Will you run out disk space capacity for your databases or/and backups in the near future?

These questions are easy to answer if you have a SQL monitoring system in place or if you use DBInsight’s SQL monitoring service (SQLRDS).

But what if you don’t have the monitoring tools in place or utilise a service like DBInsight’s monitoring service (SQLRDS)?

Well, all is not lost if you have the SQL Server Error Log files that cover at least a few weeks to a few months. Here’s how to measure your database growth using your SQL Server Error logs.

Note that every time a full database backup is created a message is written to the SQL Server like below.

Database backed up. Database: XXXXX, creation date(time): 2019/10/10(13:29:22),
pages dumped: 2159831

The relevant component of the message is the pages dumped value. So, if we can harvest the pages dump value over a number weeks or months then we have a ready made indicator of database growth.

The pages dumped can be converted to a MB or GB value – note 1 page = 8KB. The calculation for above is

2159831 pages x 8KB = 17278648 KB or 16873 MB or 16.4 GB

You can use a Powershell script to harvest all pages dumped values for each database in all of your SQL Server Error log files

Typically, I would copy the SQL Server error logs into a temporary folder and then point the Powershell script to that folder. It will produce a CSV file like below, showing the database name, date of full backup and pages dump.

Here’s a script that I use.

#Database backed up. Database: XXXXXXXX, creation date(time): 2019/08/14(17:11:47), pages dumped: 409252

$ErrorlogDir = “C:\temp \errorlogs”

# Get Errorlog files in $ErrorlogDir
$ErrorLogs =  Get-ChildItem $ErrorlogDir | Where-Object {$_.name -like “*errorlog*” } 

foreach ($File in $ErrorLogs)

{

  $ErrFile = $File.fullname

  Write-Host “–> Processing $ErrFile”

  # Parse the error log file

  get-content $ErrFile | foreach-object {if ($_ -like “*Database backed up. Database: *”) `

            {$_.substring($_.indexof(“Database backed up. Database: “)+30, $_.indexof(“creation date(time): “)-$_.indexof(“Database backed up. Database: “)-32) + “,” `

                                                + $_.substring(0,10) + “,” `

                                                + $_.substring($_.indexof(“pages dumped: “)+14, $_.indexof(“first LSN:”)-$_.indexof(“pages dumped: “)-16) } } | add-content “C:\temp\backupinfo.csv”                                      

}

You then open the CSV file in Excel and perform the Pages to MB or GB calculation, from there you can use Excel to predict the future database data file size over the desired period.

The post SQL Server Database Space Monitoring and Forecasting appeared first on SQLRDS Monitor.

Like most monitoring systems these days the SQLRDS monitor has a number of availability group (AG) health checks and alerts built in. However, what if you don’t have a monitoring system in place but you have deployed an availability group or BAG (if running Standard Edition).

Checking the health of the AG on a continuous basis is very important, otherwise, you may find your environment is not in a state to failover at that crucial moment.

You can still get alerts about your AG health and status, even if you are not using a full feature monitoring system like SQLRDS Monitor. This simple solution is based on the Database Mail component to send out alerts that are generated by :-

• An SQL Agent job that runs an AG health check script every 15 minutes. The script will alert if the AG is not healthy or if the Log Send Queue is greater than a threshold (20000 KB).
• SQL Server Alerts that checks for the change of AG roles i.e. a failover has occurred.

AG Health Script

Create a SQL Agent job on each server of the Availability Group configuration and run the job every 15 minutes. Configure the AG Health Check script below as a TSQL Job Step.

