SQL Server Dev ToolKit
3 likes678 views
This document is a comprehensive toolkit for SQL Server developers, detailing essential recovery plans, important programs, and useful scripts for database management. It emphasizes the need for good practices around index management and efficient coding while advocating for continuous learning and professional development. Additionally, it provides resources for further training and career advancement within the SQL community.
![Index Interrogation (2a)
--[Reads] = SUM(DUS.user_seeks + DUS.user_scans + DUS.user_lookups),
[Seeks]
= DUS.user_seeks,
[Scans]
= DUS.user_scans,
[Lookups] = DUS.user_lookups,
[Writes] = DUS.user_updates,
!
[Reads Per Write] = CONVERT(DECIMAL(10,2), MAX(CASE WHEN DUS.user_updates < 1 THEN 100.00
ELSE 1.00 * (DUS.user_seeks + DUS.user_scans + DUS.user_lookups) / DUS.user_updates END)),
--[Last Read] = CONVERT(CHAR(10), (SELECT MAX(v) FROM (VALUES (DUS.last_user_seek),
(DUS.last_user_scan), (DUS.last_user_lookup)) AS value(v)), 101),
[Last Seek] = DUS.last_user_seek,
[Last Scan] = DUS.last_user_scan,
[Last Lookup] = DUS.last_user_lookup
Reads Per Write can help identify indexes that have not been used since the last restart or aren't helping to improve
performance. For every 1 write to the index, you want to see as many reads as possible.
!
Indexes with a reads_per_write score of 1 means that for every 1 write, the index is also used 1 time to help with
performance. Ideally, you want to see scores much higher than that.
!
Consider dropping indexes with a reads_per_write score under 0 and strongly consider dropping ones with scores under .1.
!
This isn't a hard-and-fast rule: you may have an index that's used once per month for a single report, but that report is run by
the CEO and he wants it instantaneously. Before dropping indexes, know what they're used for or make sure alternate
indexes exist.](https://image.slidesharecdn.com/sqlserverdevtoolkitpdf-140914095136-phpapp02/85/SQL-Server-Dev-ToolKit-26-320.jpg)
SQL Server Dev ToolKit
- 1. A SQL Server Developer’s Toolkit
- 2. • Kirsten Benzel, Database Engineer at SurveyMonkey • Twitter: CausalDensity • Email: [email protected] • Blog: http://kSQL.net • SQL Server Magazine: http://sqlmag.com/author/kirsten-benzel
- 3. What is Causal Density? From a book review titled Controlling The Uncontrollable on Jim Manzi’s Uncontrolled, Kevin Drum writes: ! “If you're studying the orbit of a planet, you can pretty much assume there's only one important cause of the planet's movement: gravity. Causal density is low … in human affairs, there are lots of causes of everything, there are causes of the causes, and the causes often interact in complex ways. Causal density is very high, which means it's very hard to make sure you've accounted for everything.” ! (Emphasis mine, motherjones.com/kevin-drum/2012/05/controlling-uncontrollable)
- 4. Your computer just puked on it’s shoes. HAPPY MONDAY
- 5. Have a Recovery Plan 1. List of all your installed programs and license keys. Include versions and installation steps for difficult installs. 2. Keep your web browser Bookmarks backed up. 3. Keep a list of your mapped drives - the full extension, not the truncated version you see in My Computer (Windows Users). 4. Keep important folders backed up: SSMS Templates, my Desktop folders, my Powershell script folder, and the databases for KeePass and Stickies. 5. Write down the location of your SSMS Templates. 6. Write a .bat file that dumps all the above from your workstation to a external drive.
