Month: November 2017

Script to retrieve database backup information

by Dev on

Backup script is something which is a very common need of a DBA in day to day life.This article will provide you with a base query to retrieve database backup information.

I am not able to give credit to anybody as I have seen lot of similar queries in the internet and so could not find the original writer.

Hmm! then what is the need of having another one in this blog? Well, I have experienced that sometimes a simple query suffice your need rather than a fancy one.

The below queries will help answering  the following questions:
  • Do you have recent database backups?
  • What type of backup is being taken for particular databases?
  • What is the backup size?
  • Who has taken the backup or which application login id is being used to take backup?
  • What is the Recovery model of the databases being backup up?
  • How long it took to backup?
Backup details of all the databases of all backup types:
SELECT server_name, database_name, 
CASE 
   WHEN type='D' THEN 'Full Backup' 
   WHEN type='L' THEN 'Log Backup'
   WHEN type='I' THEN 'Diff Backup' 
END AS BackupType, 
backup_start_date, backup_finish_date, DATEDIFF(SS,  backup_start_date, backup_finish_date)/60 AS [Duration(Minutes)],
CAST(backup_size/1024/1024 AS NUMERIC(10,2)) AS 'BackupSize(MB)', user_name AS UserName  
FROM msdb..backupset 
ORDER BY backup_finish_date DESC
The following screen shot is to show you an example output:

You can use this as a base and tweak it here and there as per your requirements. Few examples are given below.

Retrieve the same information for particular database(s) and for a particular type:
SELECT server_name, database_name, 
CASE 
   WHEN type='D' THEN 'Full Backup' 
   WHEN type='L' THEN 'Log Backup'
   WHEN type='I' THEN 'Diff Backup' 
END AS BackupType, 
backup_start_date, backup_finish_date, DATEDIFF(SS,  backup_start_date, backup_finish_date)/60 AS [Duration(Minutes)],
CAST(backup_size/1024/1024 AS NUMERIC(10,2)) AS 'BackupSize(MB)', user_name AS UserName  FROM msdb..backupset
WHERE database_name IN ('master','model','msdb')
AND type='D' 
ORDER BY backup_finish_date DESC
Include recovery model of the databases in the same output list:
SELECT b.server_name, b.database_name, 
CASE 
   WHEN type='D' THEN 'Full Backup' 
   WHEN type='L' THEN 'Log Backup'
   WHEN type='I' THEN 'Diff Backup' 
END AS BackupType, 
d.recovery_model_desc AS RecoveryModel,
b.backup_start_date, b.backup_finish_date, DATEDIFF(SS,  b.backup_start_date, b.backup_finish_date)/60 AS [Duration(Minutes)],
CAST(b.backup_size/1024/1024 AS NUMERIC(10,2)) AS 'BackupSize(MB)', b.user_name AS UserName  FROM msdb..backupset b
JOIN sys.databases d
ON b.database_name = d.name
WHERE database_name IN ('master','model','msdb')
AND type='D' 
ORDER BY backup_finish_date DESC

Please let me know if it helps by leaving your comments below.

Further more you can go through the below link for some addition scripts:

https://www.mssqltips.com/sqlservertip/1601/script-to-retrieve-sql-server-database-backup-history-and-no-backups/

You might also want to look at the below scripts:
Script to get Cluster Failover Time and Node Name
Script to List All Users with db_owner role in SQL 2000

Tagged as:

{ 2 Comments }

User Defined Profiler Template Location

by Dev on

First of all let me explain you why you need to know the user defined profiler template location.

In situations you may need to define a template and share it with other DBAs to run or vice versa. You have saved the profiler and request the other person to run by providing the name of the profiler. The other person called you (when you are not in shift) since he/she is not able to find the profiler you created, but that is hard for you to believe as you are quite sure you did it right.

Once you go through this article you will get to know:

  1. How to create a user defined template?
  2. Why others are not able to see the template?
  3. How to resolve this?

You can watch the following video demo which shows the stuffs I am going to explain here:

1. How to create a user defined template?

Launch SQL Server Management Studio (SSMS) and open SQL Server profiler:

Open a new Template:

Provide a template name and choose a base template from the drop down menu. In the below example I have chosen Standard (default) template. Then click on Events Selection tab.

In the above example I have not selected the check box for “Use as a default template for selected server type”. You can choose this option if you want your template to be default template.

Choose the Events as per your requirement. Click on “Show all events” and “Show all columns” to view all the events and columns to choose from. Then click on “Column Filters” to put filter for the trace.

In the below example I have put a filter for database id since I wanted to capture the transactions only for one database. The filter may differ for your requirement.

Once you are done with the filter selection click on OK and then Save the template.
Now if you go back and check the drop down list of templates you will find your custom template.
Go to File > New Trace and then connect to any SQL Instance

The following screen shot shows the newly created template “Demo Template (user)” in the drop down list.

2. Why others can’t see this template under the same drop down?

Here is the physical location of the user defined template file:

C:\Users\UserName\AppData\Roaming\Microsoft\SQL Profiler\11.0\Templates\Microsoft SQL Server\110

The main point to note here is the template gets created under the user profile folder of the login id, used to login to the server and that is why it is not visible to others.
AppData is a hidden folder and 11.0 and 110 will change depending upon the SQL Server version.