— Configure values

Declare @ToMail varchar(max) = ‘<Your Email Address>‘

Declare @DBMailprofile varchar(128) = ‘<Your Database Mail Profile Name>‘

— Standard Declares

Declare @ErrorMsg varchar(8000) = ”

Declare @QueryStr nvarchar(max) = ”

Declare @ag_replica_server varchar(128) = ”

Declare @database_name varchar(128) = ”

set nocount on

— Get Health Data

SELECT ag.name AS ag_name, ar.replica_server_name AS ag_replica_server,

ag_replica_role = CASE

WHEN ar_state.role_desc IS NULL THEN N’DISCONNECTED’

ELSE ar_state.role_desc

END, db_name(dr_state.database_id) as database_name,

dr_state.synchronization_state_desc, dr_state.is_suspended, dr_state.synchronization_health_desc,

dr_state.suspend_reason_desc, isnull(dr_state.secondary_lag_seconds,0) secondary_lag_seconds,

isnull(dr_state.log_send_queue_size,0) log_send_queue_size, isnull(dr_state.redo_queue_size,0) redo_queue_size

into #AGStatus

FROM (( sys.availability_groups AS ag JOIN sys.availability_replicas AS ar ON ag.group_id = ar.group_id )

JOIN sys.dm_hadr_availability_replica_states AS ar_state  ON ar.replica_id = ar_state.replica_id)

JOIN sys.dm_hadr_database_replica_states dr_state on

ag.group_id = dr_state.group_id and dr_state.replica_id = ar_state.replica_id

where ar_state.is_local = 1 — check local server only

— used to put output in DBMail alert

Set @QueryStr= ‘SELECT ag.name AS ag_name, ar.replica_server_name AS ag_replica_server,

ag_replica_role = CASE WHEN ar_state.role_desc IS NULL THEN N”DISCONNECTED” ELSE ar_state.role_desc END,

db_name(dr_state.database_id) as database_name, dr_state.synchronization_state_desc, dr_state.is_suspended,

dr_state.synchronization_health_desc, dr_state.suspend_reason_desc, isnull(dr_state.secondary_lag_seconds,0) secondary_lag_seconds,

isnull(dr_state.log_send_queue_size,0) log_send_queue_size, isnull(dr_state.redo_queue_size,0) redo_queue_size

FROM (( sys.availability_groups AS ag JOIN sys.availability_replicas AS ar ON ag.group_id = ar.group_id )

JOIN sys.dm_hadr_availability_replica_states AS ar_state  ON ar.replica_id = ar_state.replica_id)

JOIN sys.dm_hadr_database_replica_states dr_state on ag.group_id = dr_state.group_id and dr_state.replica_id = ar_state.replica_id where ar_state.is_local = 1′

If exists (select * from #AGStatus where synchronization_health_desc <> ‘HEALTHY’ or synchronization_state_desc <> ‘SYNCHRONIZED’ or is_suspended = 1)

begin

   select top 1 @database_name = [database_name],  @ag_replica_server =  ag_replica_server

              from #AGStatus

   Set @ErrorMsg = ‘AG Issue Found – Databases on ‘ + @ag_replica_server + ‘ are NOT Healthy’

   EXEC msdb.dbo.sp_send_dbmail @profile_name = @DBMailprofile, @recipients = @ToMail, @subject = @ErrorMsg, @Query = @QueryStr, @attach_query_result_as_file = 0, @query_result_width = 1024

end

— Check log send queue size – if bigger than 20MB then alert

If exists (select * from #AGStatus where log_send_queue_size >= 20000  )

begin

   select top 1 @database_name = [database_name],  @ag_replica_server =  ag_replica_server

              from #AGStatus

   Set @ErrorMsg = ‘AG Issue Found – Databases on ‘ + @ag_replica_server + ‘ have a LOG SEND QUEUE >= 20000’

   EXEC msdb.dbo.sp_send_dbmail @profile_name = @DBMailprofile, @recipients = @ToMail, @subject = @ErrorMsg, @Query = @QueryStr , @attach_query_result_as_file = 0, @query_result_width = 1024

end

— End AG Health Check

SQL Server Alert Definitions

The following script will define 3 x SQL Server Alerts to determine if an AG role change has occurred or if data movement was suspended or resumed. The script also adds the Alerts to the specified SQL Operator for email notifications.

Note Database Mail, the Alerts and SQL Operator need to be configured on each server of the AG configuration.

