From 51b0124880b91c273a134f905b58fb8556bb0ae8 Mon Sep 17 00:00:00 2001 From: Jos de Bruijn Date: Mon, 12 Sep 2016 13:05:58 -0700 Subject: [PATCH] scale factor --- samples/features/in-memory/benchmark-order-processing/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/samples/features/in-memory/benchmark-order-processing/README.md b/samples/features/in-memory/benchmark-order-processing/README.md index f26ee553..20632cf9 100644 --- a/samples/features/in-memory/benchmark-order-processing/README.md +++ b/samples/features/in-memory/benchmark-order-processing/README.md @@ -17,7 +17,7 @@ Sample order processing workload that can be used for benchmarking transactional 1. Create the database, tables, and stored procedures using the T-SQL scripts in the corresponding subfolders. - - The hash indexes are sized for a 24GB database, with bucket_count double the row count. For larger database and memory size, adjust the bucket counts accordingly. + - The hash indexes are sized for a scale factor of 100, which translates to a 24GB database, with bucket_count double the row count. For larger database and memory size, adjust the bucket counts accordingly. - The max bucket_count in SQL Server 2016 is 1 billion. It is OK to have a higher row count. The benchmark performs well with bucket_count of 1 billion and row counts of 5 billion. - There are plans to publish scripts for initial populate of the tables. Timeline is TBD. - Scripts are also provided for corresponding disk-based tables and traditional stored procedures, to compare performance between disk-based and memory-optimized tables.