ApexSQL Audit performance usage - Testing case using Trace auditing technology
説明
ApexSQL Audit performance usage when using Trace auditing technology
対策
DISCLAIMER: The following sample represents an insight of resource utilization during overload cases, insinuated with a huge auditing data processing using ApexSQL Audit. Testing that was performed below provides metrics obtained by the usage of ApexSQL Audit and SQL Server within a specific testing environment with no extra processes running except for SQL engine and auditing. The obtained results can be considered as a reference guide in building the expectations on how ApexSQL Audit can impact the given environment. Testing environment specification Audited instance machine specification:
OS: Windows Server 2019 Datacenter 10.0 <X64> (Build 17763: ) (Hypervisor)
SQL Server version: Microsoft SQL Server 2017 (RTM-GDR) (KB4505224) - 14.0.2027.2 (X64) Jun 15 2019 00:26:19 Developer Edition (64-bit)
Performance testing on the audited instance machine The following data provides an insight of the testing result in a specific environment and displays overall hardware consumption in specific scenarios, showing the graph results in which a dedicated auditing agent is referred to as ApexSQL Audit processor Distributed, which is a core component for data processing and collecting on the audited instance machines.
Testing case using Trace auditing technology - low impact on resource utilization Input parameters for testing:
Testing duration: 10 minutes
Auditing technology: Trace
Auditing configuration: GDPR
SQL query average length: 80 chars long
Number of executed SQL events: almost 253.000 which stands for 422 transactions per second (TPS)
Test results: Graph below showing CPU utilization during test run with Trace auditing technology: Graph below showing memory utilization during test run with Trace auditing technology: Conducted testing shows a very low impact on CPU and memory utilization regarding ApexSQL Audit Distributed processor when the test SQL query was executed simultaneously within 10 minutes. Workload during test run per total SQL Server resources (test SQL query vs ApexSQL Audit SQL query): Comparison of applications that executed SQL queries during test run per total SQL Server resources:
Testing case using Trace auditing technology - high impact on resource utilization Input parameters for testing:
Testing duration: 6 minutes
Auditing technology: Trace
Auditing configuration: GDPR
SQL query average length: 80 chars long
Number of executed SQL events: ¬620.000 which stands for 1.722 transactions per second (TPS)
Test results: Graph below showing CPU utilization during test run with Trace auditing technology: Graph below showing memory utilization during test run with Trace: Conducted testing shows the limit, where higher impact on CPU can be expected and slightly higher memory utilization regarding ApexSQL Audit Distributed processor when test SQL query was executed simultaneously from 6 applicatios within 6 minutes. Workload during test run per total SQL Server resources (test SQL query vs ApexSQL Audit SQL query): Comparison of applications that executed SQL queries during test run per total SQL Server resources:
