Help: SQL Server

Sharing my knowlege about SQL Server Troubleshooting Skills

Solution – SQL Server Backup Failing with EXCEPTION_ACCESS_VIOLATION

Posted by blakhani on March 25, 2015


Recently someone posted on Facebook group about a problem. He informed that whenever the backups are taken in SQL Server 2014 instance, it’s failing with error as below.

Msg 3013, Level 16, State 1, Line 4
BACKUP DATABASE is terminating abnormally.

When we open ERRORLOG we saw below

***Stack Dump being sent to D:\MSSQL\LOG\SQLDump0089.txt
SqlDumpExceptionHandler: Process 1144 generated fatal exception c0000005 EXCEPTION_ACCESS_VIOLATION. SQL Server is terminating this process.
* ******************************************************************************* 

* BEGIN STACK DUMP:
*   01/01/15 18:05:25 spid 1144
* Private server build.
*
*
*   Exception Address = 00007FF85B652D45 Module(sqlmin+00000000009D2D45)
*   Exception Code    = c0000005 EXCEPTION_ACCESS_VIOLATION
*   Access Violation occurred reading address 00007FFA0C208980
* Input Buffer 116 bytes –
*             backup database test to disk=’c:\temp\test.bak’

As we can see that Access Violation is occurring while running backup database command. Whenever there is a dump generated, there is a MDMP file is also generated. Here is the stack when we analyze the dump. Yon can refer http://mssqlwiki.com/2012/10/16/sql-server-exception_access_violation-and-sql-server-assertion link to know more about identifying stack.

sqlmin!PerfmonManager::AddInstance+0x4e0
sqlmin!BackupPerfmonCounter::AddInstance+0x1f
sqlmin!BackupDevicePerfmonCounter::Init+0x13c
sqlmin!BackupFileDesc::InitPerfCounters+0xc                          <<Performance Counters Initialization.
sqlmin!BackupMediaIo::Initialize+0x29
sqlmin!BackupMedium::CreateDeviceObject+0x224
sqlmin!BackupMedium::Open+0x27
sqlmin!BackupStream::OpenForBackup+0x27
sqlmin!BackupStream::ThreadMainRoutine+0x160
sqlmin!BackupThread::ThreadBase+0x51
sqlmin!SubprocEntrypoint+0xa7f                                       << This is the child thread of main backup thread.
sqldk!SOS_Task::Param::Execute+0x21e
sqldk!SOS_Scheduler::RunTask+0xa8
sqldk!SOS_Scheduler::ProcessTasks+0x279
sqldk!SchedulerManager::WorkerEntryPoint+0x24c
sqldk!SystemThread::RunWorker+0x8f
sqldk!SystemThreadDispatcher::ProcessWorker+0x3ab
sqldk!SchedulerManager::ThreadEntryPoint+0x226
kernel32!BaseThreadInitThunk+0xd
ntdll!RtlUserThreadStart+0x1d

Here is the version of SQL Server.

Microsoft SQL Server 2014 – 12.0.2000.8 (X64)
Feb 20 2014 20:04:26
Copyright (c) Microsoft Corporation
Enterprise Edition: Core-based Licensing (64-bit) on Windows NT 6.3 <X64> (Build 9600: ) (Hypervisor)

This is identified as an issue with SQL Server 2014 RTM version. Here is the KB article – http://support.microsoft.com/en-us/kb/2973444 (FIX: "SQL Server performance counters are disabled" when you move the SQL Server resource in SQL Server 2014)

Fix: Cumulative Update 2 for SQL Server 2014 or any later version of SQL Server 2014.

If you are running higher version and above fix then you need to find out why performance monitor is not showing SQL Server related counters. The way to verify it would be to use DMV

select * from sys.dm_os_performance_counters

 

You should see many counters specifically look for “Backup/Restore Throughput/sec”. In SQL 2014 this is a new performance counter added to get backup/restore speed and that missing counter (Backup/Restore Throughput/sec) which is causing backup to fail.

  • Cheers,
  • Balmukund Lakhani
  • Twitter @blakhani
  • Author: SQL Server 2012 AlwaysOnPaperback, Kindle
  • Posted in EXCEPTION_ACCESS_VIOLATION, SQL Server, SQL Server 2014 | Tagged: , , | 2 Comments »

    Help : Unable to start SQL Service after installing Service Pack or Cumulative Upgrade or Hotfix

    Posted by blakhani on March 10, 2015


    This is one of the common issue which a DBA might have seen. Let’s first understand what happens whenever an upgrade is performed for SQL Server. During installation of patch the binaries are updated, resource database is replaced. During the restart, after successful completion, the upgrade scripts are run. This is the time when Login would fail with below error.

    2015-01-05 10:45:03.23 Logon       Error: 18401, Severity: 14, State: 1.
    2015-01-05 10:45:03.23 Logon       Login failed for user ‘NT AUTHORITY\SYSTEM’. Reason: Server is in script upgrade mode. Only administrator can connect at this time. [CLIENT: <local machine>]

    This error is normal for sometime till upgrade scripts have completed execution. It has been observed that sometimes, due to system configuration, the script don’t run successfully. If upgrade script has not run, SQL would shutdown automatically and below would be the message in SQL ERRORLOGs.

