Microsoft has recently release search connectors to IBM FileNet and EMC Documentum for MOSS 2007 and MMS 2008. The links to download the connectors are: Enterprise Search Indexing Connector 2008 for IBM FileNet and Enterprise Search Indexing Connector 2008 for EMC Documentum.
In addition to this Microsoft also release tutorials on how to install the connectors. The tutorials shows the install for small server environments. The tutorials can be viewed at: ESIC2008 for FileNet tutorial and ESIC2008 for FileNet tutorial.
The reason why I limited the title for this blog to the FileNet connector is because I have had some experience in the last few weeks in installing and using the connector in an large corporate in South Africa. I want to share some of the experiences I had with installing and configuring the connector.
The connector uses the FileNet Content Engine Client to connect to FileNet. We could install the client on the indexing server with the SharePoint administrator, but we could not configure it to connect to FileNet. We eventually used the FileNet administrator account to install and configure the FileNet Content Engine Client on the SharePoint Servers.
Our original environment consisted out of a 32 bit indexer and 64 bit web front ends. We could install the the connector on the indexer, but not the WFE's. The error on the WFE was "This product requires Windows SharePoint Services 3.0". The actual error is that the connector is not supported on 64 bit servers. You need to have the indexer and the WFE servers in 32 bit node. The reason for this limitation is that the FileNet client is only available in 32bit and you need 32bit SharePoint to use the 32bit FileNet client. The indexer uses the client to index the documents and the WFE servers uses the client to open and view the document in FileNet.
Once we managed to install the connector we had problems configuring it. We got access denied errors. Eventually after a lot of trail and error we managed to configure the connector by using an account that is both a FileNet administrator and and a SharePoint farm administrator.
We hoped that that was the end of the problems, but when we tried to crawl the FileNet object stores we got access denied again. We eventually traced all these access denied error back to the fact that the search and crawl user accounts tried to open DCOM connections from the SharePoint server to the FileNet server. We bypassed this by making the search and the crawl accounts local administrators on the FileNet server. This did the trick and we are indexing FileNet now. You also need to make sure that Network DCom is enabled on both the FileNet and SharePoint servers.
Making these accounts administrators in FileNet is far from ideal. I have theory on how to reduce the necessary permissions to do the crawls. This involves giving the crawl and search accounts launch and activation permission on the CE DCom object on the FileNet server. We are currently in a testing freeze and I will update the blog if I have proof that this works.
Happy Searching