3. How to resolve this?

Let’s see the location of SQL Server Provided profiler templates.

C:\Program Files (x86)\Microsoft SQL Server\110\Tools\Profiler\Templates\Microsoft SQL Server\110

You guessed it right! You just need to copy the user defined template from
C:\Users\UserName\AppData\Roaming\Microsoft\SQL Profiler\11.0\Templates\Microsoft SQL Server\110
To
C:\Program Files (x86)\Microsoft SQL Server\110\Tools\Profiler\Templates\Microsoft SQL Server\110

Hope you enjoyed this article and feel free to share if you think it is helpful. Please let me know if you have any questions on this.

You might also want to check Myth around Remote Query Timeout Option

Tagged as:

{ Add a Comment }

Script to get SQL Cluster Failover Time and Node Name

by Dev on

A DBA’s life becomes easier if there is a script to get the required information. This article is about a script to get SQL Cluster Failover time and Node Name prior to failover and little more.

You can go through the following link to know what and how to check the same information in SQL Server Error log. I have also explained what all other related information can be found. The output of the below query will make more sense if you read the article first.

SQL Cluster Instance Failover Time and Node Name prior to Failover

After writing the above article I thought of working on the script to scan through all the error log files. This way I can find not only the latest SQL Cluster failover time and Node name but also history of failover information. This may help in situations where you are facing frequent failover issues or you need a failover history report.

CREATE TABLE #ErrorLog(
   LogDate DATETIME,
   ErrorSource NVARCHAR(MAX),
   ErrorMessage NVARCHAR(MAX)
)

CREATE TABLE #NumberOfLogs(
   ID INT PRIMARY KEY NOT NULL,
   LogDate DATETIME NOT NULL,
   LogFileSize bigint
)

INSERT INTO #NumberOfLogs(ID,LogDate,LogFileSize)
EXEC master.dbo.xp_enumerrorlogs

DECLARE @ErrorLogID INT

DECLARE cNumberOfLogs CURSOR FOR
   SELECT ID
   FROM #NumberOfLogs

OPEN cNumberOfLogs
FETCH NEXT FROM cNumberOfLogs INTO @ErrorLogID
   WHILE @@FETCH_STATUS = 0
   
   BEGIN
       INSERT INTO #ErrorLog(LogDate,ErrorSource,ErrorMessage)
       EXEC sp_readerrorlog @ErrorLogID, 1, 'NETBIOS'
        
       INSERT INTO #ErrorLog(LogDate,ErrorSource,ErrorMessage)
       EXEC sp_readerrorlog @ErrorLogID, 1, 'SQL Server is terminating'
       FETCH NEXT FROM cNumberOfLogs INTO @ErrorLogID
   END 
   
CLOSE cNumberOfLogs
DEALLOCATE cNumberOfLogs

SELECT LogDate, ErrorMessage FROM #ErrorLog

DROP TABLE #ErrorLog
DROP TABLE #NumberOfLogs
The below screen shot shows a sample result set:

If this is helpful, please let me know by leaving your comments below. I would love to hear from you if you have any better way to get the same information.

You may also like to check the  Script to List All Users with db_owner role in SQL 2000.

{ Add a Comment }

SQL Cluster Instance Failover Time and Node Name prior to Failover

by Dev on

You may have come across situations where SQL Server cluster instance has failed over from one node to other and you were asked to find the SQL Cluster Instance failover time and node name prior to failover.

Who can ask this question? Your approver, manager or colleague anyone can ask you for different reasons, to know since when the instance is running on wrong/passive node, what alert system you have and it worked properly or not etc.

Whatever may be the reason, you want to find it quickly. In this article I am going to explain how to find the required information from SQL error log.

You need to search for the messages containing the text NETBIOS in the error log files.

Launch SSMS (SQL Server Management Studio) and start with the first error log to view as shown below:

Click on Filter and type in “netbios” against the field “Message Contains Text”. Mark the check box for Apply filter and click on OK as shown below. The message you are searching for, can be typed in Capital/Lower case.

Start checking each error logs until you find the required information as shown in the following screen shot.

The Message is like:

“The NETBIOS name of the local node that is running the server is ‘NodeName’. This is an informational message only. No user action is required.”

The message provides the information about the current Node, hosting the SQL Server Instance and the time it came up on the Node. Hence you need to keep checking the prior error logs which will show a different Node name and time in case it failed over from other node.

In the above example it shows the particular SQL Instance was running on Node “N02” prior to the failover.

The following information you can find using the NETBIOS filter:
  1. What is the current node? (N01)
  2. Is failover happened? Yes, as it shows a different node (N02) in prior error log file.
  3. Which node it was running prior to failover?(N02)
  4. What time failover started (Between 9/9/2017 11:28:10 PM the instance was on N02 and 9/9/2017 11:47:07 PM the instance started on No1. Check the error log around the time removing the filter and you will be able to find the exact time as shown in the below screen shot.)


You can also find what time SQL Instance came online/ what is the last SQL restart time. Of course you can find it from tempdb creation time? (Date 9/9/2017 11:47:07 PM the instance started on No1).

Hope this helps and I would like to hear from you if you have any other ways to find the same information quickly.

You may also want to check the following articles:
Myth around Remote Query Timeout Option
All Logins got disabled due to trigger execution

Tagged as:

{ Add a Comment }