Spotlight on SQL Server 7.0. Reporting and Trending Guide

Transcription

1 Spotlight on SQL Server 7.0 Reporting and Trending Guide

2 2010 Quest Software, Inc. ALL RIGHTS RESERVED. This guide contains proprietary information protected by copyright. The software described in this guide is furnished under a software license or nondisclosure agreement. This software may be used or copied only in accordance with the terms of the applicable agreement. No part of this guide may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording for any purpose other than the purchaser s personal use without the written permission of Quest Software, Inc. The information in this document is provided in connection with Quest products. No license, express or implied, by estoppel or otherwise, to any intellectual property right is granted by this document or in connection with the sale of Quest products. EXCEPT AS SET FORTH IN QUEST'S TERMS AND CONDITIONS AS SPECIFIED IN THE LICENSE AGREEMENT FOR THIS PRODUCT, QUEST ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY WARRANTY RELATING TO ITS PRODUCTS INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON- INFRINGEMENT. IN NO EVENT SHALL QUEST BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDENTAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS OF PROFITS, BUSINESS INTERRUPTION OR LOSS OF INFORMATION) ARISING OUT OF THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF QUEST HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Quest makes no representations or warranties with respect to the accuracy or completeness of the contents of this document and reserves the right to make changes to specifications and product descriptions at any time without notice. Quest does not make any commitment to update the information contained in this document. If you have any questions regarding your potential use of this material, contact: Quest Software World Headquarters LEGAL Dept 5 Polaris Way Aliso Viejo, CA [email protected] Refer to our web site for regional and international office information. Trademarks Quest, Quest Software, the Quest Software logo, AccessManager, ActiveRoles, Aelita, Akonix, Benchmark Factory, Big Brother, BridgeAccess, BridgeAutoEscalate, BridgeSearch, BridgeTrak, BusinessInsight, ChangeAuditor, CI Discovery, Defender, DeployDirector, Desktop Authority, Directory Analyzer, Directory Troubleshooter, DS Analyzer, DS Expert, Foglight, GPOADmin, Help Desk Authority, Imceda, IntelliProfile, InTrust, Invirtus, itoken, JClass, JProbe, LeccoTech, LiteSpeed, LiveReorg, LogADmin, MessageStats, Monosphere, NBSpool, NetBase, NetControl, Npulse, NetPro, PassGo, PerformaSure, Point, Click, Done!, Quest vtoolkit, Quest vworkspace, ReportADmin, RestoreADmin, ScriptLogic, SelfServiceADmin, SharePlex, Sitraka, SmartAlarm, Spotlight, SQL Navigator, SQL Watch, SQLab, Stat, StealthCollect, Storage Horizon, Tag and Follow, Toad, T.O.A.D., Toad World, vautomator, vconverter, vecoshell, VESI,vFoglight, vpackager, vranger, vspotlight, vstream, vtoad, Vintela, Virtual DBA, VizionCore, Vizioncore vautomation Suite, Vizioncore vessentials, Vizioncore vworkflow, WebDefender, Webthority, Xaffire, and XRT are trademarks and registered trademarks of Quest Software, Inc in the United

3 States of America and other countries. Other trademarks and registered trademarks are property of their respective owners. Spotlight on SQL Server 7.0 Reporting and Trending Guide April 2010

4 Table of Contents Spotlight Reporting and Trending 6 About Spotlight Reporting and Trending 6 Views 7 About Views 7 Group Views 8 Export Views 9 Reports 10 About Reports 10 Use Your Own Reports 11 Example Reports 11 Deploy Reports 13 Specify the Location of Your Web Reports Server 14 View Web Reports in Spotlight Schema of the 16 How the works 18 Store Data in the 20 Query the 21 Query Custom Counters 26 Query Alarm Data 27 Table Definitions 29 Stored Procedures 33 Specifications 36 Maintenance Plan for and Playback Database 43

5 Table of Contents 5 Appendix: Contact Quest 45 Contact Quest Support 45 Contact Quest Software 46 About Quest Software 46 Index 47

6 1 Spotlight Reporting and Trending About Spotlight Reporting and Trending In Spotlight Reporting and Trending, you can access views and reports to assist with long-term trend analysis and associated reporting. By configuring the Diagnostic Server to store statistics in the, you can retrieve the data via views and reports Views display data stored in the in the form of charts and tables. The charts and tables are grouped into logical categories that help you analyze the collected data. Views are different to drilldowns in that they show long-term data collected over time, not real-time data. Reports also display data stored in the and are displayed in the Spotlight Report Viewer. These reports are designed to provide you with a starting point for meeting your reporting requirements and can be modified to suit your needs. To view the data collected by Spotlight in the Reporting and Trending portal, you must configure the. See the Spotlight on SQL Server Deployment Guide for more information. Note: It is important to implement a good maintenance routine to ensure the Spotlight Statistics Repository is optimized and free of inconsistencies. See "Maintenance Plan for Spotlight Statistics Repository and Playback Database" (page 43) for more information.

7 2 Views About Views Use the Views in Reporting and Trending when you want to analyze and investigate longterm SQL Server data for planning purposes, to identify trends, or to determine potential performance issues. Views display data stored in the in the form of charts and tables. The charts and tables are grouped into logical categories, called Views, that help you analyze the collected data. Views are different to drilldowns in that they show long-term data collected over time, not real-time data. What you can do with a view depends on the type of view it is. Default views Initially, all views are grouped under the Default group. Default views cannot be saved, renamed, or deleted. They can be copied, saved in a different location or under a different filename, and exported. User views User views can be copied and pasted among groups, renamed, saved, or deleted. User views are created by either copying a default view and pasting it in the User group (or any group that you have created within the User group) or by saving a Default view to the User group. To open a view 1. Click (CTRL+R). 2. Click Views (under the Reporting and Trending bar). 3. Expand the list of either Default or User views. 4. From the list of views displayed on the left, click a view. 5. Click the link in the top right corner of the window, to connect to a Spotlight Statistics Repository. 6. Using the lists at the bottom of the window, select a connection and time frame. 7. Click Refresh.

8 Views 8 To save a view Tips: 1. Open the view. 2. Click Save As. Save the view to the User group or create a new group to save the view to. You may need to scroll to see all the charts and tables displayed in a view. You can change which you are connected to at any time by clicking the link in the top right corner of the window. To retain the connection details and time range of a view, click Save As (Default view) or Save (User view). Default views cannot be saved within the Default group. Views can be saved under the User group or any new groups you have created. For descriptions of the views available, see the online help. Group Views To create a new view group Tips: 1. Click (CTRL+R). 2. Click Views (under the Reporting and Trending bar). 3. From the list of views displayed on the left, right-click the User view group. 4. From the shortcut menu, select New view group. 5. Type a name for the view group. 6. Click OK. Right-click a group to access the available options. You can cut, copy, paste, delete, and rename groups. You can create nested groups by creating "groups within groups" under the User group. To move views between groups, use the copy or cut and paste options. Deleting groups deletes all views contained within that group.

9 Views 9 Export Views Grids are exported as.csv files. Charts are exported as.bmp files. You cannot change the name of the exported files so if you have previously exported a view and you don't want to overwrite the files, either rename the exported files (via Windows Explorer) or export the new files to a different location. To export a view 1. Open the view you want to export. 2. Click Export. 3. Select the folder where you want to save the exported files. 4. Click OK.

10 3 Reports About Reports In Spotlight, you can configure the Diagnostic Server to store the statistics it collects in a repository for reporting purposes. You can then retrieve this data in the form of views and reports. You can use the example reports provided or build on them to create your own reports by querying the data yourself using stored procedures. See "Stored Procedures" (page 33) for more information. Spotlight reports are displayed in an external viewer. When you launch the Spotlight Report Viewer it connects to the. To view a Spotlight report Tips: 1. Click (CTRL+R). 2. Click Reports (under the Reporting and Trending bar). 3. Select a report from the Reports list displayed on the right pane. 4. If SQL Server authentication is required, enter the user and password details. 5. In the Spotlight Report Viewer, select the following: A time frame for the report Diagnostic Server Connection you want to view the report for 6. Click View Report. Change reports by selecting a different report from the Report list at the top of the Spotlight Report Viewer. Complete the Diagnostic Server and Connection Name fields and then click View Report. If you change the report criteria (time frame, Diagnostic Server or Connection Name), click View Report to see an updated version of the report. Save reports in PDF, or as an.xls (Microsoft Excel) file using the Save icon on the Report toolbar. Print, search, and refresh reports using the Report toolbar. You can create your own reports for use in Spotlight.

11 Reports 11 Use Your Own Reports The Spotlight Report Viewer displays reports using SQL Server 2005 or 2008 Report Definition Language files (.rdl). You can build on the example reports or create your own reports by querying the data from the yourself. Example reports are located on the Spotlight client in the following folder: C:\Program Files\Quest Software\Spotlight on SQL Server\Plug-ins\SoSSE\Reports Example reports compatible with SQL Server 2008 Reporting Services are located in the following folder: C:\Program Files\Quest Software\Spotlight on SQL Server\Plug-ins\SoSSE\Reports\Web Reports 2008 To use your own report definition files» Set the location of your.rdl files in Spotlight. or a. Click Reporting and Trending in the Spotlight browser. b. Right-click Reports. c. From the shortcut menu, select Settings. d. Navigate to the directory containing the rdl files you want to use. e. Click OK.» Copy your.rdl files into C:\Program Files\Quest Software\Spotlight on SQL Server\Plug-ins\SoSSE\Reports. Note: If you do not have write access to the Program Files folders listed above, you can create the following folder and use it as an alternative: Windows XP - My Documents\Spotlight Reports Vista - <user>\documents\spotlight Reports Example Reports Spotlight Reporting and Trending includes a set of example reports that demonstrate how stored procedures are used to query the. Example reports are located on the Spotlight client in the following folder: C:\Program Files\Quest Software\Spotlight on SQL Server\Plug-ins\SoSSE\Reports

