At home for SQLSaturday #311
After having a very intensive training at a customer site in Copenhagen, I am returning back to Sofia just a night before the third Bulgarian SQL Saturday event.
The training in Denmark was very exciting for me! I had a group of about 18 people, all of them curious, polite and eager to understand and get everything I had to talk to them. I did a customized training from my SQL Master Academy program for DBAs and for Devs, We had some important topics on SQL Server configurations, backups and major maintenance tasks, indexes and statistics, as well as some basics concepts of monitoring and troubleshooting, They learned how to check and configure properly their SQL Server Instances, to build good maintenance tasks and to perform check in case there is a problem with their Server or database. We then had a Dev part, where they were even more curious and passionate about query tuning, execution plans, cardinality errors and different code optimization techniques.
I love those moments in my trainings when I see attendees’ look as if there were a lamp that suddenly has turned on in their heads and they become so happy. I also love the moments when they start asking questions which answers come two slides after that. You feel that you align the training with their experience, expectations and way of thinking. It is just a great feeling!
I hope to come back again soon and do some SQL stuff again with those great people.
Now its time for the third SQL Saturday in BG. Love to present at BG Conferences, that’s my audience and I always prepare something interesting and special for them. Cant wait to meet you all again!
My sessions this year will be:
Let those statistics be with you! – a demo-only session that explores statistics and tries to give you all the understanding you need on that topic
A deep dive into SQL Server Plan Cache Management – yeat another deep dive session for this year 🙂
The sessions summary:
Let those statistics be with you!
There are a lot of questions about statistics that I receive in my trainings, projects and during my sessions: Why I have to rebuild statistics, if my auto-update stats option in ON. Why I still have bad plans or high CPU usage? Should I disable it or use full scan instead? We usually rely on auto-update stats database options and sometimes perform a regular stats updates additionally in order to be sure that the QP has an up to date info about our table data. But there are some specific cases when those thinks just don’t help and even recompiling don’t boost the performance of a query. Let me list and explain for you some statistics updates exceptions cases and you will understand how to proper manage your statistics in order to gain max from them.
A deep dive into SQL Server Plan Cache Management
As an experienced DBA you have probably asked yourself questions like how the memory is consumed, is my system well-tuned, and is my memory configuration well defined. In order to understanding such processes as local and global Memory pressure, caches concurrency, adding and removing plans from cache we will go through the internal organization of the plan cache, the metadata available, how SQL Server finds a plan in cache, plan cache sizing, and the plan eviction policy. We will point out the important cache consumers that affect your workload and how to get metadata about them, what are the thresholds in Cache Size Management. At the end you will understand how to monitor the memory consumers in your SQL Server, what are wait types that we need to look for, how to understand if your system is well-tuned and the memory configuration is well defined
Session Level: Advanced