SQL Governor version 16 is now officially available for our partners and customers. With these new optimizations and enhancements for FinOps and the sustainability of your Microsoft data platform, you can improve your TCO by up to 50%.
New features
SQL Governor V16 includes enhanced, AI-based capacity planning features and several improvements in reporting side.
AI-assisted capacity planning automation
SQL Governor V16 offers a powerful auto-adjust feature to automate the selection of the most relevant target SQL Server host, VM and Managed Instance size for SQL Server and database right-sizing and consolidation. You don’t have to try out the different alternatives on your target platform anymore – instead, let the AI resolve the optimal configuration for your setup – with just one click! This is fast and simple and improves time-to-value, even if you have several hundreds of source SQL Servers.
VMware API and SQL Server metadata automation
SQL Governor now supports VMware vCenter integration. SQL Governor Proxy service queries changes in the SQL Server hosts and their virtual machine setups from VMWare vCenter on an hourly basis and then relays this information to the SQL Governor server for further processing.
SQL Server metadata automation is based on VMware vCenter queries. It automatically updates the SQL Server configurations regarding the hardware setup and VM configurations, whenever they change in the system, with a maximum one-hour lag. This makes keeping up the SQL Server configuration metadata effortless in SQL Governor when it comes to VMware-hosted virtual machines, even if there are hundreds or more SQL Server VMs.
CMDB API
CMDB API offers a configurable set of source queries to fetch the SQL Server, instance and database level properties into the 3rd party CMDB software such as ServiceNow. This helps in automating SQL Server related metadata in CMDB software.
Note! The CMDB API requires case-by-case customization, which must be agreed upon separately with the SQL Governor team.
Improved data platform support in capacity planning
SQL Governor supports now not only physical hardware, VMware virtualization, Azure IaaS / PaaS and AWS IaaS, but also Nutanix platforms: In capacity planning, we have improved Nutanix cluster target host / VM support up to Gen9 hardware benchmarks and host settings, such as CVM reserve (CPU, RAM). Also, the benchmark repository is regularly refreshed with the latest CPU benchmarks.
Create your own VMware or Nutanix clusters and plan optimal VM-, instance- and database distribution:
- Save overall host and VM capacity significantly
- Improve TCO
- Create more sustainable data platforms
- Avoid CPU contention
Enhanced reports
Capacity planning
Capacity planning reports have experienced a facelift: The main capacity planning reports are now combined and simplified in terms of visualization. The reports have more information such as the latency prediction spread chart, which helps in addressing the latency issues.
These are the remaining reports on capacity planning:
Database latency prediction spread chart:
Monitoring
Some monitoring reports have improved: For example, the environment load report shows not only the CPU contention but also the CPU count over time and capacity planning max CPU in addition to the monitoring max CPU for each SQL Server. The difference between these values is that capacity planning related max CPU is aggregated (avg) on consecutive 5-minute intervals from 1-minute delta data and is calculated on hourly basis, whichever 5-minute interval is highest for that hour. In contrast, SQL Server max CPU sampling interval is 1 minute, so the highest minute is a representative of the monitoring max CPU value. That’s why it is always higher or same as the capacity planning CPU. This is because in capacity planning the aggregated values are more meaningful in terms of the sufficiency of the computational resources (only 5 min CPU peaks are meaningful).
CPU contention over time:
Max server monitoring CPU usage with capacity planning max CPU usage:
Deprecated features
Short-term pattern recognition
The short-term pattern recognition feature has been deprecated and will be most likely to be substituted with enhanced long-term pattern recognition feature in the future. The challenge with short-term pattern recognition was twofold:
- It was built to forecast only very short-term behavioral patterns (5 to 60 minutes) for the SQL Server and instance performance counters and therefore did not add much value to the more relevant, longer-term monitoring, like anomaly regressions are.
- Because the feature was machine learning -based, the algorithm training sometimes lasted from several hours to days in very large environments with high CPU consumption.
SQL Governor roadmap – what’s next
SQL Governor Azure SaaS
SQL Governor V17 is planned to be provided as a multi-tenant SaaS service from Azure. It should scale significantly better than SQL Governor V16. Planned new features for the V17 are such as:
- Nutanix virtualization API
- New intuitive dashboards
- Enhanced T-SQL query tracing
- Improved user experience
Target release date for the V17 is during Q2/2025.
Note! We are still going to support on-premises installation of the SQL Governor software.
For more information, please contact:
Jani K. Savolainen
jani.savolainen@sqlgovernor.com
CEO & Chairman
