monitoring ssas cube performance and usage

26
Monitoring SSAS cube performance and usage Amit Bansal CTO | Peopleware India (unit of eDominer Systems) www.amitbansal.net | www.WeTogether.in | www.peoplewareindia.com

Upload: vera

Post on 24-Feb-2016

58 views

Category:

Documents


0 download

DESCRIPTION

Monitoring SSAS cube performance and usage. Amit Bansal CTO | Peopleware India (unit of eDominer Systems) www.amitbansal.net | www.WeTogether.in | www.peoplewareindia.com. Amit Bansal . CTO, eDominer Systems & Peopleware India Corporate Trainer/Consultant & Evangelist - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Monitoring SSAS cube performance and usage

Monitoring SSAS cube performance and usage

Amit BansalCTO | Peopleware India (unit of eDominer Systems)www.amitbansal.net | www.WeTogether.in | www.peoplewareindia.com

Page 2: Monitoring SSAS cube performance and usage

<about> Amit Bansal </about>

CTO, eDominer Systems & Peopleware IndiaCorporate Trainer/Consultant & EvangelistConducted more than 200 workshops on SQL Server & BI for top notch IT companies world wideMicrosoft MVP for SQL Server, UG LeadMicrosoft Certified Trainer Advisory Council memberSpeaker at TechED India, TechED US & TechED EuropeTechnical Reviewer – MSL courses on SQL ServerSME – SQL Server 2008 certificationsManager – www.WeTogether.in, www.BlogBoard.inPresident – SQLServerGeeks.com

Page 3: Monitoring SSAS cube performance and usage

IntroductionWhen you create large, enterprise-scale databases, you may encounter performance problems caused by poor database design or hardware limitations.When you have performance issues, it is critical to first properly diagnose the problem before you take corrective measures.

Page 4: Monitoring SSAS cube performance and usage

Shared resourcesCPU

For processing & queryingDuring partition processing, CPU is used mainly to build aggregations

MemoryAlmost all processes within SSAS need memory

I/OSSAS generates I/O operations both directly and indirectly

Page 5: Monitoring SSAS cube performance and usage

Using Task Manager to monitor CPU usage

DEMO

Page 6: Monitoring SSAS cube performance and usage

ToolsTask ManagerWindows Reliability And Performance MonitorSQL Server ProfilerDynamic Management Views

Page 7: Monitoring SSAS cube performance and usage

Windows Reliability & Performance Monitor

Provides tools to analyze operating system, application, and hardware performance.Consists of three components:

Resource ViewReliability Monitorand Performance Monitor

Page 8: Monitoring SSAS cube performance and usage

Windows Reliability & Performance Monitor

What do you monitor?CPU

% Processor TimeMemory

Pages/sec is the number of pages read from and written to disk to resolve hard page faultsCommitted Bytes is the amount of virtual memory requested by all running processes

Page 9: Monitoring SSAS cube performance and usage

Windows Reliability & Performance Monitor

What do you monitor?I/O

Current Disk Queue Length represents the current number of queued requests to the disk. If this number is constantly above the number of spindles plus 2, the disk subsystem could be improved by adding spindles (to a RAID 0/10 configuration) or by using faster disksDisk Read Bytes/sec & Disk Write Bytes/sec

Page 10: Monitoring SSAS cube performance and usage

Windows Reliability & Performance Monitor

What do you monitor?SSAS counters:

MSAS 2008: Memory – Memory Limit High KBMSAS 2008: Memory – Memory Limit Low KBMSAS 2008: Memory – Memory Usage KBMSAS 2008: ProcessingMSAS 2008: Proc AggregationsMSAS 2008: Proc Indexes

Page 11: Monitoring SSAS cube performance and usage

Monitoring Processing PerformanceSSAS counters:

MSAS 2008: ProcessingTotal Rows ConvertedTotal Rows ReadTotal Rows Written

SQL Server ProfilerCommand Events

Command BeginCommand End

Progress ReportsProgress Report BeginProgress Report EndProgress Report Error

Page 12: Monitoring SSAS cube performance and usage

Monitoring using Performance Monitor

DEMO

Page 13: Monitoring SSAS cube performance and usage

Monitoring Processing PerformanceUsing Dynamic Management Views (DMVs)SELECT TOP 10 *FROM $System.DISCOVER_OBJECT_ACTIVITYORDER BY OBJECT_CPU_TIME_MS DESC

Page 14: Monitoring SSAS cube performance and usage

Monitoring using DMVs

DEMO

Page 15: Monitoring SSAS cube performance and usage

Monitoring Query PerformanceRegardless of the tool we use, trace data is the most important source of information on query related operations.Trace data provides information on the internal operations of the Storage Engine and the Formula Engine, for example showing if aggregations are used or not or if calculations are evaluated in bulk mode or not

Page 16: Monitoring SSAS cube performance and usage

Monitoring Query PerformanceProgress Report Begin (5) / Progress Report End (6)

Query (14) shows when the Storage Engine accesses a partition or aggregation to get data. This will only happen if the data required is not available in the Storage Engine cache.

Query Begin (9) / Query End (10) – are raised at the start and end of query evaluation.

MDXQuery (0) shows the MDX statement sent to Analysis Services. For the Query End event, the Duration column shows the overall amount of time taken to run the query and return its results back to the client.

Page 17: Monitoring SSAS cube performance and usage

Monitoring Query PerformanceCalculate Non Empty Begin (72) / Calculate Non Empty End (74) – have no related subclass events.The Calculate Non Empty events are raised when Analysis Services performs non empty filtering operations, for example when the NonEmpty MDX function or the Non Empty statement is used in a query, and these operations are often the cause of slow query performance.

Page 18: Monitoring SSAS cube performance and usage

Monitoring Query PerformanceGet Data From Aggregation (60) – has no related subclass events. This event is raised when the Storage Engine reads data from an aggregation.Get Data From Cache (61) – is raised when data is read from cache.

Page 19: Monitoring SSAS cube performance and usage

Monitoring Query PerformanceQuery Cube Begin (70) / Query Cube End (71)Query Dimension (81) - this event is raised when queries retrieve members from dimension hierarchiesQuery Subcube (11)Query Subcube Verbose (12) – is functionally identical to Query Subcube, but it adds more information about the multidimensional

Page 20: Monitoring SSAS cube performance and usage

Monitoring Query Performance using SQL Server Profiler

DEMO

Page 21: Monitoring SSAS cube performance and usage

Monitoring Query PerformanceCan you monitor query performance using Performance Monitor?

MSOLAP: CacheMSOLAP:MDX

Page 22: Monitoring SSAS cube performance and usage

Summary

Monitoring Processing PerformancePerformance MonitorSQL Server ProfilerDMVs

Monitoring Query PerformancePerformance MonitorSQL Server ProfilerDMVs

Page 23: Monitoring SSAS cube performance and usage

Feedback / QnAYour Feedback is Important!Please take a few moments to fill out our

online feedback form at: << Feedback URL – Ask your organizer for this in advance>>

For detailed feedback, use the form at http://www.connectwithlife.co.in/vtd/helpdesk.aspx

Or email us at [email protected]

Use the Question Manager on LiveMeeting to ask your questions now!

Page 24: Monitoring SSAS cube performance and usage

SQLServerGeeks.comBangalore – 20th November 2010Mumbai – 28th November 2010Gurgaon – Dec 5, 2010

Visit www.SQLServerGeeks.com to register

Page 25: Monitoring SSAS cube performance and usage

Join my networkJoin my network:Personal Site – http://www.amitbansal.netLinkedIn – http://www.linkedin.com/in/AmitBansal2010Forum – http://www.WeTogether.inBlog – http://www.BlogBoard.in/AmitBansalTwitter – http://www.twitter.com/A_BansalFaceBook – http://www.facebook.com/people/Amit-Bansal/525339346

Page 26: Monitoring SSAS cube performance and usage

© 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.

The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after

the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.