I am trying to insert some data into a table in my database in sql server. It has huge amount of data, I am talking about millions of records.
I kept getting error 9002
The transaction log for database 'GCVS2' is full. To find
out why space in the log cannot be reused, see the log_reuse_wait_desc
column in sys.databases.
When I tried inserting data yesterday, it was fine with no problem, although it did take some time.
I tried it again today but kept getting this error. I checked the log file for my database, and it's auto increment is set to 10% ,unlimited. Is there any way to fix this?
You Can Truncate the transaction log. use the below query
BACKUP LOG databasename WITH TRUNCATE_ONLY
DBCC SHRINKFILE ( databasename_Log, 1)
Check here for more details
You will need to check the Recovery mode of your database. Put it in Full Recovery mode. After that, make sure there is a transaction log backup in place for your database. You will need to dig through it and make a Maintenance plan, depending upon how critical your data is. That will be the long term solution.
For time being you can shrink your log files using following DBCC command -
BACKUP LOG DBName WITH TRUNCATE_ONLY
DBCC SHRINKFILE ( DBNameLog, 1)
Or you can do it through Object Explorer. Refer to this link for details. But you will have to set your Database to Simple Recovery model to use the Shrink command
I have 2 databases. I have created a logic where firstly i delete all the data from Database2 with Truncate & then copy all the data from Database1 to Database2 with INSERT INTO.
This process runs every 2 days. The size of Database1 is around 1 GB.
This was all working good but now suddenly i started running out of space. My C: drive just got full & the reason i found was Transaction Log of Database2. Every time i did the above mentioned process which runs with MVC Website application, the Transaction Log goes increasing & increasing.
I can afford to lose data from Database2 didn't want Transaction Logs.
Is there any solution for this?
One option is to shrink your log file.
USE YourDatabaseName;
GO
-- Truncate the log by changing the database recovery model to SIMPLE.
ALTER DATABASE YourDatabaseName
SET RECOVERY SIMPLE;
GO
-- Shrink the truncated log file to 1 MB.
DBCC SHRINKFILE (YourDatabaseName_Log, 1);
GO
-- Reset the database recovery model.
ALTER DATABASE YourDatabaseName
SET RECOVERY FULL;
GO
Further reading: How do you clear the SQL Server transaction log?
Turn your transaction logging on the database 2 to simple:
https://technet.microsoft.com/en-us/library/ms175987(v=sql.105).aspx
I have a SQL Server 2008 database with a .mdf file with 1 GB and a .ldf file (log) with 70 GB. I really don't know what took my log file to be so big in a week and to stop to increase, but my main issue is to fix this problem.
I'm used to reduce the log file shrinking it, but I can only shrink IF I backup it first. If I try to shrink without backuping first (using SSMS), nothing happens, even with SSMS showing that available free space is big. I can try shrinking many times but it will work only if I backup first.
The problem is that I can't backup it this time because I don't have free space (the total size of my HD is 120 GB).
Note 1: my database is set to use the full recovery model because I need to be able to do point-in-time recoveries.
Note 2: I know that shrink increases the index fragmentation. After shrinking, I can use REBUILD in indexes to avoid this.
You can temporarily set recovery model to simple and truncate log
you will lose point-in-time recovery ability in time period between last successful log backup and end of the next differential backup that you can take after log cleanup. But point in time recovery possible after backup time onwards
You also need to find the long running transaction that is active and find the root cause
You can see here http://blog.sqlxdetails.com/transaction-log-survival-guide-shrink-100gb-log/
With the help of below command you can clear transaction log file. command is well commented.
-- see the log size
DBCC SQLPERF(LOGSPACE);
--taking backup for log file before shrink it
BACKUP LOG MyTestDB
TO DISK = 'E:\PartProcForOld_log_backup\MyTestDB.TRN'
GO
-- this command will tell you the log file name
SELECT name
FROM sys.master_files
WHERE database_id = db_id()
AND type = 1
--- these below command will alter database with actual shrink
GO
-- Truncate the log by changing the database recovery model to SIMPLE.
ALTER DATABASE MyTestDB
SET RECOVERY SIMPLE;
GO
-- Shrink the truncated log file to 1 MB.
DBCC SHRINKFILE (MyTestDB_log, 1);
GO
-- Reset the database recovery model.
