Monthly Archives: May 2014

Setting Up Your Custom Data Collector Set

Problem:

From my previous post, we learned how to set up Data Collection in SQL Server. This is most suitable for monitoring purposes and reporting on disk usage and server activities. But what if you want to monitor other areas of your database server aside from the default data collection sets?

Solution:

Aside from the default data collection sets, you can also set up your own customised data collection sets. For example your requirement would be to monitor all failed SQL Agent jobs on a daily basis across all your SQL database servers. Lets use this as an example.

In creating a customised data collection set, the step is composed of 3 parts;

1. defining the data collection container – which contains header parameters such as name of the data collection set, description, logging and schedule to run.  You will need to use the stored procedure sp_syscollector_create_collection_set. See example below;

EXEC [msdb].[dbo].[sp_syscollector_create_collection_set]
                                    @name=N’Failed SQL Jobs’,
                                    @collection_mode=1,
                                    @description=N’Collects data about failed jobs for all servers.’,
                                    @logging_level=0,
                                    @days_until_expiration=180,
                                    @schedule_name=N’CollectorSchedule_Every_6h’,
                                    @collection_set_id=@collection_set_id_1 OUTPUT,
                                    @collection_set_uid=@collection_set_uid_2 OUTPUT

2. defining the data collector type – this is the part you will define the data collector type for your custom data collection. There are predefined data collector type already setup for use. For most usual cases we will use the Generic T-SQL Query Collector Type.   To retrieve this value you may use the statement below;

Declare @collector_type_uid_3 uniqueidentifier
                  Select @collector_type_uid_3 = collector_type_uid
                  From [msdb].[dbo].[syscollector_collector_types]
                  Where name = N’Generic T-SQL Query Collector Type’;

3. Define the data collection item – this is the part wherein you define your parameters for your data collection set.  This will also contain the actual query to retrieve all SQL failed jobs. As an example see the code below

Declare @collection_item_id_4 int
EXEC [msdb].[dbo].[sp_syscollector_create_collection_item]
                                    @name=N’Failed SQL Jobs Item’,
                                    @parameters=N'<ns:TSQLQueryCollector xmlns:ns=”DataCollectorType”><Query><Value>
                                    SELECT  @@ServerName AS [ServerName],
                                                      [sJOB].[name] AS [JobName],
                                                      [sJOBH].[run_date]  AS [LastRunDateTime],
                                                      CAST([sJOBH].[run_duration] AS VARCHAR(6)) AS [LastRunDuration (HH:MM:SS)],
                                                      [sJOBH].[message] AS [LastRunStatusMessage],
                                                      CAST([sJOBSCH].[NextRunDate] AS CHAR(8)),
                                                      [sJOBSCH].[NextRunDate] AS [NextRunDateTime]
                                    FROM [msdb].[dbo].[sysjobs] AS [sJOB] LEFT JOIN (SELECT [job_id], MIN([next_run_date]) AS [NextRunDate]
                    , MIN([next_run_time]) AS [NextRunTime]
                FROM [msdb].[dbo].[sysjobschedules]
                GROUP BY [job_id] ) AS [sJOBSCH] ON [sJOB].[job_id] = [sJOBSCH].[job_id]
                 LEFT JOIN (SELECT [job_id] , [run_date] , [run_time]          , [run_status], [run_duration], [message], ROW_NUMBER() OVER (PARTITION BY [job_id] ORDER BY [run_date] DESC, [run_time] DESC) AS RowNumber FROM [msdb].[dbo].[sysjobhistory] WHERE [step_id] = 0 ) AS [sJOBH] ON [sJOB].[job_id] = [sJOBH].[job_id] AND [sJOBH].[RowNumber] = 1    WHERE  [sJOBH].[run_status] = ”0”                                    ORDER BY [LastRunDateTime] DESC

                        </Value><OutputTable> FailedJobs</OutputTable> </Query><Databases UseSystemDatabases=”true” UseUserDatabases=”true” /> </ns:TSQLQueryCollector>’, 
                                    @collection_item_id=@collection_item_id_4 OUTPUT,
                                    @frequency=60,
                                    @collection_set_id=@collection_set_id_1,
                                    @collector_type_uid=@collector_type_uid_3;

The <OutputTable> tag contains the destination table for your query results. This table will be created with default schema of custom_snapshots inside your Management Data Warehouse database.

