Your Browser is Out of Date

Nytro.ai uses technology that works best in other browsers.
For a full experience use one of the browsers below
Reference Architecture Guide—Consolidate and Simplify Mixed Database Workloads
Executive Summary Solution Architecture and Design Overview Validation Test Goals, Configuration, and Use Cases Validation Test Results
Overview
Average CPU utilization
TPM
NOPM
Storage IOPS
Storage latency
Throughput
Combined performance of IOPS, latency, and throughput
Data Domain Backup and Recovery Solution Conclusions from Test Results References Solution Hardware and Software Design and Configuration Details

Throughput

Throughput

  • In addition to running OLTP workloads, we ran a DSS workload using the HammerDB TPC-H–like benchmark.

    Note: The “TPC-H-like” test means that the results are not certified.

    The DSS workload test simulates ad-hoc queries that are designed to assist the business with decision analysis. In addition, the test also simulates concurrent data modifications, in which multiple sets of data are modified in parallel. The queries are complex, reflecting that the database must join and aggregate (filter or group) large sets of data to assist the business with decisions analysis.

    We used a Scale Factor (SF) of 1,000 for the SQL Server and 3,000 for Oracle DSS testing. SF defines the database size. For example, an SF of 1 indicates 1 GB. Because we used an SF of 1,000 in our test, the database size was 1,000 GB for SQL and 3,000 GB for Oracle. The SF also defines the minimum number of query streams. For example, it specifies a minimum of seven query streams for an SF of 1,000 and a minimum of eight query streams for an SF of 3,000. A query stream is a set of queries that must be executed serially, one after another. In the case of each SQL Server database, we ran only one query stream and executed 17 of the 22 queries. For each Oracle database we ran one query stream that executed a subset of queries.

    We did not gather formal TPC-H–like metrics such as Throughput@Size because the validation test scope focused only on storage throughput.

    Our focus in running the DSS workload was to generate throughput on the PowerMax 2000 array. Throughput is the amount of sustained data that is transferred as supported by the infrastructure.

    The following table shows the throughput test results that came from the PowerMax storage report for the data LUNs of SQL Server and Oracle:

    Table 18.        Throughput test results

    DSS database

    OLTP and DSS in parallel

    OLTP, DSS, and snapshot OLTP in parallel

    IOPS

    Host MB/s

    IOPS

    Host MB/s

    SQL Server 1

    6,783

    631

    6,375

    644

    SQL Server 2

    6,720

    625

    7,587

    714

    Oracle

    13,842

    1,731

    13,688

    1,712

    Both SQL Server and Oracle databases showed stable or improved throughput and IOPS as the workload increased and became more complex. Hence, our tests prove that the throughput level improves with the increased size and complexity of the workload.