ALTER DATABASE MyTestDB
SET RECOVERY FULL;
GO
I consider myself a very experienced SQL person. But I'm failing to do these two things:
Reduce the size of the allocated log.
Truncate the log.
DBCC sqlperf(logspace)
returns:
Database Name Log Size (MB) Log Space Used (%) Status
ByBox 1964.25 30.0657 0
The following does not work with SQL 2008
DUMP TRANSACTION ByBox WITH TRUNCATE_ONLY
Running the following does nothing either
DBCC SHRINKFILE ('ByBox_1_Log' , 1)
DBCC shrinkdatabase(N'bybox')
I've tried backups. I've also tried setting the properties of the database - 'Recover Model' to both 'FULL' and 'SIMPLE' and a combination of all of the above. I also tried setting the compatibility to SQL Server 2005 (I use this setting as I want to match our production server) and SQL Server 2008.
No matter what I try, the log remains at 1964.25 MB, with 30% used, which is still growing.
I'd like the log to go back down near 0% and reduce the log file size to, say, 100 MB which is plenty. My database must hate me; it just ignores everything I ask it to do regarding the log.
One further note. The production database has quite a few replicated tables, which I turn off when I perform a restore on my development box by using the following:
-- Clear out pending replication stuff
exec sp_removedbreplication
go
EXEC sp_repldone #xactid = NULL, #xact_segno = NULL,
#numtrans = 0, #time = 0, #reset = 1
go
Trying:
SELECT log_reuse_wait, log_reuse_wait_desc
FROM sys.databases
WHERE NAME='bybox'
Returns
log_reuse_wait log_reuse_wait_desc
0 NOTHING
How can I fix this problem?
Looking at this and setting the recovery model to FULL I have tried the following:
USE master
GO
EXEC sp_addumpdevice 'disk', 'ByBoxData', N'C:\<path here>\bybox.bak'
-- Create a logical backup device, ByBoxLog.
EXEC sp_addumpdevice 'disk', 'ByBoxLog', N'C:\<path here>\bybox_log.bak'
-- Back up the full bybox database.
BACKUP DATABASE bybox TO ByBoxData
-- Back up the bybox log.
BACKUP LOG bybox TO ByBoxLog
which returned:
Processed 151800 pages for database 'bybox', file 'ByBox_Data' on file 3.
Processed 12256 pages for database 'bybox', file 'ByBox_Secondary' on file 3.
Processed 1 pages for database 'bybox', file 'ByBox_1_Log' on file 3.
BACKUP DATABASE successfully processed 164057 pages in 35.456 seconds (36.148 MB/sec).
Processed 2 pages for database 'bybox', file 'ByBox_1_Log' on file 4.
BACKUP LOG successfully processed 2 pages in 0.056 seconds (0.252 MB/sec).
Perfect! But it's not.
And DBCC SHRINKFILE ('ByBox_1_Log' , 1) now returns with
DbId FileId CurrentSize MinimumSize UsedPages EstimatedPages
7 2 251425 251425 251424 251424
and DBCC SQLPERF(LOGSPACE) still reports 30% usage.
I think I may have to resign myself to the fact there could well be a bug in SQL Server 2008, or that my log file has been corrupted in some manner. However, my database is in good working order, which leads me to think there is a bug (shudders at the thought).
In my situation, I had a 650 MB database with a 370 GB log file in SQL Server 2008. No matter what I tried, I could not get it to shrink down. I tried everything listed as answers here but still, nothing worked.
Finally, I found a very short comment somewhere else that did work. It is to run this:
BACKUP LOG DatabaseName TO DISK = N'D:\Backup\DatabaseName_log.bak'
GO
DBCC SHRINKFILE('MyDatabase_Log', 1)
GO
This caused the log file to shrink from 37 GB down to 1 MB. Whew!
I found DBCC SHRINKFILE (Transact-SQL) (MSDN).
The following example shrinks the log file in the AdventureWorks database to 1 MB. To allow the DBCC SHRINKFILE command to shrink the file, the file is first truncated by setting the database recovery model to SIMPLE.
USE AdventureWorks;
GO
-- Truncate the log by changing the database recovery model to SIMPLE.
ALTER DATABASE AdventureWorks
SET RECOVERY SIMPLE;
GO
-- Shrink the truncated log file to 1 MB.