- 6. Essential Programs SQL Server Management Studio and Local SQL Instance Visual Studio (may need different versions for SSIS packages) Anything other than Notepad: TextPad, WordPad, Notepad++ Adept SQL Diff adeptsql.com/download.htm 7-Zip 7-zip.org/download.html Stickies zhornsoftware.co.uk/stickies/download.html Softerra LDAP Browser ldapadministrator.com/download.htm KeePass keepass.info/download.html
- 7. SSMS Mastery
- 8. ALT + F1
- 11. Alter SSMS StartUp Behavior
- 16. Oh, Crap! (1)
- 17. Oh, Crap! (2)
- 18. Oh, Crap! (3)
- 19. SQL Job Info
- 20. Space Used Interrogation (1)
- 21. Space Used Interrogation (2)
- 22. Space Used Interrogation (3)
- 23. Space Used Interrogation (4) (Space Used by File)
- 26. Index Interrogation (2a) --[Reads] = SUM(DUS.user_seeks + DUS.user_scans + DUS.user_lookups), [Seeks] = DUS.user_seeks, [Scans] = DUS.user_scans, [Lookups] = DUS.user_lookups, [Writes] = DUS.user_updates, ! [Reads Per Write] = CONVERT(DECIMAL(10,2), MAX(CASE WHEN DUS.user_updates < 1 THEN 100.00 ELSE 1.00 * (DUS.user_seeks + DUS.user_scans + DUS.user_lookups) / DUS.user_updates END)), --[Last Read] = CONVERT(CHAR(10), (SELECT MAX(v) FROM (VALUES (DUS.last_user_seek), (DUS.last_user_scan), (DUS.last_user_lookup)) AS value(v)), 101), [Last Seek] = DUS.last_user_seek, [Last Scan] = DUS.last_user_scan, [Last Lookup] = DUS.last_user_lookup Reads Per Write can help identify indexes that have not been used since the last restart or aren't helping to improve performance. For every 1 write to the index, you want to see as many reads as possible. ! Indexes with a reads_per_write score of 1 means that for every 1 write, the index is also used 1 time to help with performance. Ideally, you want to see scores much higher than that. ! Consider dropping indexes with a reads_per_write score under 0 and strongly consider dropping ones with scores under .1. ! This isn't a hard-and-fast rule: you may have an index that's used once per month for a single report, but that report is run by the CEO and he wants it instantaneously. Before dropping indexes, know what they're used for or make sure alternate indexes exist.
- 28. ! Demo Scripts github.com/kbenzel/sql-scripts/
- 30. Gracefully deal with change (and different personalities)
- 31. Gracefully deal with change (and different personalities) Be an advocate for using the best tool for the job.
- 32. Legacy code and less-than-optimal current code. Don’t disparage. Teach.
- 33. Write elegant, good code. Legacy code and less-than-optimal current code. Don’t disparage. Teach.
- 34. Compartmentalize. Have home and work balance.
- 35. Deal with Make the most of criticism. Compartmentalize. Have home and work balance.
- 37. Advancing Your Career credit buzzfeed
- 38. #sqlhelp Twitter got me my job at SurveyMonkey. RT = Re-Tweet MRT = Modified Re-Tweet TL;DR = Too Long, Didn’t Read # = This is a Hashtag
- 39. Training & Keeping Current Follow SQL professionals on Twitter. They will post their blogs AND re-tweet other articles and news! ! Sign up for free newsletters. ! Brent Ozar Unlimited® Weekly Links - BrentOzar.com ! ! ! SQLSkills Insiders Weekly Newsletter - SQLSkills.com/Insider
- 40. Training & Keeping Current Online Training • PluralSight.com • Free WebCasts (I hear about them on Twitter) ! Conferences and In-Person • SQLSkills (Paul Randal and Kimberly Tripp) • sqlskills.com/sql-server-training/ • IEPTO1: Immersion Event on Performance Tuning and Optimization – Part 1 (formerly IE1)
- 41. sqlpass.org/summit/2014/Home.aspx PASS Summit! #sqlpass #summit14 #sqlkilt #sqlrun #sqllongrun #ArgenisWithoutBorders bit.ly/1pJhn11
- 42. A SQL Server Developer’s Toolkit Questions? • Demo Scripts: github.com/kbenzel/sql-scripts/ • Kirsten Benzel, Database Engineer at SurveyMonkey • Twitter: CausalDensity • Email: [email protected] • Blog: http://kSQL.net • SQL Server Magazine: http://sqlmag.com/author/ kirsten-benzel