12 Reports 12 Example reports compatible with SQL Server 2008 Reporting Services are located in the following folder: C:\Program Files\Quest Software\Spotlight on SQL Server\Plug-ins\SoSSE\Reports\Web Reports 2008 The input parameters are the same for all example reports. The Input parameters are: Time start Time end Diagnostic Server Server name (monitored object) Database Growth The Database growth report contains the following charts: Data Growth Trend Shows the following statistics for each database on the specified server: Data size Data allocated Log Growth Trend Shows the following statistics for the transaction log for each database on the specified server: Log size Log allocated Server Alarms The Server Alarms report shows a list of all alarms that have occurred on the specified server for a set time period. SQL Server Health The Server Health report shows the following charts for the specified server for a set time period: Number of Logged on Users Batches, Compiles and Recompiles Buffer Cache Hit Ratio Procedure Cache Hit Ratio Page Life Expectancy Server Memory

13 Reports 13 Windows Server Health The Windows Server Health report shows the following charts for the specified server for a set time period: Total CPU Utilization chart Context Switches/Sec Paging Rate Network Bandwidth Utilization Disk Queue Length See "Deploy Reports" in this section for information on how to deploy the example reports. Deploy Reports This topic outlines how to use the SQL Server 2005 and 2008 Reporting Services web interface to deploy the example reports included with Spotlight Reporting and Trending. See "Example Reports" (page 11) for more information. To complete the following steps, you need report server administrator privileges. 1. In the Reporting Services web interface, create a shared data source by clicking New Data Source. 2. Complete the fields on the New Data Source page. You will need the following information: Data Source name The example reports included with Spotlight Reporting and Trending use a shared data source called. Connection type Select Microsoft SQL Server. Connection string Enter the connection string for the SQL Server instance that you installed the on. For example: Data Source=YourServerHostname;Initial Catalog=Spotlight Statistics Repository Connection credentials Including user name and password, if required.

14 Reports Upload the report definition (.rdl) files. To do this: a. In the Reporting Services web interface, click Upload File. b. Click Browse to navigate to the.rdl file. Spotlight report definition files are located on the Spotlight client in the following folder: C:\Program Files\Quest Software\Spotlight on SQL Server\Plug-ins\SoSSE\Reports Report definition files compatible with SQL Server 2008 Reporting Services are located in: C:\Program Files\Quest Software\Spotlight on SQL Server\Plugins\SoSSE\Reports\Web Reports 2008 c. Click OK. You can view web-based reports within Spotlight by configuring the Web Reports portal to point to your web-based reports server. Specify the Location of Your Web Reports Server You can view web-based reports within Spotlight by configuring the Web Reports portal to point to your web-based reports server. To specify the location of your Web reports server 1. Select View Options. 2. Click Diagnostic Server. 3. Click Specify the location of your Web Reports server. 4. Enter the URL of your web reports server. View Web Reports in Spotlight You can view web-based reports within Spotlight by configuring the Web Reports portal to point to your web-based reports server. Review the following for information on how to use the SQL Server Reporting Services web interface to deploy the example reports included with Spotlight Reporting and Trending: Deploy Reports (page 13) Example Reports (page 11)

15 Reports 15 To view Web reports in Spotlight 1. In the Spotlight browser, click the Reporting and Trending bar. 2. Click Web Reports.

16 4 Schema of the The is based on a star schema because of the simplicity it offers for data storage and retrieval. Star schemas are a simple schema consisting of a centralized fact table connected to multiple dimension tables. Fact tables hold factual data and the dimension tables hold descriptive data. One of the benefits of a star schema is the speed of data retrieval. With star schemas, you can use relatively simple SQL queries to return particular information about, in this case, a SQL Server instance. Later in this chapter we ll show you how you can query the using Spotlight API stored procedures and T-SQL. In the, the fact table (spotlight_perfdata) stores facts related to SQL Server performance, and the dimension tables hold definition data such as SQL Server instance information and attributes.

17 17 The schema of the is as follows: The main dimension tables in the are as follows: Spotlight_datasources Contains data used to categorize the collections that are performed by Spotlight. There is one row in this table for each category of data collected. Examples of categories are SQL Server, Windows, and Diagnostic Server. Spotlight_domains Contains data about the Diagnostic Servers feeding the. In the case of multiple Diagnostic Servers feeding the, there is one row per Diagnostic Server. The Domain_description column contains the Diagnostic Server host name. Spotlight_monitored_objects Contains data about each of the SQL Server or Windows servers that a Diagnostic Server is monitoring. There is one row in this table for each server being monitored by each Diagnostic Server. Spotlight_techonologies Contains data that categorizes the collections that are performed. There is one row in this table for each category of data collected.

