When a data warehouse is being updated by periodical ETL load, there could be significant performance impact on the end users' reports while the underlying tables are touched. The data becomes volatile and the query results could be different between updates. In my experience, The relevant technologies that I could think of are MS Snapshot and Data Replication. For platform neutral techniques, using backup/restore and views/ synonyms could somehow switch underlying data dynamically without obvious data outage. What's the best practice in resolving the data availability issues, especially when the DW is big and near real time (eg. hourly frequency).
- Posts : 528
Join date : 2010-05-07
Location : Brisbane, Australia
Permissions in this forum:You cannot reply to topics in this forum