Data – Page 16 – TroubleshootingSQL

T-SQL Tuesday #13 – What the Business Says Is Not What the Business Wants

This is my second post for T-SQL Tuesday. This time around, T-SQL Tuesday is being hosted by Steve Jones (Twitter | Blog), The Voice of the DBA on SQLServerCentral.com this month. The topic this time around is not a technical topic but nonetheless very important in the grander scheme of things.

I have seen multiple scenarios where the teams involved in deploying solutions/projects work in silos. There is the management team which decides what the final outcome of the project should look like and what the end deliverables are. Then the set of developers who can throw together gazillion lines of code (in multiple languages) to put together a front-end, maybe a middle-tier, which talks to a database server fetches data and presents it to the end-user. And finally a set of DBAs who are always guilty of the post-deployment issues “till the database is proven innocent” as most applications hit a database to fetch the data at one time or another.

There are some key things to consider and contemplate about before you start writing code and deploying your solutions.

What’s the business?

Knowing your business is like knowing your backyard. The bigger it is the more lethargic you tend to become. Business acumen is not something that is missing in individuals but it has got to do more with “I want to” versus “I don’t want to”! The positive side of knowing enough about your business is that it can make your life easier. This will prevent you from spending additional hours troubleshooting database performance issues since you deployed something that was not needed for the business and unfriendly to your current database environment. So unless you have a fair idea of why you are implementing what you are implementing, you are going to invite trouble later if not sooner!

Requirement Analysis: The Who, What and How

If you have background in Software Design or have been part of a project, then you will know what RA is. However, due to the time-to-market commitments becoming tighter and tighter in a competitive environment, RA becomes a non-entity or a formality with developers developing without knowing why they are doing what they are doing. A definite recipe for disaster! Unless you understand WHO is going to be using what you are developing, WHAT purpose is it going to serve and HOW they will be using it; you will not be in a position to explain why a module or feature may be an overkill or completely unnecessary. Figuring out the Who, What and How requires a certain understanding of the business (my first point) and without this, you are bound to face challenges!

Good-to-have and Must-have

You must have heard this a million times. What is a must-have as opposed to a good-to-have. Without knowing this, you might not be able to land a project where end deliverables are met. It is never a good idea to over-commit. Timelines and deliverables need to be frozen before you start developing which would make me point back to the planning phase. Just because the business asks for it doesn’t mean that the business needs it. Find out if that is going to serve the end-goal. What is being prioritized may not be a must-have but actually a good-to-have. Must-haves in a project are non-comprisable (if the right RA has been done) and the good-to-have is well… (the phrase is self explanatory)

The shops where there are smart people who can look beyond their egos and work efficiently in a cross-team environment usually don’t land up in a soup post-deployment; using band-aids to fix a “gone-to-hell” deployment.

Conclusion: Cohesion in communication among teams is required to formulate a concise design goal to achieve or solve a specific business need. Without this, you will be back to the drawing board in no time after spending substantial amount of effort and $$ in a futile exercise.

Note: The above post is not penned to finger-point at any set of individuals or a group. It is just an attempt to help understand the need for cross collaboration between business groups. Additionally, it is equally pertinent to know your backyard before you deploy something. This is applicable to all business groups and not just DBAs!

Technorati Tags: #TSQL2sDay,Developers,Project Management,Software Development

December 14, 2010

Volume Shadow barfs with 0x80040e14 code

I was replying on a MSDN forum thread recently and found that the a VSS backup was failing due to large number of database volumes residing on a particular volume.