18 18 The remaining dimension tables contain data about the collections being stored in the Spotlight Statistics Repository. See "Table Definitions" (page 29) for more information on each of the tables in the. How the works The Diagnostic Server collects data about the SQL Server instance and then writes it to the. How often this data is written to the Spotlight Statistics Repository depends on the scheduling options set for each data collection in Spotlight. The data is inserted into the relevant dimension tables and then the fact is written into the fact table (spotlight_perfdata). Let s use the example of a Diagnostic Server writing data about the sqlblockinglist collection to the. In the following chart, the Diagnostic Server, DS123, is collecting data about SQL Server instance, SQLServer789, running on Windows host, Windows01.

19 19 The following chart shows how the data flows into the.

20 20 Store Data in the There are a number of options that must be configured for Spotlight to store data in the Spotlight Statistics Repository. To configure the 1. In Spotlight, select View Options. 2. Click Diagnostic Server. 3. Click Configure the. 4. Select Enable use of the. 5. Complete the remaining options.

21 21 See Configure the in the Spotlight on SQL Server Deployment Guide for more information. To store data in the 1. Click (Ctrl+M). 2. Click Schedules. 3. Select Store Reporting Data. 4. From here, you can configure how often data is stored in the Spotlight Statistics Repository for each collection. See Schedule Data Collection in the online help for more information. Query the Before querying the we need to know how the repository is structured and what data it contains. You can use the Spotlight API stored procedures to browse the dimension tables in the repository. You can then use additional stored procedures or custom T-SQL to retrieve data. Scenario Your manager has asked you to provide statistics on database performance for the last quarter. You decide to query the to return file I/O statistics for the databases your manager is interested in. To retrieve this data, you first browse the dimension tables to become familiar with the data in the. Once familiar, you write queries using the Spotlight stored procedures to get data from the fact table. You then refine the queries using custom T- SQL to retrieve more specific data from the. First up, we re going to use some of the Spotlight API stored procedures to browse the dimension tables and get the range of values available for use in our queries. The Spotlight stored procedures used below are explained in detail in this guide. Review the following for additional information: Stored Procedures (page 33) Specifications (page 36) In SQL Server Management Studio, create a new query against the Spotlight Statistics Repository. We want to know the names of the available data sources and the names and domains of the monitored objects. Let s start exploring the by looking at datasources. Datasources are broad categories of data.

22 22 To get a list of data sources in the repository, run the following query: exec spotlight_get_datasources This returns the following results: We can see here that we have data from Windows, SQL Server, the Diagnostic Server and alarms (availability). Now we want to know the name of the Diagnostic Server storing data in the Spotlight Statistics Repository. We can determine this by running: exec spotlight_get_domains To get the following result: We can see here that there is one Diagnostic Server writing data to this repository. There can be multiple Diagnostic Servers, each one resulting in a row. Finally, we want to know information about the servers being monitored. We can do this by running the following query: exec spotlight_get_monitored_objects To get the following results: A common question is what is the range of timestamps available for a monitored server? The following query returns the time ranges for all monitored servers: select so.monitored_object_name, min(st.timecollected) as 'start',

23 23 max(st.timecollected) as 'end' from group by spotlight_timestamps st join spotlight_monitored_objects so on st.monitored_ object_id = so.monitored_object_id so.monitored_object_name Now that we have some information about the Diagnostic Server, monitored servers and categories of collections in the, we want to know what statistics are available to us. The spotlight_get_tables stored procedure returns a list of tables in the repository for a given data source. Since we are looking for SQL Server data, from our first query above, we know that the datasource is sqlserver_spotlight, so we will use that as the parameter in the following query: exec spotlight_get_tables 'sqlserver_spotlight' The following results are returned. We re going to look at virtual file stats as this will give us the file IO statistics we need for our report. To find out what columns are available to us in the virtual file stats table use the following query. With this query we need to specify the datasource and table name. exec spotlight_get_table_columns 'sqlserver_ spotlight','virtualfilestats'

24 24 We need to produce the report over a time range, but for curiosity s sake, we would like to find out the time range of all stored data for a particular domain name, monitored object, and table combination. To do this, use the spotlight_get_table_span stored procedure and specify the domain name, monitored object, and table name. exec spotlight_get_table_span 'DS123:3843','Windows01_ SQLServer789_sqlserver','virtualfilestats' Now we can put it all together and query the fact table. We ll use the spotlight_get_table_range stored procedure and specify the following: start date end date domain name monitored object table name Spotlight_get_table_range returns data for the requested time range for a table, for a particular monitored object and domain. This gives us the following query: exec spotlight_get_table_range ' :15:16.153',' :15:17.113', 'DS123:3843','Windows01_SQLServer789_ sqlserver', 'virtualfilestats' Which returns all columns for the table virtualfilestats : Although there is a lot of useful data in the query results, it s not exactly what we are looking for. We need to refine the data returned from the. Following on from the example above, if we want to retrieve only certain columns from the virtualfilestats table, we can use custom T-SQL to return specific data.

