Great to hear! That barely scratched the surface...this is everything: www.red-gate.com/simple-talk/books/sql-server-execution-plans-third-edition-by-grant-fritchey/
Excellent Explanation. What if there is no clustered index or Primary column. Will it cost more to search another column which is LastAccessDate which is not included in the non clustered index and also query uses non clustered index and is it look for look up column by that time ?
No Clustered Index makes the table a HEAP, which will almost always be harder to search for extra columns that are not in the NC index it is using now. www.brentozar.com/blitz/heaps-tables-without-primary-key-clustered-index/
Brilliant video :) Quick question, what would be the best way to optimise a query that uses wildcards? E.g. select * from table where name like 'Kevin%'; Would you have to reevaluate the query to see if bringing back all the Kevins is the right approach?
Thanks Ryan! You would want a non-clustered index on [name] for a 'Kevin%'...and you might see an index scan instead of a seek, depending on how many variations of Kevin there are. If you try '%Kevin%', your index will no be used. That leading wildcard changes everything. Always test and evaluate periodically. Missing and Unused index queries are my favorite ways to evaluate my index strategy
There is no way to tell...it totally depends on what the procedure is doing, and how the Query Optimizer decide it is going to execute the code in the proc, get the data, allocate memory, etc.
I decided to dive deep into query optimization / index thing today, let's see how it goes. Thanks for the video.
There's a LOT of layers to SQL Server performance tuning...have fun!
Thanks Kevin for the simplicity and examples!!!
You're very welcome!
Many thanks for posting. Surprised the likes and views are so low but for me it's been very useful .
This is the best video I have seen . Easy to understand . Just started to learn and was lost . This helped me a lot . Thank you 🙏
You're very welcome!
This is Pure Gold (as Manosphere Highlights Daily channel would say).
Hope it helps you :)
Right to left and from top to bottom.
Thanks kevin!
The best explanation I have seen! I can't thank you enough for it.
This is exactly what I have been working on the last 2 weeks! Nice breakdown!
Working on such a video, or learning to read Plans?
@@Kevin3NF Im going through our software and optimizing queries. Especially missing indexes.
@@simusprime Awesome! Good luck...and don't over-index! Run an "Unused Index" query as well to look for possible dead weight.
Finally I understood execution plans. Thanks!
Great to hear!
That barely scratched the surface...this is everything:
www.red-gate.com/simple-talk/books/sql-server-execution-plans-third-edition-by-grant-fritchey/
wow, you explained everything i was looking for a while , Thanks a lot
Glad it was helpful!
Very clearly put, thanks. Looking for more of your examples and explanations
Excellent Explanation.
What if there is no clustered index or Primary column. Will it cost more to search another column which is LastAccessDate which is not included in the non clustered index and also query uses non clustered index and is it look for look up column by that time ?
No Clustered Index makes the table a HEAP, which will almost always be harder to search for extra columns that are not in the NC index it is using now.
www.brentozar.com/blitz/heaps-tables-without-primary-key-clustered-index/
Thank you Kevin for your valuable time.
omg you saved my energy and everything thanks alot
Glad it helped!
Great example with clear explanation. Definitely a subscribe!
Thanks Kevin... Can u post performance tuning in wild card searches....
Very well explained with examples. Thank you, Kevin!
Very welcome!
Very nicely explained, thanks.
You are awesome!!! looking for this from a long time.. Simple and Great explanation
Well done, very clear explanation
Thank you for the good explanation and the examples!
Brilliant video :) Quick question, what would be the best way to optimise a query that uses wildcards? E.g. select * from table where name like 'Kevin%';
Would you have to reevaluate the query to see if bringing back all the Kevins is the right approach?
Thanks Ryan! You would want a non-clustered index on [name] for a 'Kevin%'...and you might see an index scan instead of a seek, depending on how many variations of Kevin there are. If you try '%Kevin%', your index will no be used. That leading wildcard changes everything. Always test and evaluate periodically. Missing and Unused index queries are my favorite ways to evaluate my index strategy
Kevin Good job. Please upload more videos
Would have been more helpful to put that link to your other video in your summary where it could have been clickable...
Thanks Jerry...I have added the link to the description of this one
great stuff, thanks Kev!
This is exactly what I was looking for.
Awesome! You should get Grant Fritchey's newest edition of his query plan book...
Are the cost of each operation accumulated from right to left or are they independent ?
Each operator has its own cost, and the entire query is shown as a cost in the query stats DMVs:
www.scarydba.com/2017/02/20/estimated-costs-queries/
Good explanation..thanks
Glad you liked it
Awesome video.
Well done, I actually learned something!
Thank you. Good Job.
best explanation
if cost =20% for stored proedure this good or not
There is no way to tell...it totally depends on what the procedure is doing, and how the Query Optimizer decide it is going to execute the code in the proc, get the data, allocate memory, etc.
@@Kevin3NF i mean excution plan the estimaton cost is 20%
Thanks sir, I'm from India
hi
10:27
This is very helpful, Thank you.