I re-created the issue on my box by limiting the number of Max Worker Threads to 128 on my SQL Server 2008 instance. After making the change to the number of worker threads, I created 99 databases on the D: drive of my server using the following script:

set nocount on
declare @cntr int = 1,@sql varchar(8000)
while @cntr &amp;lt; 100
begin
set @sql = 'CREATE DATABASE [db'+cast(@cntr as varchar(5))+'] ON&amp;nbsp; PRIMARY

( NAME = N' 'db'+cast(@cntr as varchar(5))+''', FILENAME = N' 'D:\DBFiles\db'+cast(@cntr as varchar(5))+'.mdf' ')
LOG ON
( NAME = N' 'db'+cast(@cntr as varchar(5))+'_log' ', FILENAME = N' 'D:\DBFiles\db'+cast(@cntr as varchar(5))+'.ldf'')'

set @cntr = @cntr + 1
exec(@sql)
end

I then attempted to backup the DBFiles folder on the D: drive using Windows Server Backup and it failed.

After that I looked into the SQL Server Errorlog and found the following errors:

2010-12-06 03:39:15.280 Backup Error: 3041, Severity: 16, State: 1.2010-12-06 03:39:15.280 Backup BACKUP failed to complete the command BACKUP DATABASE db68. Check the backup application log for detailed messages.

2010-12-06 03:39:15.370 spid53 Error: 18210, Severity: 16, State: 1.2010-12-06 03:39:15.370 spid53 BackupVirtualDeviceFile::PrepareToFreeze: failure on backup device ‘{DB840D1B-DD17-4D73-83F1-7390F676DEE4}1’. Operating system error 995(The I/O operation has been aborted because of either a thread exit or an application request.).

I had already documented earlier that the infamous 995 error message could be a problem with the sqlvdi.dll but in this case, it is not!

The next thing I did was to pick the errors from the Application Event logs and I found the following errors repeated multiple times:

Error 12/6/2010 3:39:15 AM SQLVDI 1 None SQLVDI: Loc=SignalAbort. Desc=Client initiates abort. ErrorCode=(0). Process=3920. Thread=11604. Client. Instance=SQL2008. VD=Global\{DB840D1B-DD17-4D73-83F1-7390F676DEE4}1_SQLVDIMemoryName_0.Error 12/6/2010 3:39:15 AM SQLWRITER 24583 None Sqllib error: OLEDB Error encountered calling ICommandText::Execute. hr = 0x80040e14. SQLSTATE: 42000, Native Error: 3013

Error state: 1, Severity: 16

Source: Microsoft SQL Server Native Client 10.0

Error message: BACKUP DATABASE is terminating abnormally.

SQLSTATE: 42000, Native Error: 3224

Error state: 1, Severity: 16

Source: Microsoft SQL Server Native Client 10.0

Error message: Cannot create worker thread.

If you look closely at the above error, then you will find that the following error message stands out like a sore thumb: “Cannot create worker thread”. After this issue occurs, you will find that the SQLWriter shows the following when you execute the command: “vssadmin list writers” from a command prompt window:

Writer name: ‘SqlServerWriter’ Writer Id: {a65faa63-5ea8-4ebc-9dbd-a0c4db26912a}

Writer Instance Id: {9075f235-fdee-4445-851b-a69c54bd8b33}

State: [8] Failed

Last error: Non-retryable error

Ensure that you do not have any memory related errors in the SQL Server Errorlogs or any DBCC MEMORYSTATUS outputs printed in the Errorlog during the time of the backup failure. If there aren’t, then the above issue occurs due to the SQL instance hitting a worker thread limit.

The reason for this is explained by Rob Dorr (Principal SQL Server Escalation Engineer at Microsoft) in his blog post:

The controller and database worker come from the worker pool. So if your max worker thread is configured for 255 the answer would be a max of 255 / 2 = 127 databases. You probably won’t get to 127 with this configuration as you have system workers (lazy writer, checkpoint, log writer, and others) that already are using workers.

On my SQL instance, I have 120 database and I need about 240 worker threads to complete this operation. However, I have only 128 worker threads configured for my SQL instance. If you have Dedicated Administrator Connection enabled for the SQL instance, then you can this troubleshooting further and execute a select against the DMV sys.dm_os_schedulers (for SQL Server 2005 or above) and get an output of the following type:

status counts
———————- ——–

HIDDEN ONLINE 168

VISIBLE ONLINE 8

VISIBLE ONLINE (DAC) 1

Query used:

select status, count(*) as counts
from sys.dm_os_schedulers
group by status

As you will see that there are about 168 hidden online schedulers created for performing the Snapshot backups. But SQL instance in question doesn’t have sufficient work threads to facilitate this operation.