25 25 To return only the iorate, disk, readsrate, and writesrate columns from the virtualfilestats table, we can use the following custom T-SQL: select from where group by order by sp.timecollected, max(case when sn.statistic_name = 'dbname' then sp.raw_value end) as 'dbname', max(case when sn.statistic_name = 'dbfilename' then sp.raw_value end) as 'dbfilename', max(case when sn.statistic_name = 'disk' then sp.raw_ value end) as 'disk', max(case when sn.statistic_name = 'iorate' then sp.raw_value end) as 'iorate', max(case when sn.statistic_name = 'readsrate' then sp.raw_value end) as 'readsrate', max(case when sn.statistic_name = 'writesrate' then sp.raw_value end) as 'writesrate' spotlight_perfdata sp join spotlight_stat_classes sc on sp.statistic_class_ id = sc.statistic_class_id join spotlight_stat_names sn on sp.statistic_name_id = sn.statistic_name_id join spotlight_monitored_objects so on sp.monitored_ object_id = so.monitored_object_id sc.statistic_class_name = 'virtualfilestats' and so.monitored_object_name = 'Windows01_ SQLServer789_sqlserver' and sp.timecollected between ' ' and ' ' sp.timecollected, sp.statistic_key_id sp.timecollected

26 26 This gives us the following results: Query Custom Counters Scenario You have created a custom counter in Spotlight on SQL Server and now want to create a report that contains data from that counter. Querying custom counters is exactly the same as querying other data in the Spotlight Statistics Repository. SQL Server custom counters are stored in the statistic class sqlcustomcounters and Windows custom counters are stored in the class windowscustomcounters. You can use the API stored procedures to retrieve date ranges and column names with these like any other data collection. The following example of T-SQL retrieves SQL Server custom counter values: select sp.timecollected, max(case when sn.statistic_name = 'countername' then sp.raw_value end) as 'countername', max(case when sn.statistic_name = 'countervalue' then sp.raw_ value end) as 'countervalue' from spotlight_perfdata sp join spotlight_stat_classes sc on sp.statistic_class_id = sc.statistic_class_id join spotlight_stat_names sn on sp.statistic_name_id = sn.statistic_name_id

27 27 join spotlight_monitored_objects so on sp.monitored_object_id = so.monitored_object_id where sc.statistic_class_name = 'sqlcustomcounters' and so.monitored_object_name = 'Windows01_SQLServer789_sqlserver' and sp.timecollected between ' ' and ' ' group by sp.timecollected, sp.statistic_key_id order by sp.timecollected The following shows the results returned for the custom counter Pages Allocated. Query Alarm Data Alarm data can be queried from the. Querying alarm data is the same as querying other data stored in the repository. Alarm data is stored in a statistic class called alarms. The following is an example of a T-SQL statement that can be used to retrieve alarm data. You need to specify the monitored object name and start and end times as parameters. select sp.timecollected, max(case when sn.statistic_name = 'severity' then sp.raw_value end) as 'severity', max(case when sn.statistic_name = 'text' then sp.raw_value end) as 'text', max(case when sn.statistic_name = 'rule' then sp.raw_value end) as 'rule', max(case when sn.statistic_name = 'action' then sp.raw_value end) as 'action',

28 28 max(case when sn.statistic_name = 'key' then sp.raw_value end) as 'key', max(case when sn.statistic_name = 'raw text' then sp.raw_value end) as 'raw text' from spotlight_perfdata sp join spotlight_stat_classes sc on sp.statistic_class_id = sc.statistic_class_id join spotlight_stat_names sn on sp.statistic_name_id = sn.statistic_name_id join spotlight_monitored_objects so on sp.monitored_object_id = so.monitored_object_id where sc.statistic_class_name = 'alarms' and so.monitored_object_name = 'Windows01_SQLServer789_sqlserver' and sp.timecollected between ' ' and ' ' group by sp.timecollected, sp.statistic_key_id order by sp.timecollected The above T-SQL statement returns the following results:

29 29 Table Definitions spotlight_datasources spotlight_datasources contains data used to categorize the collections that are performed. There is one row in this table for each category of data collected. Column datasource_id datasource_name Description Generated ID. Name of the data source. spotlight_domains spotlight_domains contains data about Diagnostic Servers. If you have multiple Diagnostic Servers feeding the, there is one row per Diagnostic Server in this table. Column Description domain_id Generated ID. domain_ description The name of the domain (analogous to Diagnostic Server host name means that this can hold data from multiple diagnostic servers). spotlight_monitored_objects spotlight_monitored_objects contains data about the SQL Servers and Windows hosts that a Diagnostic Server is monitoring. There is one row in this table for each SQL Server or Windows host being monitored by each Diagnostic Server. Column monitored_ object_id monitored_ object_ name monitored_ object_ display_ name Description Generated ID. Name of monitored object (Windows host name or SQL Server Instance Name etc). A user-friendly interpretation of the encoded name that the Diagnostic Server uses to uniquely identify a monitored object. This name is suitable for display on reports or other user-visible outputs.