After successfully creating your custom data collection set, you must manually start the data collection. You can either do this via GUI from the Data Collection menu or by executing below under the msdb database.

EXEC sp_syscollector_start_collection_set @collection_set_id = <yourcollectionsetid>;

After the initial upload of data thats the time the output table for your data collection set will be created. Now that you have the data for all failed SQL Agent jobs across your database servers on a daily basis, you can create a report for this via Reporting Services and added to your monitoring reports.

There you have it, your first customised data collection set. From here on you can create more monitoring reports for your dashboard and impress your team lead or boss. 🙂

 

Advertisements

Developer’s Life – Every Developer is a Spiderman


Data Collection in SQL Server

Problem : As a DBA, one of your primary job description is to handle management and administration of 1 or 100+ SQL Server databases. One of the top skill required for this is being proactive in your managing task and being able to fix something even before it even happens. For this you will need a tool to help in accomplishing this. Good thing in SQL Server it is already available for you to use. Ease to setup and aside from its default reporting you can add your own custom data collection set.

Solution : There is a built in tool for this in SQL Server called Data Collection Set. The idea for this you have data warehouse repository that collects data from all your SQL Servers and generates a report on the current state of your SQL Servers.

If you are Formula 1 Grand Prix fanatic like me, imagine your engineering team collecting all kinds of data from your car and presents it in a dashboard format of reporting so that they will know how to fix the car so you, the driver can come out on top of the race. It is the similar principle here for SQL Server.

Setting up your Data Collector

For configuring your data collector server, you need to configure a server that will act as the data collector and also generate the management data warehouse report. Also you will need a service account for data collection and grant this account data collection roles.

Configure the Management Data Warehouse

In your SQL Server that will act as the data collector, you need to setup the Management Data Warehouse database.

From your SSMS, go to Management menu and right click on Data Collection.

DataCollectionMenu

Select the Configure Data Management Warehouse. This will launch the Configure Data Management Warehouse wizard.  Click on Create or upgrade a management data warehouse option. Click Next.

ConfigureMDWWizard

Enter the server name, that will act as your data collector or the SQL Server on which you are running this wizard. Enter the database name that will act as the repository database from which your report will gather data. For the cache directory, this is optional and for cases when you want your data collector server to collect data but not yet upload to the MDW server.

ConfigureMDWStorage

That’s all that is needed and you can click Finish. Click Close.

ConfigureMDWProgress

Refresh your Data Collection Menu. You should be able to see the following default data collection sets configured.

ConfigureAfter

Each data collection set will automatically create a job in your SQL Agent. You may see this  starting with job name ‘collection_setxxxxxxx’.

At this point, you have completed the setup of your Data Collection in SQL Server. Now to configure the properties of your data collection sets, you right click on each data collection set and select Properties.

DataCollectionSetProperties

From the Properties dialog window, you can configure the schedule of data collection, the retention period of the data collected

Data Collector Permissions and Roles

There are 3 database roles within the MDW database. The service accounts from the database server that have client data collector configured must be added to the logins and mapped to one of the following database role.

  • mdw_admin – members of this role have rights to manage the data collection set properties. Members of this role also have read and write access to the MDW database.
  • mdw_reader – members of this role have read access the MDW database
  • mdw_writer – members of this role have write access the MDW database

From your client database server that is sending the data for data collection set, the least privilege must be granted. The service account of the client database server must be added to the MDW server and must be added to the database role mdw_writer.