    2015-01-05 10:45:07.94 spid9s      Error: 3417, Severity: 21, State: 3.
    2015-01-05 10:45:07.94 spid9s      Cannot recover the master database. SQL Server is unable to run. Restore master from a full backup, repair it, or rebuild it. For more information about how to rebuild the master database, see SQL Server Books Online.

    Above message is very scary and suggest you to do something which you may not want. Now, If you are into situation where upgrade is not complete, there are few things you should try. Very first thing is look at ERRORLOG and locate a place where you would find some error during script upgrade. Here are few sample errors which you might see.

    Error 1 – CREATE DATABASE failed

    2015-01-05 10:45:02.55 spid9s      Database ‘master’ is upgrading script ‘sqlagent90_sysdbupg.sql’ from level 0 to level 2.
    2015-01-05 10:45:07.94 spid9s      Error: 5133, Severity: 16, State: 1.
    2015-01-05 10:45:07.94 spid9s      Directory lookup for the file "F:\Program Files\Microsoft SQL Server\MSSQL.1\MSSQL\DATA\temp_MS_AgentSigningCertificate_database.mdf" failed with the operating system error 3(The system cannot find the path specified.).
    2015-01-05 10:45:07.94 spid9s      Error: 1802, Severity: 16, State: 1.
    2015-01-05 10:45:07.94 spid9s      CREATE DATABASE failed. Some file names listed could not be created. Check related errors.
    2015-01-05 10:45:07.94 spid9s      Error: 912, Severity: 21, State: 2.
    2015-01-05 10:45:07.94 spid9s      Script level upgrade for database ‘master’ failed because upgrade step ‘sqlagent90_sysdbupg.sql’ encountered error 598, state 1, severity 25. This is a serious error condition which might interfere with regular operation and the database will be taken offline. If the error happened during upgrade of the ‘master’ database, it will prevent the entire SQL Server instance from starting. Examine the previous errorlog entries for errors, take the appropriate corrective actions and re-start the database so that the script upgrade steps run to completion.
    2015-01-05 10:45:07.94 spid9s      Error: 3417, Severity: 21, State: 3.
    2015-01-05 10:45:07.94 spid9s      Cannot recover the master database. SQL Server is unable to run. Restore master from a full backup, repair it, or rebuild it. For more information about how to rebuild the master database, see SQL Server Books Online.

    Okay. Problem here is that F:\Program Files\Microsoft SQL Server\MSSQL.1\MSSQL\DATA folder is missing as we are getting error “The system cannot find the path specified”. The database default path is stored in registry and this is explained here. All we need to do is change the registry to correct value and start SQL Services.

    Error 2 – affinity mask related error

    2015-01-01 00:00:06.79 spid7s      Database ‘master’ is upgrading script ‘sqlagent100_msdb_upgrade.sql’ from level 0 to level 3.
    2015-01-01 00:00:06.79 spid7s      —————————————-
    2015-01-01 00:00:06.79 spid7s      Starting execution of PREINSTMSDB100.SQL
    2015-01-01 00:00:06.79 spid7s      —————————————-
    ..
    2015-01-01 00:00:06.97 spid7s      Setting database option COMPATIBILITY_LEVEL to 100 for database msdb.
    2015-01-01 00:00:07.01 spid7s      Configuration option ‘allow updates’ changed from 1 to 1. Run the RECONFIGURE statement to install.
    2015-01-01 00:00:07.01 spid7s      Configuration option ‘allow updates’ changed from 1 to 1. Run the RECONFIGURE statement to install.
    2015-01-01 00:00:07.01 spid7s      Error: 5833, Severity: 16, State: 1.
    2015-01-01 00:00:07.01 spid7s      The affinity mask specified is greater than the number of CPUs supported or licensed on this edition of SQL Server.
    2015-01-01 00:00:07.01 spid7s      Error: 912, Severity: 21, State: 2.
    2015-01-01 00:00:07.01 spid7s      Script level upgrade for database ‘master’ failed because upgrade step ‘sqlagent100_msdb_upgrade.sql’ encountered error 5833, state 1, severity 16. This is a serious error condition which might interfere with regular operation and the database will be taken offline. If the error happened during upgrade of the ‘master’ database, it will prevent the entire SQL Server instance from starting. Examine the previous errorlog entries for errors, take the appropriate corrective actions and re-start the database so that the script upgrade steps run to completion.
    2015-01-01 00:00:07.02 spid7s      Error: 3417, Severity: 21, State: 3.
    2015-01-01 00:00:07.02 spid7s      Cannot recover the master database. SQL Server is unable to run. Restore master from a full backup, repair it, or rebuild it. For more information about how to rebuild the master database, see SQL Server Books Online.