— 1480 – AG Role Change (failover)

EXEC msdb.dbo.sp_add_alert

       @name = N’AG Role Change’,

       @message_id = 1480,

    @severity = 0,

    @enabled = 1,

    @delay_between_responses = 60,

    @include_event_description_in = 1;

GO

EXEC msdb.dbo.sp_add_notification

       @alert_name = N’AG Role Change’,

       @operator_name = N’<YourSQLOperator>‘,

       @notification_method = 1;

GO

— 35264 – AG Data Movement – Suspended

EXEC msdb.dbo.sp_add_alert

       @name = N’AG Data Movement – Suspended’,

       @message_id = 35264,

    @severity = 0,

    @enabled = 1,

   @delay_between_responses = 60,

    @include_event_description_in = 1;

GO

EXEC msdb.dbo.sp_add_notification

       @alert_name = N’AG Data Movement – Suspended’,

       @operator_name = N’<YourSQLOperator>‘,

       @notification_method = 1;

GO

— 35265 – AG Data Movement – Resumed

EXEC msdb.dbo.sp_add_alert

       @name = N’AG Data Movement – Resumed’,

       @message_id = 35265,

    @severity = 0,

    @enabled = 1,

    @delay_between_responses = 60,

    @include_event_description_in = 1;

GO

EXEC msdb.dbo.sp_add_notification

       @alert_name = N’AG Data Movement – Resumed’,

       @operator_name = N’<YourSQLOperator>‘,

       @notification_method = 1;

GO

The post Availability Group Heath Monitoring appeared first on SQLRDS Monitor.

Ever encountered a GB+ sized SSAS log file, well this Powershell command provides a quick way to check the tail of the log.

I’m showing the last 500 lines

Enter Powershell prompt, then run
get-content <yourpath>\OLAP\Log\msmdsrv.log -tail 500

Got some locking going on? Connect to SSAS via SSMS and run the following query.

select * from $system.discover_locks

You may need to know the commands and sessions also

select * from $system.discover_sessions

select * from $system.discover_connections

select * from $system.discover_commands

The below query combines some of the internal views for a more convenient check and allows you to run from a standard Query Window connected to a SQL Server Engine.

Create a linked server on a SQL Server, replace the linked server name in the definition and then run the below query to create the linked server.
EXEC master.dbo.sp_addlinkedserver @server = N'SSASCheck', @srvproduct='MSOLAP', @provider=N'MSOLAP',
@datasrc=N'<Your SSAS server name>'

Then run the Openquery statement below to combine data from some of the relevant SSAS internal views.

SELECT ssasSessions.SESSION_SPID AS [SPID],

ssasSessions.SESSION_USER_NAME AS [User ID],

ssasSessions.SESSION_CURRENT_DATABASE AS [Database],

ssasConnections.connection_host_name AS ‘Client IP address’,

ROUND(CAST(ssasCommands.COMMAND_CPU_TIME_MS AS int)/1000, 2) AS [Command Time(in sec)],

ssasCommands.COMMAND_START_TIME AS [Command Start Time],

ssasCommands.COMMAND_END_TIME AS [Command End Time],

ssasCommands.COMMAND_TEXT AS [MDX Command],

connection_host_application AS ‘Application’,

CHARINDEX(”,ssasCommands.COMMAND_TEXT) AS ProcessRunning

FROM OPENQUERY( SSASCheck ,’select * from $system.discover_sessions’) AS ssasSessions

LEFT JOIN OPENQUERY(SSASCheck,’select * from $system.discover_commands’) AS ssasCommands ON ssasSessions.SESSION_SPID = ssasCommands.SESSION_SPID

LEFT JOIN OPENQUERY( SSASCheck ,’select * from $system.discover_connections’) AS ssasConnections ON ssasSessions.SESSION_connection_id = ssasConnections.connection_id

ORDER BY [Command Start Time] DESC

The post Analysis Services Diagnostics appeared first on SQLRDS Monitor.