30 30 Column Description parent_ monitored_ object_id technology_ id domain_id The monitored_object_id value of another monitored object upon which this one depends, or null if no such dependency exists. For example, a SQL Server database s parent will be the Windows host upon which it runs. The technology of this monitored object. The domain (that is, Diagnostic Server) this monitored object belongs to. spotlight_perfdata spotlight_perdata is the fact table into which all of the data is stored. Each fact is effectively matrixed by the hierarchy of attributes that denote which Diagnostic Server, monitored object, collection and collection column the datum originally came from. Column perfdata_id monitored_ object_id statistic_ class_id statistic_ name_id statistic_ key_id timecollected raw_value num_ observations sum_of_ values max_value min_value sum_of_ squares Description Generated ID. The monitored object that this fact came from. The collection that this fact came from. The column in the collection that this fact came from. The key column of the collection that this fact came from. The time this fact was collected from the monitored object. The raw value returned by the collection for that column. Internal use only. Internal use only. Internal use only. Internal use only. Internal use only.

31 31 spotlight_stat_classes spotlight_stat_classes contains the names of collections being stored in the Spotlight Statistics Repository. There is one row for each table/data source combination. Column statistic_class_id datasource_id statistic_class_name Description Generated ID. The data source that this collection belongs to. The name of the collection. spotlight_stat_keys spotlight_stat_keys contains key values for all of the collections stored in the Spotlight Statistics Repository. Each key value/table/data source combination has a row in this table. The purpose of this table is to allow queries to get data from the fact table based on a key value without having to scan values in the fact table itself. Column statistic_ key_id statistic_ class_id datasource_ id statistic_ name_id statistic_ key_name Description Generated ID. The statistic_class (Collection) this column belongs to. The data source of the collection that this column belongs to. Generated ID. The name of the key column. spotlight_stat_names spotlight_stat_names contains data about the columns belonging to collections. There is one row in this column for each column/table/data source combination. Column statistic_ name_id Description Generated ID.

32 32 Column statistic_ class_id datasource_ id statistic_ name datatype_ guide derived Description The statistic_class (collection) this column belongs to. The data source of the collection that this column belongs to. The name of the column. This column provides a broad guide to the type of data that can be expected in facts for this named statistic. Values can be string, numeric or date. Whether this is a derived column or not. spotlight_technologies spotlight_technologies contains data used to categorize the collections that are performed. There is one row in this table for each category of data collected. Column technology_id technology_name Description Generated ID. The name of the technology. spotlight_timestamps spotlight_timestamps contains a summary of times for which data was collected, qualified by monitored entity. It allows Spotlight to quickly determine the times for which historical data is available for individual monitored entities, without the overhead of scanning the potentially large fact table spotlight_perfdata. There is one row in this table for each category of data collected. Column timestamp_id timecollected domain_id monitored_ object_id statistic_ class_id Description Generated ID. The time this fact was collected from the monitored object. The domain that this fact belongs to. Generated ID. The collection that this fact came from.

33 33 Stored Procedures The following stored procedures return either metadata about the or data from within the. Stored procedure spotlight_get_datasources @table_name spotlight_get_technologies Description Returns a list of all data sources in the repository (effectively the contents of spotlight_datasources). Returns a list of all the diagnostic servers putting data into this repository. Returns a list of names of all the monitored objects that have data in this repository for a given domain_name. If domain_name is NULL all monitored objects are returned. Also returns the monitored object display name, domain description, and technology name. Returns a list of columns for a table belonging to a data source (effectively the contents of spotlight_stat_names). Returns the rows and columns from a table with a timestamp that most closely time. Returns a list of keys for a table belonging to a data source (effectively the contents of spotlight_stat_keys). Returns the rows and columns from a table Returns the time span of all stored data for a particular domain_name, monitored object, and table combination. Returns a list of tables in the repository for a given data source (effectively the contents of spotlight_stat_classes). Returns a list of technologies in the repository (effectively the contents of spotlight_ technologies).

34 34 Stored procedure @column_name @monitoredobject @maxrows spotlight_ssr_version Description Returns data for the requested time range for the Workload view. Returns data for the requested time range for a table and associated columns for multiple monitored objects in a domain. Returns data for the Database File IO Statistics view. Returns the same data as spotlight_rt_get_ batch_data but for a particular point in time. Returns the top n rows of spotlight_rt_get_ point_data based on the column sort order passed in. Returns data for the view Wait Statistics view. Returns the worst performing SQL recorded in the over the supplied time span for a particular SQL Server instance. Returns the version number of the Spotlight Statistics Repository. The following stored procedures are used for the example reports: Stored procedure spotlight_rep_alarms Description Returns a list of alarms for a chosen monitored object from the repository. Used in the Server Alarms report. This is a further example of how to query the Spotlight Statistics Repository to return tabular information.

