Migrating MOSS Databases from Testing to Production

The following procedures assumes that two WFE exist: one for serving content and the other an indexing server. This procedure should successfully migrate from testing to production using database backup and restore for the MOSS applications.

1. In SQL Server Enterprise Manager, backup all Office SharePoint Server 2007 related databases from the previous database server.

2. On the front-end web server, disconnect MOSS from the far using the Sharepoint Products and Technologies Configuration Wizard.

3. In SQL Server Enterprise Manager, restore all Office SharePoint Server 2007 related databases except the configuration database and the central administrator content database and the search databases to the production database server.

Central Administration Configuration (Front-End Web Server and Index Server)

1. On the front-end web server, run the Sharepoint Products and Technologies Configuration Wizard.

a. Do not connect to an existing server farm. Enter the database and server name information for the production database server.

b. After the configuration wizard has completed, do not accept the offer to create a shared service provider or portal site. Instead create it manually.

c. Next, disconnect the index server from the farm using the same method as above.

d. Connect to the new pilot site from the index server.

e. Turn on the indexing service for the index server.

2. Restore the Shared Service Provider.

a. From the Application Management page click Create or configure this farm's shared services.

b. Click Restore SSP.

c. Add the necessary configuration information for the Shared Service Provider web application, My Site web application, Shared Service Provider database name and search database name.

d. Click OK.

e. Start the search crawl from the Search Setting page in the Shared Service Provider administrator console.

Event-driven updating of the BDC index.

A typical business scenario for using business data search would be one where users need to access data on new customers provided by the company’s Siebel implementation. Though the index will provide a snapshot of the customers that existed at the time of the last crawl, new customers might have been entered since that time that might be of high interest. Unfortunately, no built-in event raising mechanism exists to signal the index that new records have been entered and an incremental crawl is needed. One workaround to this issue, and one, moreover that accords with the way crawler is intended to work, is to create an IDEnumerator method that could detect changed records and kick off an incremental crawl. From a performance viewpoint, it would be undesirable to have a crawl initiated whenever there was a new record, but this impact could be minimized if only changed records were indexed at convenient intervals.

Features in the Chinks

SharePoint 2007 allows developers to insert new UI and functionality as features in an easy-to-deploy and granular way. See the following post to learn about how useful features can be created and deployed: http://blogs.msdn.com/sharepoint/archive/2007/04/10/debugger-feature-for-sharepoint.aspx

Features in the Chinks

SharePoint 2007 allows developers to insert new UI and functionality as features in an easy-to-deploy and granular way. See the following post to learn about how useful features can be created and deployed: http://blogs.msdn.com/sharepoint/archive/2007/04/10/debugger-feature-for-sharepoint.aspx

MOSS 2007 Gaps

The following are some of the major gaps I have discovered so far in planning large ECM projects with MOSS 2007: 1) First and foremost, the lack of email indexing and manipulation 2) The lack of built-in realistic document lifecycles that define a set of stages in a document's life. MOSS has a few basic workflows, but these must be considerably extended to implement realistic lifecycles. For instance, it is often necessary to require digital signatures at various points in a document's life - while MOSS supports signatures through Office, requiring a signature can only be done programmatically. Moreover, there is no server-side support for this. 3) A built-in commenting/annotating mechanism for documents in review 4) No virtual documents - a built-in ability to assemble several subdocuments into a single virtual document. 5) Linked documents - a way to show which documents are associated with a particular document. As time allows, I'll start to contribute ideas and workarounds for these limitations - and I welcome your contributions!

A Little Philosophy

SharePoint 2007 is major step in the right direction, but those of us who have to implement sites for a living need to know the exact distance between the rubber and the road. A few added millimeters here and there adds up to thousands of extra hours figuring out the SharePoint object model or an extra week or two of vacation. It matters. The purpose of this blog is explore how SharePoint works as a content management tool for those of us who create enterprise content management systems for a living. As much as possible we will try to focus on issues and gaps and how to cross them. Comments to the point are always welcome, as are suggestions about how to spend the extra vacation time.