What can you do in such situations:

1. Increase the number of worker threads to account for the additional worker threads required for completing the VSS backups. This can be counter-productive as it can lead to non-Buffer Pool memory contention on 32-bit instances of SQL Server. The higher the number of max worker threads on your SQL Server instance, more the amount of non-BPool memory is consumed for the thread stack. This results in depletion of the available Buffer Pool memory.

2. Change the physical location of the SQL Server database files so that all the database files do not reside on the same volume.

3. Perform a VSS Backup with SQL Server VSS Writer service turned OFF. However, you cannot recover any SQL Server database files from such a Volume Shadow Backup.

4. Use SQL Native backups or other VDI backup tools (non-Snapshot) to perform SQL Server database backups.

Note: This also happens when you select any file on a drive which has a high number of SQL Server database files. This is because the SQL Server VSS Writer tries to enumerate the database files for the SQL Server instance that reside on the volume that has been selected for the Snapshot backup.

Technorati Tags: VSS,VDI,Backups,Database Files,Windows

December 6, 2010

Wiki: SQL Database Engine Startup Failures for Clusters

I had recently written a Wiki article on my blog for a troubleshooting startup methodology for SQL Server startup failures for Standalone instances. In this article, I shall cover a similar roadmap for troubleshooting database engine startup failures for clustered SQL Server instances. Some of the steps are going to be the same as troubleshooting start-up failures for stand-along instance but there are some minute differences. So here goes…

The first thing that you need to do is to try and bring the SQL Server resource online from the Failover Cluster Administrator in case you are using a Windows Server 2003 cluster or below. If you are using a Windows Server 2008 or above, you would need to use the Failover Cluster Management Snap-in. If the SQL Server resource fails to come online, then check the SQL Server Errorlogs for any errors that were reported.

Locate the SQL Server Errorlog location and check if any errors were observed in the SQL Server Errorlog. This can be done by looking up the –e startup parameter value for the SQL Server instance in question. If you are using SQL Server 2005 and above, then right-click on the Service Name in Configuration Manager –> Properties –> Advanced Tab. Alternatively, you can use a VBScript (or Powershell) to fetch this information. Here is an example for obtaining the Startup Parameters using WMI and VBScript.

Verify that the files pointed to by the –d (master data file location), –e (Errorlog location) and –l (master log file location) actually exist and have the correct file paths. Starting from SQL Server 2005, the Errorlog will have the startup parameters printed in it when the SQL instance starts up. If you find no errors in the SQL Errorlogs, look into the Windows Application Event logs and ensure that there are no errors related to the cluster resource for the SQL instance or the cluster group (applicable for Windows 2003 clusters) that the SQL resource is a part of.

If the Windows Application Event logs, don’t provide any useful errors, then look at any warnings or errors pertaining to the SQL Server resource in the Windows Cluster logs.

Refer the following post to generate the Cluster log on a Windows Server 2008 Failover cluster or higher. For a Windows Server 2003 Cluster, the default location of the cluster logs is: <system driver>:\<windows folder>\Cluster\Cluster.log.

Now try to start the SQL Server instance using command prompt using the sqlservr.exe to check if the service comes online without bringing any user database online using -T3608.

For a default SQL Server instance:

sqlservr.exe –c –m –T3608

For a named instance of SQL Server:

sqlservr.exe –c –m –T3608 –s<instance name>

If the above fails, then the console window should provide you with an error message as to why the SQL instance fails to start. If the above succeeds, then try and start the service using the Configuration Manager.

Note: Make sure that you are not falling prey to a Desktop Heap exhaustion issue.

