White Papers

38 BP1062 | Best Practices for Decision Support Systems with Microsoft SQL Server 2012 using Dell EqualLogic PS Series
Storage Arrays
(Six users, one database, two
arrays)
•
6 users (132 queries) from Benchmark Factory
• Database volumes: Refer to Figure 16
Data Warehouse Workload Parameters
Database load & Workload
generation
TPC-H from Benchmark Factory
Database size 300 scale (~620 GB including data and indexes)
Number for users/streams 6 (As per TPC-H standard, the minimum required streams/users to be
run for a 300 scale Database is 6)
Queries
22 TPC-H queries per user
SQL Server Memory (Max
memory allocation) in GB
117.7 GB
SQL Server parameters
-E, -T1117 and lock pages in memory
CPU
4* Intel® Xeon® Processor E5-4620 @2.20 GHz,8 Cores per socket
This test checked the performance improvement while adding a second array to the existing pool. The
database was spread across both arrays while six users with 132 queries were constantly run on one array.
The test was repeated after scaling to two arrays to gather the performance differences. Figure 17 shows
the read throughput and query completion time improvements after the second array was added.
Read throughput, query completion time- performance comparison Figure 17
The storage throughput increased by 24% for the constant user load by scaling the array. This
improvement was due to the increase in storage processing resources such as number of disk drives,
895
1116
433
347
0
50
100
150
200
250
300
350
400
450
500
0
300
600
900
1200
1xarray 2xarrays
Time (minute)
Throughput (MB/sec)
Average read throughput and completion time
Average Read Throughput
Query Completion Time