Designing a Microsoft SharePoint 2010 Infrastructure Vol 1 part 48. This course is intended for IT Professionals who use Microsoft SharePoint 2010 in a team-based, medium-sized to large environment. While they may have implemented a SharePoint deployment, they have limited experience in designing a SharePoint infrastructure. They likely work as a senior administrator who acts as a technical lead over a team of administrators. Members of this audience should have at least 6 months experience with SharePoint 2010. | Designing an Enterprise Search Strategy 9-29 Improve query performance. Plan capacity for FAST search. MCT USE ONLY. STUDENT USE PROHIBITED 9-30 Designing a Microsoft SharePoint 2010 Infrastructure Capacity Sizing for SharePoint Search Key Points To plan capacity for SharePoint search you must recognize each of the components that require storage the storage medium and the sizing requirement. Most search sizing calculations are a factor of the size of your corpus which is the total body of content that you plan to crawl and index. Use the figures and calculations below as an estimate or starting point for the environment that you are planning. Due to the nature of various content sources and types of content you may discover that you require more or less than your estimate. Index Requirements Index partitions are stored on query servers. To determine the total index size use the following formula TotallndexSize CorpusSize x This estimates the total index. To calculate the storage requirements for each query server you must know the number of index partitions and mirrors that you require. From a capacity perspective an index partition should hold no more than MCT USE ONLY. STUDENT USE PROHIBITED Designing an Enterprise Search Strategy 9-31 10 million items. SharePoint automatically divides the index evenly among all active nonmirrored index partitions. You can calculate the storage requirement for a partition or mirror by using the following formula QueryComponentSize TotallndexSize v NumberOfPartitions For each query server you must also allow additional disk space for index updates and index merging. To calculate disk requirements use the following formula for each query server IndexStorage QueryComponentSize x 3 x NumberOfQueryComponents In addition you should plan enough RAM for the query server to hold a third of all active index partitions. For example if the query server holds two active partitions each of which is 10 GB the server should have at least