Data Collector SQL Server Clients

For all your SQL Server that you want to collect data from, you will need to run the same configuration wizard. And you need to add the SQL Server service account from this client server to your MDW server and make it a member of mdw_writer in the MDW database.

From your client SQL Server, select the option ‘Set up Data Collection’. Click Next.

ConfigureMDWWizard

In the below wizard dialog box, enter the name of your data collector server and the name of your repository database.

ConfigureMDWStorage

Click Finish and you are set. You will need to this for all the SQL Servers that you want to monitor.

Built in Reports and Dashboard

You may have a hard time finding the built-in reports for MDW because they must be used once before they automatically appear in the SSMS menu.

To get started:

  • Open SSMS and go to one of the MDW databases.
  • Right-click and select the Reports menu item.
  • Select Management Data Warehouse menu item.
  • Select Management Data Warehouse Overview menu item.

You should be able to see main reporting dashboard below. The below dashboard contains the server or instance name for all your client SQL Servers and the last data collection date for the 3 default data collector sets

MDWDefault

 

Click on each column to view the different dashboard for your default data collection set.

I have been using the Data Collection only recently but i find it useful for SQL Server monitoring especially the Disk Usage report and dashboard. It really helps on forecast and disk usage trending of all the servers that I am monitoring. I hope you find it useful in your case too.

Aside from the default data collection set that SQL Server automatically creates, you can create your own custom data collection set. There maybe a future requirement for all failed SQL Agent jobs report in your environment or something like an auditing report for all your logins across all your SQL Server Production environment. This can be created too and I will tackle that on my next post!

For more information on data collection, you may visit this link.

 

 


That “Cannot Generate SSPI Context” headache

Problem : You have completed an SQL Server setup on a new instance. The setup is successful, you can connect locally and and everything is fine. No headaches. Now to test your SQL Server, you try to connect from a client PC and you get that “Cannot Generate SSPI Context”  error. Now what?

Most likely you have the same situation as below;

  • You are using Integrated Security.
  • Kerberos authentication is used to perform the security delegation. You can verify this with your domain admin.
  • You are not using Local System account in your SQL Server service which is a good practice.

Solution:

What SSPI?

Security Support Provider Interface (SSPI) is a set of Windows APIs that allows for delegation and mutual authentication over TCP/IP sockets. Therefore, SSPI allows for a computer that is running a Windows operating system to securely delegate a user security token from one computer to another over any transport layer that can transmit raw bytes of data.

What SPN?

Another term that we have to define here is SPN. SPN stands for Service Principal Name. Kerberos authentication uses an identifier named “Service Principal Name” (SPN). Consider an SPN as a domain or forest unique identifier of some instance in a server resource. You can have an SPN for a web service, for a SQL service, or for an SMTP service. You can have a multiple SPNs if you have multiple instance sitting on your database server.

Setting SPN

If you are using a domain user account for your SQL Server service, you need to manually register the SPN for this instance to create the SPN in your Active Directory. The SPN for SQL Server has the following elements;

  • ServiceClass: This identifies the general class of service. This is always MSSQLSvc for SQL Server.
  • Host: This is the fully qualified domain name DNS of the computer hosting SQL Server.
  • Port: This is the port number that the service is listening on. Unless you have made changes to the default port during your setup, this value is 1433.

To register the SPN for the domain user account in your Active Directory, you may run the following command in your cmd window;

SETSPN -A MSSQLSvc/<SQLServerName>:1433 domain\useraccount

The trick here is that you need to issue this command twice. One for the NETBIOS name, which is the above command, and one for the FQDN (Fully Qualified Domain Name) to allow for the setup to succeed.

So you will have to run also the below as your next step.

SETSPN -A MSSQLSvc/<SQLServerName.domain.com>:1433 domain\useraccount

For multiple instances, you will then need to run it twice for each instance that you have. There are times, when you will need the help of your friendly domain admin to run the setspn command because your current permission is not enough. Also to make sure this is successful, you will need to check that the SQL Browser service is running during the setup, because it will be looking for it (especially for multiple instances) and validate it.

