1

I am trying to find out better way deal with large volume of records within single table. Given that a table have 50M records and records are inserted about ~1.5M records a day.

Which options are available to avoid further performance struggling after month/year?

Please point me to the right direction. Currently I am thinking about having separate table (or maybe views) with different sets of records (like deleted/old/new) and feed EF with appropriate table/view in appropriate scenario (just for example if it simple user he might not need to work with all records like deleted or 2 years old and that might give some performance issues)

Vladimirs
  • 8,232
  • 4
  • 43
  • 79
  • Well it will really depend on your data policy, but it already would seem a good idea to have an "archive" table (or database) where to move the "old" records. what can be considered an old record is up to your business to determinate... – Laurent S. Oct 23 '13 at 12:19
  • How many rows/entities are you planning to load at any one time? Assuming that is reasonably low then the amount of rows in the table is pretty much irrelevant as long as you find them with an index. – Martin Smith Oct 23 '13 at 12:34
  • Bartdude, thank you for reply. I am just thinking is there any already implemented mechanisms for archiving data with EF or it should be simply stored procedure and task? – Vladimirs Oct 23 '13 at 12:45
  • Martin Smith, thank you for your comment. Sorry, I might not clearly understood your question. In DB records will be loaded one by one, but than they will be displayed using filters and paging and also some statistics will be collected daily/weekly. So I am afraid that at some point index will not deal good enough with large data volumes. – Vladimirs Oct 23 '13 at 12:52

0 Answers0