Monday, 13 October 2003 The Anatomy of a Data Modification Subset of: What everyone should know about Performance, Recovery & Logging Monday, 13 October 2003 – 08:00 SDB 308 Kimberly L. Tripp President, SYSolutions, Inc. – SQLSkills.com Principal Mentor, Solid Quality Learning – SolidQualityLearning.com

Monday, 13 October 2003 The Anatomy of a Data Modification 1.User sends UPDATE Update is highly selective (only 5 rows) Indexes exist to aid in finding these rows efficiently The update is a SINGLE statement batch NOT enclosed in BEGIN TRAN…COMMIT TRAN block therefore this is IMPLICIT transaction 2.Server receives the request and locates the data in cache OR reads the data from disk into cache Since this is highly selective only the necessary pages are read into cache (maybe a few extra but that’s not important here) Let’s use an example where the 5 rows being modified are located on 3 different data pages

Monday, 13 October 2003 What it looks like - Data UPDATE… Data Log Server… Cache

Monday, 13 October SQL Server proceeds to lock the necessary data Locks are necessary to give us a consistent point FOR ALL rows from which to start If any other transaction(s) have ANY of these rows locked we will wait until ALL locks have been acquired before we can proceed. In the case of this update (because it’s highly selective and because indexes exist to make this possible) SQL Server will use row level locking. The rows are locked but there are also “intent” locks at higher levels to make sure other larger locks (like page or table level locks) are not attempted (and fail) There are a few locks that have already occurred – within indexes, etc. to read the data – but they are not significant here This sounds complex but it’s not too bad… The Anatomy of a Data Modification

Monday, 13 October 2003 What it looks like - Locks Cache Page Row Update Lock

Monday, 13 October SQL Server can now begin to make the modifications – for EVERY row the process will include: 1. Change the lock to a stricter lock (eXclusive lock) An update lock helps to allow better concurrency by being compatible with other shared locks (readers). Readers can read the pre-modified data as it is transactionally consistent The eXclusive lock is required to make the change because once modified no other reads should be able to see this un-committed change 2. Make the modification to the data row (yes, in cache) 3. Log the modification to the transaction log pages (also in cache) The Anatomy of a Data Modification

Monday, 13 October 2003 What it looks like - Modifications Cache Page Row Update LockExclusive Lock x x Update LockExclusive Lock x x Update LockExclusive Lock x x Update LockExclusive Lock x x Update LockExclusive Lock x x L

Monday, 13 October Finally, the transaction is complete – this is the MOST critical step All rows have been modified There are no other statements in this transaction – i.e. Implicit transaction Steps are: 1.Write all log pages to transaction log ON DISK 2.Release the locks 3.Send a message to the user: (5 Rows Affected) The Anatomy of a Data Modification

Monday, 13 October 2003 What it looks like Write-Ahead Logging L Data Log Server… Cache ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~ Sequential writes Change … Log 5 Rows Affected After the log entries are made and the locks are released…

Monday, 13 October 2003 So now what? The transaction log ON DISK – is up to date The data in CACHE – is up to date But when does the data get written from cache to disk? CHECKPOINT It’s important to realize that the sole purpose of checkpoint is NOT just to write committed pages… Instead a checkpoint writes ALL pages which have changed since they were brought into cache – regardless of the state of the transaction which changed them!

Monday, 13 October 2003 Transaction Recovery and Checkpoints Transactions… Action Required if restart recovery None CheckpointSystem Failure Roll forward Roll back Roll forward Roll back LD L/D L L Time

Monday, 13 October 2003 Check out for information about upcoming events, useful downloads and excellent scripts! There are quite a few resources and/or links to use. MSPress title: SQL Server 2000 High Availability Authors: Allan Hirt with Cathan Cook, Kimberly L. Tripp, Frank McBath ISBN: Check out the main page of for a sample chapter to download! Resources

Monday, 13 October 2003 Resources From Books Online “Home Page” select White Papers to get to msdn For Tech Net articles use: /sql/default.asp?frame=true See for all sorts of useful links, resources and whitepapers, etc. Support Resources listed: asp asp

Monday, 13 October 2003 Articles in SQL Server Magazine, Aug/Sept 2003: Recovering from Isolated Corruption, InstantDoc #39657 The Best Place for Bulk_Logged, InstantDoc #39782 Articles in SQL Server Magazine, Sept 2002: Before Disaster Strikes, InstantDoc ID#25915 Log Backups Paused for Good Reason, InstantDoc #26032 Restoring After Isolated Disk Failure, InstantDoc #26067 Filegroup Usage for VLDBs, InstantDoc #26031 Articles in TSQLSolutions Journal, Oct 2001: TSQLTutor Quick Tip: Saving Production Data from Production dBAs, InstantDoc ID#22073 Resources

Monday, 13 October 2003 BOL Favorites Make sure to get the latest version of the BOL. See Technical Resources, Product Documentationwww.microsoft.com/sql Functions: DATABASEPROPERTYEX, OBJECTPROPERTY, SERVERPROPERTY “Using Recovery Models” “Checkpoints and the Active Portion of the Log” “Virtual Log Files” “Shrinking the Transaction Log” “Optimizing Transaction Log Performance” “Switching Recovery Models” “ALTER DATABASE”

Monday, 13 October 2003 Kimberly L. Tripp President, SYSolutions, Inc. Website: Principal Mentor, Solid Quality Learning Website: Thank you!