|
|
|
UDP Broadcaster
      
Group: General Forum Members
Last Login: Monday, June 10, 2013 4:11 AM
Points: 1,474,
Visits: 2,344
|
|
Koen Verbeeck (7/20/2012) I like the intent of the question, but alas poorly executed.
Agreed.
|
|
|
|
|
Ten Centuries
      
Group: General Forum Members
Last Login: Today @ 10:50 AM
Points: 1,230,
Visits: 1,364
|
|
|
|
|
|
SSCrazy
      
Group: General Forum Members
Last Login: Today @ 5:26 AM
Points: 2,797,
Visits: 1,745
|
|
This implies two things: 1. Any open transaction before full DB backup READ but not committed or rollbacked will be included in the full backup. [It sounds strange to me as ideally backups should not have consider any open transaction] 2. New transaction which is opened after full DB backup READ starts, which is also not committed or rolled back even after full DB backup READ ends, should not be included in the backup. [Which looks ok to me]
Could someone help me with the first point ?
|
|
|
|
|
Ten Centuries
      
Group: General Forum Members
Last Login: Thursday, June 06, 2013 4:06 PM
Points: 1,219,
Visits: 13,509
|
|
WayneS (7/19/2012) The full backup will also contain the uncommitted transactions from transaction 4 at the time that the backup finishes. If a restore is necessary, the UNDO portion of the restore will roll back those transactions. Therefore, the full backup will contain transactions from all four of the transactions.
A small error in the question: transaction 1 never is committed, and transaction 2 is committed twice (which can't happen). Obviously, one of those transaction 2 commits is for transaction 1.
+1
rfr.ferrari DBA - SQL Server 2008 MCITP | MCTS
remember is live or suffer twice!
|
|
|
|
|
SSC Eights!
      
Group: General Forum Members
Last Login: Today @ 11:10 AM
Points: 856,
Visits: 1,278
|
|
| Wow, interesting. I'd been under the false impression that the full backup used an internal snapshot to capture the state of the database at the time the backup started. I'm intrigued that that is not the case, and changes to pages made after the backup has read the page are captured from the log at the end of the backup process. This seems a bit awkward to me, as it makes the point in time of the backup a bit fuzzy depending on how long it takes to process.
|
|
|
|
|
SSCertifiable
       
Group: General Forum Members
Last Login: Today @ 10:28 AM
Points: 5,314,
Visits: 11,305
|
|
A full backup contains enough of the transaction log so the database can be made consistent on restore. The way the question is worded transaction 4 commits before the end of the backup so would be included in the backup.
---------------------------------------------------------------------
|
|
|
|
|
Valued Member
      
Group: General Forum Members
Last Login: Friday, June 14, 2013 10:11 AM
Points: 56,
Visits: 492
|
|
| Deleted by author (I neglected to read the first post before posting).
|
|
|
|
|
Hall of Fame
       
Group: General Forum Members
Last Login: Today @ 11:13 AM
Points: 3,250,
Visits: 65,553
|
|
|
|
|
|
Hall of Fame
       
Group: General Forum Members
Last Login: Yesterday @ 7:53 PM
Points: 3,367,
Visits: 1,577
|
|
I should apologize. I read the silly thing several times unfortunately I missed the fact that I failed to commit transaction 1 at all. To much copy and paste I guess. Obviously I need to read it a few more times next time.
Kenneth Fisher I strive to live in a world where a chicken can cross the road without being questioned about its motives. -------------------------------------------------------------------------------- For better, quicker answers on T-SQL questions, click on the following... http://www.sqlservercentral.com/articles/Best+Practices/61537/ For better answers on performance questions, click on the following... http://www.sqlservercentral.com/articles/SQLServerCentral/66909/
Link to my Blog Post --> www.SQLStudies.com
|
|
|
|
|
Mr or Mrs. 500
      
Group: General Forum Members
Last Login: Tuesday, April 02, 2013 8:28 PM
Points: 555,
Visits: 2,137
|
|
I just focussed on the last three lines and counted the 'begins' and 'commits': 4 begin, 3 commit and only the transaction 4 had been commit after the backup reading process... 12:18 AM - Full backup data reading of QotDDatabase ends 12:19 AM - Transaction 4 on QotDDatabase commits 12:20 AM - Full backup of QotDDatabase ends Nice and simple logic...
Thanks!
|
|
|
|