutes after reaching the start of the window. At the close of the window, the search component will finish full-text indexing the currently processing documents (which takes approximately 5–10 minutes) before stopping.

To remove a full-text indexing window, click Remove next to the listing.

To change an existing full-text indexing window, remove the window and then add another window with the settings you want.

[top of page]

Manually Launching Full-Text Indexing

To manually create or update the full-text indexes at your organization, complete these steps:

1. Point to the Setup link, and then click System Options.
2. Click Indexing Settings.
3. If you want to fully re-create index files (and not just update existing index files), click Update Collection XML in the Full Text Indexing section.

4. To start full-text indexing updates immediately for all storage collections (regardless of the scheduled time window), click Rebuild Index Now under Full Text Indexing.

   Indexing starts according to the scheduled time window for full-text indexing.

[top of page]

Manually Launching Full-Text Indexing by Storage Collection

You can manually update the full-text indexes for individual storage collections. This feature is useful if re-indexing one particular collection is more urgent than another or if you only want one storage collection to be re-indexed.

To manually update the full-text indexes for individual storage collections, complete these steps:

1. Point to the Setup link, and then click HTML Storage.
2. In the left frame, click Details and Indexing.
3. In the right frame, click the Full Text Index Now icon next to the storage collection you want to index.

Indexing starts according to the scheduled time window for full-text indexing.

[top of page]

Modify and Administer Storage Collections

You can modify the settings for a storage collection. With the search component, you can also get document counts and retry failed documents.

To modify or administer a storage collection,

1. Point to the Setup link, and then click HTML Storage.
2. In the left frame, click Details and Indexing.
3. In the right frame, click the Details icon next to the storage collection you want to modify or administer.

To modify a storage collection, click Modify, make the changes you want under Collection Details, and click Save.

To view the number of documents in the storage collection that are indexed in the search component, click Run Search under Get Collection Document Count.

Actions

• To retry full-text indexing of failed documents in the storage collection, click Retry Failed Docs. When the full-text indexing window opens, the search component will attempt to reprocess the failed documents.

  Note: The search component will automatically retry failed documents according to the NovusIndexRetryAttempts property in Thomson.FTSearch.Server.exe.config. The Retry Failed Docs function could be useful, for example, if the search component was down and you want to manually launch reprocessing attempts.

• The Update functions (e.g., Update Stop Words, Update Non-Indexed Words, Update Equivalency List, Update Fields) are typically necessary only after an upgrade, but will have no ill effect if used at other times.
• To update the storage collection’s XML files, which contain high-level data for the storage collection, click Update Collection XML. This function is used only for troubleshooting purposes, but will have no ill effect if used at other times.
• To manually start the ReclaimProcess for the storage collection, regardless of the scheduled time window, click Reclaim Merge. This process optimizes index files for searching.
• Do not use the Destroy Content function without the assistance of a West km representative. This function deletes all search component index files and document content from the search component databases. It should be used only in certain troubleshooting situations.

[top of page]