DBCC SHRINKFILE (AdventureWorks_Log, 1);
GO
-- Reset the database recovery model.
ALTER DATABASE AdventureWorks
SET RECOVERY FULL;
GO
Beware of the implications of changing recovery models!!
And now for one more sobering thought for all you production DBAs thinking about using the script:
BEFORE YOU CHANGE THE RECOVERY MODEL FROM FULL TO SIMPLE... there's no worries if you're in a development/QA environment. But if you're in a production environment where you're responsible to ensure full recovery of data in the event of an issue, you may want to take a closer look at what BOL says regarding doing this (see BOL under: "Managing Databases" -> "Transaction Log Management" -> "Recovery Models and Transaction Log Management"):
A database can be switched to another recovery model at any time. However, switching from the simple recovery model, is unusual. Be aware that if you switch to the full recovery model during a bulk operation, the logging of the bulk operation changes from minimal logging to full logging, and vice versa.
After switching from the simple recovery model
If you switch from the full or bulk-logged recovery model to the simple recovery model, you break the backup log chain. Therefore, we strongly recommend that you back up the log immediately before switching, which allows you to recover the database up to that point. After switching, you need to take periodic data backups to protect your data and to truncate the inactive portion of the transaction log.
After switching to the simple recovery model
If you switch from the full or bulk-logged recovery model to the simple recovery model, you break the backup log chain. Therefore, we strongly recommend that you back up the log immediately before switching, which allows you to recover the database up to that point. After switching, you need to take periodic data backups to protect your data and to truncate the inactive portion of the transaction log.
Really? "We strongly recommend that you back up the log immediately before switching, which allows you to recover the database up to that point." I cannot understand why this little tidbit of information is hidden in a section named "After switching to the simple recovery model" making most "normal people" think they can go ahead and switch it and then continue or come back and read this after changing it.
Rant
To Microsoft: So, please correct me if I'm wrong, if I fail to do the t-log backup BEFORE changing from FULL to SIMPLE and lo and behold my database gets corrupted somehow (ever heard of Murphy's law?) right before I'm able to take a backup... then I'm screwed, right? If switching the recovery model of my ****production**** database from FULL to SIMPLE is something that can break the backup log chain such that if I fail to take a transaction log backup before doing it (like it suggests above) I'm potentially going to lose data, then WHY THE HECK AREN'T YOU HIGHLIGHTING THAT IN A BLINKING MARQUEE MAKING IT A BIGGER DEAL than you seem to be??! You should literally be grabbing me by the shirt and slaping me to get my attention (so to speak) and warning me of the importance of this UPFRONT!!
Another easy way to reduce the size of the log file, is to:
Backup logs
Full backup of the database
Shrink the logs file
Backup logs again
Shrink the logs file again
In this way you don't have to modify any database parameters and your logs file is 1 MB sized.
I've always hated the way SQL Server handles the physical shrinking of log files. Please note that I've always done this via Enterprise Manager/SQL Server Management Studio, but it seems that when you shrink/truncate the log file, the physical size of the log file will not reduce until after doing a full backup on the database's data file, and then backing up the log file again. I could never nail down the exact pattern, but you could try and see what the exact sequence is. However, it has always involved doing a full backup of the data file.
Found the solution!
I added a load of data to the database, so the log was forced to expand. I then removed the uneeded data to get my database back to how it was.
Backup and voila, a perfect 0% log.
So the solution is to make the log expand.
Try running
DBCC OPENTRAN
to check if there are any open transactions.
This will sound kind of stupid, but I find that I have to perform two full database and log file backups to be able to shrink the database.
When using SQL Server Management Studio, after doing a full database backup followed by a full transaction log backup, the shrink files page shows plenty of free space available, but it will not truncate the log files.
However, when I run a second full backup of both the database and the log file, I find that the full backup of the transaction log is much much smaller as it appears to only be backing up the new changes since the last backup.
Once this second backup is complete, I run the shrink tool again within the management studio. It still shows that there is plenty of available free space, but this time when I click OK, the log file reduces in size.
So, try the following.
Take a full backup of both the database and the log file. If you check within the shrink tool, you should see that there is now plenty of available free space in the log file. However clicking OK won't remove the free space.
Take a second full backup of both the database and the log file. You should find the full backup of the database is similar in size to the first full backup. The size of the full transaction log backup should be much smaller.