35 35 Stored procedure Description spotlight_rep_cache_hit_ratios spotlight_rep_cpu_utilization spotlight_rep_disk_growth spotlight_rep_disk_statistics spotlight_rep_growth_data_ files spotlight_rep_sql_batches_rate Returns hit rates for the buffer and procedure caches for all monitored objects in the repository. Used in the SQL Server Health report. This is an example of querying two columns from a table over time from the. Returns the amount of CPU being used by the specified server against the total amount being used by all Windows processes. Used in the Windows Server Health report. This is an example of querying one column from a table over time from the. Returns information on disk space usage over time. Returns disk statistics (percentage busy and disk queue length) for the specified server. Used in the Windows Server Health report. This is an example of querying multiple columns with multiple rows from a table over time from the Spotlight Statistics Repository. Returns the following statistics for the specified server: Data size Data allocated Index size Table size Log size Log allocated Used in the Database Growth report. This is an example of querying multiple columns with multiple rows from a table over time from the repository. Returns the rate at which batches of SQL statements are being submitted to SQL Server for execution.

36 36 Example stored procedures The following stored procedures can be used as examples of how to retrieve data from the. spotlight_rt_get_allsqlagentjobs spotlight_rt_get_custom_counter spotlight_rt_get_failedsqlagentjobs spotlight_rt_get_mostcommonblock spotlight_rt_get_sqlblockedmost spotlight_rt_get_sqlblockingmost Specifications The following section contains specifications for each of the stored procedures. spotlight_get_datasources Returns a list of all data sources in the repository (effectively the contents of spotlight_ datasources). Syntax spotlight_get_datasources Input parameters There are no input parameters for this stored procedure. Returned values ColumnName ColumnType Description datasource_id int ID of data source datasource_name nvarchar(255) Name of data source

37 37 spotlight_get_domains Returns a list of all the diagnostic servers putting data into this repository. Syntax spotlight_get_domains Input parameters There are no input parameters for this stored procedure. Returned values ColumnName ColumnType Description domain_id int ID of domain domain_description nvarchar(255) Descriptive name of domain spotlight_get_monitored_objects Returns a list of names of all the monitored objects that have data in this repository for a given domain_name. If domain_name is NULL all monitored objects are returned.

38 38 Returned values ColumnName ColumnType Description monitored_object_name nvarchar(255) Name of monitored object monitored_object_ display_name monitored_object_ display_technology_ name nvarchar(255) nvarchar(255) A user-friendly interpretation of the encoded name that the Diagnostic Server uses to uniquely identify a monitored object. This name is suitable for display on reports or other user-visible outputs. Combines the columns monitored_object_ display_name and technology_name to display a user-friendly name that includes the monitored object name and the technology. domain_description nvarchar(255) Descriptive name of domain technology_name nvarchar(255) Descriptive name of technology spotlight_get_table_columns Returns a list of columns for a table belonging to a data source (effectively the contents of spotlight_stat_names). Returned values ColumnName ColumnType Description statistic_name nvarchar(255) Name of table column

39 39 spotlight_get_table_data_instance Returns the rows and columns from a table with a timestamp that most closely Returned values The returned values are dependent on the columns implicit in the table represented by parameter. spotlight_get_table_keys Returns a list of keys for a table belonging to a data source (effectively the contents of spotlight_ stat_keys). Returned values ColumnName ColumnType Description statistic_key_value nvarchar(255) Key value for procedure

40 40 spotlight_get_table_range Returns the rows and columns from a table @table_name @table_name Returned values The returned values are dependent on the columns implicit in the table represented by parameter. spotlight_get_table_span Returns the time span of all stored data for a particular domain_name, monitored object, and table combination. @table_name @table_name Returned values The returned values are dependent on the columns implicit in the table represented by parameter.

41 41 spotlight_get_tables Returns a list of tables in the repository for a given data source (effectively the contents of spotlight_stat_classes). Syntax Input Returned values ColumnName ColumnType Description statistic_class_name nvarchar(255) Name of table spotlight_get_technologies Returns a list of technologies in the repository (effectively the contents of spotlight_ technologies). Syntax spotlight_get_technologies Input parameters There are no input parameters for this stored procedure. Returned values ColumnName ColumnType Description technology_id int ID of technology technology_name nvarchar(255) Descriptive name of technology

42 42 spotlight_sqlanalysis_results Returns the worst performing SQL recorded in the over the supplied time span for a particular SQL Server instance. @maxrows @maxrows Returned values Returns information about the worst performing SQL recorded in the Spotlight Statistics Repository. spotlight_ssr_version Returns the version number of the. Syntax spotlight_ssr_version Input parameters There are no input parameters for this stored procedure. Returned values Returns the version number.