If you determine that the SQL Server registry entries have an erroneous entry, then you would need to follow the steps mentioned in the KB Article below to rectify the problem (due to Registry Synch operations that happen on a cluster):
http://support.microsoft.com/kb/953504

You might also find that the SQL Server resource entries in the Cluster Registry hive (HKLM\Cluster\Resources\GUID) are missing. You would then have to re-create the same using the steps mentioned in the KB Article below (applicable for SQL Server 2000 and SQL Server 2005 only):
http://support.microsoft.com/kb/810056

NOTE: Remember that modifying registry keys incorrectly on a Windows Server can have disastrous effects. Make sure that you take a registry backup before making any registry modifications.

Apart from the following common startup failure root causes, there are an additional few which are applicable to clustered SQL Server instances only (in BOLD below):

The most common root causes for service startup failures are:

Access denied on the folders for the SQL Server Service Account on the DATA folder which contains the SQL Server system databases or due to some other process (like anti-virus holding a lock on the SQL database files)
Insufficient permissions granted to the SQL Server Service account.
File not found errors due to either accidental deletion of files or disk failures.
System database files having inconsistencies prevent SQL Server startup.
Password of the service account was changed but not updated on the server that has the SQL instance installed.
Startup parameters have incorrect file path locations
The dependent resources for the SQL Server clustered instance are failing to come online. (Eg. Shared disks on which the database files reside on)
Incorrect (read:messed up) Registry entries in the CLUSTER registry hive for the concerned SQL instance which have been checkpointed already and due to which the instance fails to start.
Operating system/Resource crunch/Environment related issues on one or more nodes due to which the instance comes online on one node but not on the other.

Additional references:
T-SQL Tuesday: Common SQL Server myths – Series I–Setup and Services
SQL SERVER 2000 CLuster FAILS TO COME ONLINE ON OTHER NODE
SQL SERVER 2005 RESOURCE FAILS TO COME ONLINE ON CLUSTER AFTER CHANGING THE SAN DRIVE TO ADD MORE DISK SPACE
Sql Server 2005 Resource Taking long time to come online on Windows Cluster with Resource Manager Creation Failed reported in Errorlog
Sql Server Resource Fails to come online on Cluster with Error Error Locating Server/Instance Specified [xFFFFFFFF]
SQL Server 2008 service fails to start, after applying Service Pack 1

Versions that this roadmap applies to:
SQL Server 2000, SQL Server 2005, SQL Server 2008, SQL Server 2008 R2

Technorati Tags: SQL Startup Failures,Database Engine,Failover Cluster

December 5, 2010

Wiki: SQL Database Engine Startup Failures

The startup failures for SQL Server can be classified into two types:

1. Service startup failures for stand-alone instances

2. Services startup failures for clustered instances

I shall document a road-map for a step-by-step troubleshooter for handling SQL Server startup failures for both stand-alone and clustered instances. By no means, is this the only way to troubleshoot startup failures but for a novice SQL Server user (hopefully, you don’t attempt the steps below for the first time on a production instance), this would be a good starting point in determining the root cause of the startup failures.

Services startup failures for stand-alone instances

If you are using SQL Server 2000, then try and start the SQL instance from the Services Manager and check if the service starts up. If you are using SQL Server 2005 or above, then try and start the service using SQL Server Configuration Manager.

If the service fails to start, locate the SQL Server Errorlog location and check if any errors were observed in the SQL Server Errorlog. This can be done by looking up the –e startup parameter value for the SQL Server instance in question. If you are using SQL Server 2005 and above, then right-click on the Service Name in Configuration Manager –> Properties –> Advanced Tab (Refer Pic for more details). If you want to act smart in front of your co-worker Smile with tongue out , then you can use a VBScript (or Powershell) to fetch this information. Here is an example for obtaining the Startup Parameters using WMI and VBScript.

Verify that the files pointed to by the –d (master data file location), –e (Errorlog location) and –l (master log file location) actually exist and have the correct file paths. Starting from SQL Server 2005, the Errorlog will have the startup parameters printed in it when the SQL instance starts up.

