Microsoft Office 2007 System Search Training, Day 1
As you may have read already, I am in DC for some way cool Enterprise Search training. It's 3 days of intense search training centered around SharePoint's search capabilities as well as many of the integration points available to us with some of the office tools.
We started the day off looking at some of the differences between the various versions of SharePoint. Most of day 1 is intended to be light material with the real meat of the training taking place late today and tomorrow and Thursday. Without going into too much detail, I would say the biggest takeaway for me here was that the ability to search the BDC (Business Data Catalog) does not exist in the Standard Search SKU. This kind of makes sense tough considering the BDC itself is only made available with the Enterprise SKU of Office SharePoint Server.
We talked about many of the improvements made in this version with respect to the end user experience like Relevance and Line of Business (LOB) and People search. We also discussed some of the administrative improvements that were made in areas like Indexing Scopes, Properties, Customization, Query Logging and Performance. This all sounds very dry on paper, I know, but it really is very interesting.
We also discussed Security with respect to query time trimming, pluggable authentication (I am skeptical here), some of the new features and Reporting.
At this point we dove right into Search Architecture and Topology. With respect to architecture we talked about the query engine, the indexing engine, protocol handlers, ifilters, ranking, keywords, best bets, schemas, scopes, crawl log, content sources, word breakers, stemmers and content indexes. Lots and lots of talking points here, this section took an hour and a half. The topology discussion was great. Apparently there is one unsupported topology. I was under the impression that any topology was supported.
We talked about how WSS search and MOSS search are the same, but different. Supported topologies for both were discussed as well. The concept of the Shared Service Provider was discussed at length and a conversation about having multiple ones ensued. There is a debate around when to use multiple SSP's vs. when to use multiple farms. Remember, there is only one physical index per SSP and that seems to be the crux of the confusion.
We then dove into exactly how the crawl process works as well as how the query process works. It was very enlightening.
That's it for day 1, we were supposed to have a module on relevancy, but we ran over, so we will be starting there on day 2.