43 43 Maintenance Plan for Spotlight Statistics Repository and Playback Database It is recommended that a maintenance plan for both the and the playback database is implemented. Both the and playback database support a large number of monitored servers and automatically maintain the age of data available. Due to the nature of the automatic purging of old data, the data may become fragmented and this may decrease the efficiency of data retrieval operations over time. This topic provides guidance on how to configure and automate the maintenance of the and the playback database. Database Configuration The and the playback database do not require the database to be run under the Full Recovery model. However, since Spotlight uses the model system database to create the repository, your Spotlight repository may inherit the Full Recovery setting. Unless you are required to use the Full Recovery model and are prepared to perform regularly scheduled log file backups, we recommend you change the repository recovery model to Simple (in SQL Server 2000, this option is called Truncate Log on Checkpoint). Doing so will help maintain a considerably smaller transaction log file. To change the repository database to Simple recovery, run the following command: SQL Server 2000 EXEC [dbo].[sp_dboption] 'SpotlightStatisticsRepository','trunc. log on chkpt.','true' SQL Server 2005 and 2008 ALTER DATABASE [SpotlightStatisticsRepository] SET RECOVERY SIMPLE; Fragmentation and Index Performance The implements a star schema because of the data warehousingstyle simplicity it offers for data storage and retrieval. The star schema implements a primary fact table that references a number of dimension tables. The fact table holds collections and alarm data, and the dimension tables hold definition information on everything from instance information to collection names. Spotlight maintains the age of the data in the repository, but the continuous INSERT and DELETE operations performed for data insertion and purging can lead to fragmentation. Fragmentation will eventually lead to increased CPU and I/O resource consumption.

44 44 Although the playback database uses a simpler schema, its data will also become fragmented over time as new data is inserted and old data purged. To best address performance concerns, fragmentation should be minimized and index statistics should be kept up-to-date. Spotlight's scheduled maintenance feature does this automatically. By default, the Diagnostic Server runs maintenance procedures daily at 3am for the playback database and. The maintenance schedule can be changed in Spotlight Options. See the online help for more information. Database Backup The and playback database should be included in the list of important databases which have a disaster recovery plan associated with them. The implementation of this task is dependent on the policies and infrastructure of your organization.

45 Appendix: Contact Quest Contact Quest Support Quest Support is available to customers who have a trial version of a Quest product or who have purchased a Quest product and have a valid maintenance contract. Quest Support provides unlimited 24x7 access to SupportLink, our self-service portal. Visit SupportLink at From SupportLink, you can do the following: Retrieve thousands of solutions from our online Knowledgebase Download the latest releases and service packs Create, update and review Support cases View the Global Support Guide for a detailed explanation of support programs, online services, contact information, policies and procedures. The guide is available at: Spotlight Support When contacting Quest Support about a Spotlight product, please have the version and build number handy. You can obtain both by choosing the About option from the Help menu in the relevant Spotlight application. If your question is about an error message, write the message down and have it available for the technical support representative. Whenever you have a question about Spotlight, click Support Bundle or Contact Support on the Help menu. This creates a file called SpotlightSupport.zip. This file contains a snapshot of your Spotlight installation. Send this file and your request for assistance to Quest Software ([email protected]). Note: You can find SpotlightSupport.zip in different locations under different operating systems. Windows XP: C:\Documents and Settings\<user>\Application Data\Quest Software\Spotlight\Default\Support Vista: C:\Users\<user>\AppData\Roaming\Quest Software\Spotlight\Default\Support If you find that information in the Help system or printed user guides is incorrect, unclear, or incomplete, please report the problem to Quest Support.

46 Appendix: Contact Quest 46 Contact Quest Software Mail Web site Quest Software, Inc. World Headquarters 5 Polaris Way Aliso Viejo, CA USA See our web site for regional and international office information. About Quest Software Now more than ever, organizations need to work smart and improve efficiency. Quest Software creates and supports smart systems management products helping our customers solve everyday IT challenges faster and easier. Visit more information.

47 Index A about Quest Software 46 alarm data 27 C Contact Quest Software 46 custom counters 26 Customer support 45 D definition files 11 E example reports 11 export views 9 M maintenance plan 43 Q query alarm data 27 custom counters Quest Software 46 Quest support 45 R reporting 6 reports about 10 create 11 customize 11 Database Growth 11 definition files 11 deploy 13 examples 11 open 10 Server Alarms 11 SQLServer Health 11 web 14 Windows Server Health 11 S schema 16 Spotlight Reporting 6 alarm data 27 custom counters 26 how it works 18 maintenance plan 43 query 21 schema 16 specifications 36 store data 20 stored procedures 33 table definitions 29 store data 18, 20

48 Index stored procedures 33 Support 45 T table definitions 29 Technical support 45 trending 6 V views copy 7 default 7 display 7 export 9 group 8 move 8 open 7 save 7 user 7 W web reports 14 server location 14