To validate that the setspn command was successful, in your cmd window;

setspn – L domain\useraccount

This should list all the instances you have in your server.

One less headache… 🙂

 


Allowing your App team to run SQL Agent Jobs

Problem : In most of your projects, there is always that situation that the application team have that need to run their own application jobs in the SQL Agent. But how do you allow them to let them have that right (and fun) without letting them make any changes to the SQL Agent jobs?

Solution : DBA can allow this by granting them certain permissions and making them members of any 3 SQL Agent database roles in the system database msdb.

SQL Agent fixed database roles

There are 3 SQL Agent fixed database roles in msdb namely;

  • SQLAgentOperatorRole – Of all the 3 fixed database roles, this is the most privileged role. Members of this role can manage and run jobs that are owned by them or owned by other accounts. They also have access and enumerate proxies and alerts and delete job history. They automatically inherit the privileges of SQLAgentReaderRole and SQLAgentUserRole.
  • SQLAgentReaderRole – Members of this role can list all multi server jobs, their properties and their history. and not those jobs and schedules that they own. Members of this role automatically inherit the permissions for SQLAgentUserRole.
  • SQLAgentUserRole – Members of this role has the least privilege of the 3 fixed roles. They are allowed to view local jobs, operators and schedules. They only have access to local jobs that they own.

Aside from making the App team members of any of the 3 fixed database roles in msdb database, you will need to grant them permission to execute in some Agent extended stored procedures.

To check if the Agent extended procedures are enabled in your system,

select value, value_in_use from sys.configurations where name = ‘Agent XPs’

The Agent extended procedures that I needed to grant them viewing of the Agent jobs and schedules under SQL Agent menu in SSMS are the following.

use master
go
grant select on master.dbo.sysperfinfo to [YourAppTeamGroup]
go
grant execute on master.dbo.xp_sqlagent_notify to [YourAppTeamGroup]
go
grant execute on master.dbo.xp_sqlagent_enum_jobs to [YourAppTeamGroup]
go
grant execute on master.dbo.xp_sqlagent_param to [YourAppTeamGroup]
go
grant execute on master.dbo.xp_sqlagent_is_starting to [YourAppTeamGroup]
go
grant execute on master.dbo.xp_instance_regenumvalues to [YourAppTeamGroup]
go

After executing the above statement, your App team should be able to see the SQL Agent GUI, view their jobs and job properties, but not make any changes in them.

Now, you have less complains and you made the App team happy!

 


SHE GOES GLOBAL

Travel the world, change your life.

My Time to Travel

The travels of an old(er), solo, woman

Blog Home for MSSQLDUDE

The life of a data geek

unfoldthecreativity

Traveller Observer

The SQL Pro

(Ayman El-Ghazali - www.thesqlpro.com)

Meels on Wheels

Meels for breakfast, lunch and dinner.

juliansherman.net/

Building A Business While Having A Life

Paul Turley's SQL Server BI Blog

sharing my experiences with the Microsoft data platform, SQL Server BI, Data Modeling, SSAS Design, Power Pivot, Power BI, SSRS Advanced Design, Power BI, Dashboards & Visualization since 2009

TroubleshootingSQL

Explaining the bits and bytes of SQL Server and Azure

Coding Tips

We are crazy about programming and we want to share our craziness with you!!!

SQL Studies

 Live, Learn, Share

Sql And Me

My Experiments with SQLServer

Dimitrios Kalemis

I am exactly like Jesus Christ: an atheist and anarchist against society and bad people with influence and power (judges, social workers, politicians, priests, and teachers).

Clint Huijbers' Blog

Senior Certified Microsoft BI Consultant

Blog of Many Useless Wonders

Where Uselessness Abounds!

Steve Spevack's Blog

IT Recruitment