This documentation relates to an earlier version of the SharePoint Connector.
View

Unknown macro: {spacejump}

or visit the current documentation home.

There are two options for integrated search with the SharePoint Connector.

  • Integrated Index Search
  • Federated Search

Both options involve having SharePoint show the search results from both SharePoint and Confluence (and potentially other sources). Confluence can redirect users to the SharePoint search results page using the SharePoint Theme.

This page should help you decide between these two options, but if you are still having difficulty deciding you can try both at the same time. They should both be able to operate together.

Integrated Index Search

This is the original option provided with the SharePoint Connector. Using this option involves SharePoint crawling and indexing Confluence content then showing Confluence search results along with other search results. All search results are intermixed and sorted by relevance or modified date.

This option is more involved and therefore may be more tricky to configure properly, but it is supported with the SharePoint Connector.

Prerequisites

TODO

Show screenshot thumbnail.

Federated Search

This is using new capability provided with Search Server 2008 which has been rolled into MOSS 2007 as part of the July 2008 Infrastructure Update. Using this option involves creating a federated location to Confluence using an unsupported OpenSearch plug-in from Lunatech. Confluence is not crawled or indexed by SharePoint. Instead, SharePoint passes off the query to Confluence and shows the results in a Federated Search web part. These results are shown separately from other search results and are therefore not intermixed by relevance or modified date.

This option is less involved, but it is not supported with the SharePoint Connector.

Prerequisites

TODO

Show screenshot thumbnail.

Decision Matrix

Feature

Integrated Index Search

Federated Search

Description

Support

(tick)

(error)

Integrated Index Search is supported as part of the SharePoint Connector. Federated Search is part of SharePoint, but the OpenSearch plug-in from Lunatech is not supported.

Intermixed Heterogeneous Results

(tick)

(error)

Integrated Index Search shows intermixed heterogeneous search results from Confluence along with SharePoint and potentially other sources all sorted by relevance or modified date. Federated Search shows search results separately in another web part.

Some might consider separating the search results a good thing. This could be more of a preference than a pro/con.

Security Trimming

(tick)

(error)

Integrated Index Search supports security trimming the Confluence search results to only show URLs that are accessible by the current user. Federated search does not currently allow per-user security trimming, but it does allow you to trim a single account (some "lowest common denominator account").

If both Confluence and SharePoint were using Kerberos, per-user security trimming for Federate Search would be possible, but this is not a supported configuration with the SharePoint Connector. This limitation is mentioned near the bottom of Security for Federation. While there is currently no provided solution, see CSI-410 for the feature request details.

Troubleshooting

(error)

(tick)

Federated Search is a simpler solution from a technical standpoint (there are less moving parts) and therefore may be easier to get working properly. Integrated Index Search may be harder to get to work properly requiring some fine tuning or troubleshooting.

Advanced/Property Search within Confluence

(error)

?

Federated Search may allow for advanced search capability such as searching on a label or space. Integrated Index Search does not currently support this (see CSI-238 for this feature request).

Need to determine this for Federated Search.

Multiple Confluence Installation Search

(error)

(tick)

Federated Search allows showing search results from multiple Confluence installations; the results for each instance of Confluence show in a separate web part (not intermixed). Integrated Index Search does not currently support this (see CSI-109 for this feature request).

Single Index Backup

(tick)

(error)

Integrated Index Search stores the index from SharePoint and Confluence in one place allowing for a single backup operation of the crawled content. Federated Search has the SharePoint crawled content in SharePoint database and index server, but the Confluence crawled content in the Confluence database.

Single Confluence Crawl

(error)

(tick)

Federated Search allows for Confluence content to only be crawled by Confluence. Integrated Index Search incurs extra resources and drive/backup space since Confluence is being crawled by both SharePoint and Confluence.

Index/Crawl Scheduling

(tick)

?

Integrated Index Search allows SharePoint to determine how often to index and crawl. Federated Search leaves this up to Confluence/Lucene.

Is this configurable with Confluence/Lucene?

  • No labels