Run the shrink tool on the log file and with any luck, the log file should reduce in size. The last time I did this, it reduced from 180 GB to 12 MB and the shrink tool states that there is still 10 MB of available free space within the file.
I finally found a solution for the logfile shrink problem. All of the previous options did not work for me and did not shrink the logfile to the required size. The solution I found is:
Backup your database
Set the recovery mode to simple
Detach the database with SQL Server Management Studio
Delete the logfile
Attach the database without the logfile. In the lower half of the "add screen" you get a line which says that the logfile is missing
Click remove for this line and OK for the attach
Set the recovery mode back to full
This can be a pain, and there are many things it could be. The first thing you should make sure of is that there is not a "stuck" transaction. If you have a transaction that never closes, you cannot ever shrink the log. Run "DBCC OPENTRAN" to find the longest running transaction.
Also, make sure you reorganize (I think that's the proper term) and move everything to the beginning of the file before shrinking.
I know what you mean - SQL server can be a bit maddening that way. Here is some sample code to try out. In essence, it truncates the log file and then tries to shrink the file. Let me know how it works. One other thing...you wouldn't have uncommitted transactions, would you?
Use YourDatabase
GO
DBCC sqlperf(logspace) -- Get a "before" snapshot
GO
BACKUP LOG BSDIV12Update WITH TRUNCATE_ONLY; -- Truncate the log file, don't keep a backup
GO
DBCC SHRINKFILE(YourDataBaseFileName_log, 2); -- Now re-shrink (use the LOG file name as found in Properties / Files. Note that I didn't quote mine).
GO
DBCC sqlperf(logspace) -- Get an "after" snapshot
GO
Update: Simon notes that he is getting an error on the BACKUP command. I didn't realize that "Truncate_only" has been discontinued in SQL Server 2008 when I answered earlier. After a bit of research, the recommended steps to shrink the log file is to (a) Change the Recovery Model to Simple and then (b) shrink the file using DBCC ShrinkFile as above. Unfortunately, you mention that you already tried setting the recovery model to Simple so I assume that you also ran the DBCC Shrinkfile afterward. Is this correct? Please let me know.
Reading the answers I hardly believe that they are written by DBAs. Basic golden rules:
In any database before you performing any maintenance, including
shrinking the log, you should perform a full backup.
Carry out any database maintenance, including shrinking, when
nothing else is happening on that particular database for the whole
duration of the maintenance. If it is necessary, suspend
non-essencial applications. Always keep it in mind that the healthy
database is the soul of any applications interacting with it.
After all this, the following commands for shrinking the database's transaction log always worked fine with me on SQL Server 2005 and later SQL Server versions:
USE DatabaseName
GO
-- Truncate the Transaction log
ALTER DATABASE DatabaseName SET RECOVERY SIMPLE
CHECKPOINT
ALTER DATABASE DatabaseName SET RECOVERY FULL
GO
-- Shrink the Transaction Log as recommended my Microsoft.
DBCC SHRINKFILE ('database_txlogfilelogicalname', [n -size to shrink in MBytes])
GO
-- Pass the freed pages back to OS control.
DBCC SHRINKDATABASE (DatabaseName, TRUNCATEONLY)
GO
-- Tidy up the pages after shrink
DBCC UPDATEUSAGE (0);
GO
-- IF Required but not essential
-- Force to update all tables statistics
exec sp_updatestats
GO
I have finally come to the conclusion that there is a bug in SQLServer 2008.
I've tried everything, and every combination I can think of. I've backed up the database, dropped it, re-created it, restored it. Exact same problem.
I also ran:
DBCC CHECKDB
DBCC UPDATEUSAGE (bybox)
And all checks out ok.
Roll on the next service pack is all I can say.
SQL Server 2012: I had an issue where no log file (and all were already in SIMPLE recovery) would shrink.
This worked for me... I restarted the SQL Server instance (because I could) and every one of those bad boys shrunk.
Whatever was holding it up from shrinking was released with the restart. This is only good for emergencies (or when it's your server), not a regular long-term solution.
Please run:
SELECT name, log_reuse_wait, log_reuse_wait_desc FROM sys.databases
and see what is the log_reuse_wait for your problem db
if it is replication that this is error you need this value 0, 2 or 4