    To fix above problem, we can start the Instance by skipping the post upgrade script by adding trace flag –T902 to the startup parameters. Steps to enable trace flag can be found in this KB Once SQL is started and we should be able to connect to SQL Server normally. Then connect via SSMS and check the checkbox “Automatically set processor affinity mask for all processors” under instance properties. This could have also been archived using the below TSQL.

    ALTER SERVER CONFIGURATION SET PROCESS AFFINITY CPU = AUTO
    GO
    
    

    Once configuration change as been done, we need to remove the trace flag and then start SQL normally. Post restart, we need to check SQL Server ERRORLOG and make sure we have below line.

    Recovery is complete. This is an informational message only. No user action is required

    This message during upgrade scenario means that post upgrade script have been executed successfully.

    If you have encountered any error during upgrade, please comment and let me know.

  • Cheers,
  • Balmukund Lakhani
  • Twitter @blakhani
  • Author: SQL Server 2012 AlwaysOnPaperback, Kindle
  • Posted in SQL Server, Tips and Tricks, Trace Flag | Tagged: | Leave a Comment »

    Solution – Getting Error while running rebuild index on CRM database – Unable to process object ID <ID> (object <Name> because it is a synonym. If the object referenced by the synonym is a table or view, retry the operation using the base object that the synonym references."

    Posted by blakhani on February 11, 2015


    While rebuilding all indexes on a CRM database, one of my friend reported below error.

    Msg 2547, Level 16, State 1, Line 15
    Unable to process object ID 2007014231 (object "AttributeTypes") because it is a synonym. If the object referenced by the synonym is a table or view, retry the operation using the base object that the synonym references."

    First let’s understand the meaning of the error. The error message clearly says that error is due to the fact that we are running DBCC command for a synonym which is not possible. we need to run that on base object. Here is the simple repro of the error.

    use tempdb
    go
    if object_id('TestTable') is not null
    drop table TestTable
    go
    if object_id('TestTableSyn') is not null
    drop synonym TestTableSyn
    go
    create table TestTable (i int)
    go
    CREATE SYNONYM [dbo].[TestTableSyn] FOR [dbo].TestTable
    GO
    dbcc DBREINDEX('TestTableSyn')
    go
    --dbcc CHECKTABLE('TestTableSyn')
    --go
    

    You would get below error in the result.

    Msg 2547, Level 16, State 1, Line 15

    Unable to process object ID 2007014231 (object "AttributeTypes") because it is a synonym. If the object referenced by the synonym is a table or view, retry the operation using the base object that the synonym references."

     

    Let’s have a look at script what he was using.

    USE DBName   --Enter the name of the database you want to reindex 
    
    DECLARE @TableName varchar(255) 
    
    DECLARE TableCursor CURSOR FOR 
    SELECT table_name FROM information_schema.tables 
    WHERE table_type = 'base table' 
    
    OPEN TableCursor 
    
    FETCH NEXT FROM TableCursor INTO @TableName 
    WHILE @@FETCH_STATUS = 0 
    BEGIN 
    PRINT 'Reindexing ' + @TableName 
    DBCC DBREINDEX(@TableName,' ',90) 
    FETCH NEXT FROM TableCursor INTO @TableName 
    END 
    CLOSE TableCursor 

    Do you see any problem? It looks good but notice that schema name is not considered in the script. I executed below and that would explain the cause.

    Select * from sys.objects where name = 'AttributeTypes'
    

    Looking at output, we can tell that there are two objects with same name in two different schemas (1 and 5). Schema 1 would be dbo and 5 was Metadata. In the script which he was using, the schema name was not considered. Due to this DBCC DBREINDEX is running for dbo.AttributeTypes which is synonym (2007014231)and due to that we are seeing error.

    I provided below modified script.

    DECLARE @TableName SYSNAME
            ,@schema_name SYSNAME
            ,@fullname VARCHAR(8000)
    DECLARE TableCursor CURSOR
    FOR
    SELECT  TABLE_NAME
            ,TABLE_SCHEMA
    FROM information_schema.tables
    WHERE table_type = 'base table'
    OPEN TableCursor
    FETCH NEXT FROM TableCursor INTO @TableName ,@schema_name
    WHILE @@FETCH_STATUS = 0
    BEGIN
            SET @fullname = @schema_name + '.' + @TableName
            PRINT @fullname
            DBCC DBREINDEX (@fullname,' ',90)
            FETCH NEXT FROM TableCursor INTO @TableName ,@schema_name
    
    END
    CLOSE TableCursor
    DEALLOCATE TableCursor
    go
    
    

    I tested the script and it worked fine for CRM database. Same issue was blogged by my colleague John over here and it has some more suggestions to try.

    On a side note, ideally you should use Maintenance plan available with SQL Server to rebuild the indexes.

    Hope this helps!

  • Cheers,
  • Balmukund Lakhani
  • Twitter @blakhani
  • Posted in Error, SQL Server | Tagged: , , , | 3 Comments »

     
    Follow

    Get every new post delivered to your Inbox.

    Join 947 other followers