Voici comment repérer et résoudre les problèmes de performances dans votre équipe d’administrateurs de base de données.

In the real world, we do have some monitoring tools (e.g., Grafana, Zabix,...) to track the overall database performance such as CPU, Memory, Disk,... Once we notice an unusual alert, we will have to look deeper into all the queries that the database has run in that time slot. 1. If it is currently occurring, then look at sp_WhoIsActive to see whether some processes were blocked or not. 2. If it is already done, then look at the history data from Query Store (should enabled from SQL Server 2016+) for the "Top Resource Consuming Queries" ||"Regressed Queries"||"Query Wait Statistic". This could be a good starting point for figuring out the problem.

Identifying performance issue is some time a critical tasks. But if proper process is in place along with the coordination from different teams then it helps in identification and resolution of the performance issues more quickly. From DBA perspective analyzing query and related objects should come first along with resource utilization time taken or duration of the query. Benchmarking each of these parameters and making sure DBAs are aware of them for quick identification and resolution.

After pointing out the real performance issue, there are several steps we need to implement. 1. If some queries were blocked resulting in resource contention, then figure out which are blocking the others now and why. It could be some lack of indexes results in a Full Table Scan, or maybe some modifications on the same rows at 1 time. 2. If some queries take so long to be executed or the number of executions is numerous leading to some CPU burning, then let's have a look at the execution plan to see the bottleneck. In most the cases, let's start with the highest operator's cost. 3. Don't forget to check which wait has occurred the most in that timeslot, it could be due to some network issue, disk IO latency, Memory Inefficient, etc.

From the current or ongoing issues there is always a potential to be proactive in next time. DBA and lead make sure that issue has been discussed within team. With.open ended discussion, new ideas or way of resolution might pop up which can be useful in resolving the issue more efficiently and effectively.

Implementing fixes for performance issues in a team of database administrators requires a structured and timely approach. Once root causes are identified, develop and prioritize a clear action plan. Assign specific tasks to team members based on their expertise and ensure they have the necessary resources. Communicate the changes effectively, providing context on the issue and the expected outcomes. Monitor the impact of these fixes through performance metrics and feedback, adjusting as needed to ensure the solutions are effective. Regular follow-ups ensure that the fixes are sustained and integrated into standard practices, fostering continuous improvement.

Once query resolution is implemented. It's always good to perform impact analysis for next 1 or 2 weeks and see the issue not reoccurring. It's important to get the report of impact analysis I.e. results before and after resolution. Report should include resource utilization, query duration, etc.

After solving the problem, I often provide training to my colleagues. That helps us have a quality team and ready to support each other. This training also needs to take place regularly to improve the professional knowledge of everyone in the team.

1-Collaboration and Communication: - Regular Meetings: Hold regular team meetings to discuss ongoing issues, share knowledge, and brainstorm solutions. - Cross-Functional Collaboration: Work closely with other teams (e.g., development, operations) to ensure a holistic approach to performance management. 2-Process Improvement: - Standard Operating Procedures (SOPs): Develop and enforce SOPs for common tasks and performance tuning. - Continuous Improvement: Regularly review and update processes based on feedback and performance data. 3-Training and Development: - Skill Enhancement: Provide training on the latest database technologies and performance tuning techniques. - Mentorship: Pair less experienced DBAs with senior team members.