Once you have located the current ERRORLOG (file with the same name will be present by the same name in the LOG folder), check for any errors present in the SQL Server Errorlog. An example of an error would be:

2010-11-20 07:50:58.250 Server Error: 17113, Severity: 16, State: 1.
2010-11-20 07:50:58.250 Server Error 2(The system cannot find the file specified.) occurred while opening file ‘C:\Program Files\Microsoft SQL Server\MSSQL10_50.MSSQLSERVER\MSSQL\DATA\master.mdf’ to obtain configuration information at startup. An invalid startup option might have caused the error. Verify your startup options, and correct or remove them if necessary.

The above error message clearly indicates that the master.mdf file was not found in the location that it was expected to be found. If your system database files are not consistent, the SQL Server instance will not startup.

Note: If you see that the timestamp in the SQL Server Errorlog file doesn’t reflect the current system time when you attempted to start the SQL instance, then it is highly likely that the path to the SQL Errorlog is incorrect. If that is true, then the SQL Server instance will fail to start and a pretty unhelpful error message will be shown to you. It is always one of the two:

“The service failed to respond in a timely fashion”

OR

“The service failed to start”

The most common root causes for service startup failures are:

Access denied on the folders for the SQL Server Service Account on the DATA folder which contains the SQL Server system databases or due to some other process (like anti-virus holding a lock on the SQL database files)
Insufficient permissions granted to the SQL Server Service account.
File not found errors due to either accidental deletion of files or disk failures.
System database files having inconsistencies prevent SQL Server startup.
Password of the service account was changed but not updated on the server that has the SQL instance installed.
Startup parameters have incorrect file path locations.
The SQL Server service is disabled in Service Control Manager.

Find out if the SQL Server service account has all the necessary permissions required to start the SQL Server service. This would cover the ACLs, Security Privileges for the Windows Account, and Windows Registry permissions. Refer the Books Online article which documents all the permissions required by a SQL Server service account.

Even if that doesn’t help, then check the Windows System and Application Event logs and check for any errors related to the SQL Server instance or service that is currently failing to start.

Note: Make sure that you are not falling prey to a Desktop Heap exhaustion issue.

Still stuck!! Wow!! Aren’t we in a soup. No errors in the Errorlog (very very rare case scenario) or in the Windows Event Logs.

Then you need to try and start the service using command prompt in console mode. There is yet another Books Online article which explains how to start the SQL Server instance using sqlservr.exe located in the BINN folder.

Example for Default instance:

sqlservr.exe –c –m -T3608

Example for Named instance:

sqlservr.exe –c –m –s <instance name> -T3608

Check if the command prompt window spews out any helpful error message which will point you to the root cause of the issue. The Trace Flag 3608 prevents any database startup apart from the master database. This will help you determine if the SQL instance is starting up correctly and the problem might lie with the other system databases.

During my course of troubleshooting, I have also encountered some pretty weird startup failures. Of the top of my head, here are a few which I could find links to:

By Balmukund Lakhani (Blog)

http://blogs.msdn.com/b/blakhani/archive/2009/11/24/sql-server-2005-express-setup-failure-with-error-an-error-occurred-during-encryption.aspx

The famous (infuriating) TDSSNICLIENT client initialization failures during startup. The SQL Protocols team has documented a bunch of them here:

http://blogs.msdn.com/b/sql_protocols/archive/2006/04/28/585835.aspx

The case of the obnoxious locks value:

https://troubleshootingsql.com/2011/04/01/the-case-of-the-obnoxious-locks-value/

In case you have any more interesting startup failures that you find blog posts about, feel free to leave a comment for this post.

Startup Failures for Clustered Instances in another blog post.

Versions that this road-map applies to:

SQL Server 2000, SQL Server 2005, SQL Server 2008, SQL Server 2008 R2

Additional references:
T-SQL Tuesday: Common SQL Server myths – Series I–Setup and Services

Technorati Tags: SQL Startup Failures,Troubleshooting SQL Issues,SQLWikis