Safeguarding Your SAP System Availability And Performance

Safeguarding your SAP System
Availability and Performance

Claricent, Inc

Agenda

 The challenges
– Managing system performance and stability

 The solutions:
– Canary Code – system performance monitoring
– Fitness Check

2
2 © 2008

System growth and complexity is
ongoing

 More and more
customers run mission
critical applications on
SAP.
 “67% of mission-
critical applications
have more than 9
hours of downtime per
month.”
– Source : SAP-The Standish Group.

3 © 2008

Business impact of an outage is
substantial

 Comair lost $20
million of revenue
after a crew
scheduling application
failed during the
December holiday
travel season.
– IT Risk, Turning Business Threats into Competitive
Advantage, Harvard Business School Press.

4 © 2008

How widespread is the perceived issue?

 58% of line-of-
business managers
are unsatisfied with
performance of their
enterprise applications
 Challenged to identify
issues before end-
users were impacted.
– Monitoring the End-User Experience: Improving Business
Performance through Application Management,
AberdeenGroup, 2009

5 © 2008

Historically, what is monitored?

6 © 2008

Shouldn’t the focus be on the business
application?

7 © 2008

The solution – CanaryCode

 Proactive and near
real time monitoring
and alerting to:
– Safeguard Availability
– Safeguard
Performance; before it
impacts the business.
 Focus on meaningful /
business related
indicators

8 © 2008

Architecture

Systems to monitor The KPI Groups

R/3
Detailled analysis
Canary Code
BW
Central Instance
BWA
(ABAP based
solution installed on
an existing server
APO e.g. BW or SolMan)
Alerting

Logged technical details
for root cause analysis
CRM

© 2008

Transaction /DVD/Mon

 Meaningful
KPIs
 Out of the box
and user
defined KPIs
 KPIs can be
grouped

10 © 2008

Transaction /DVD/Mon

11 © 2008

Example of indicators
(not all are available, nor on all db/versions)

System activity Objective
 Work process activity (busy, free, running programs)  Detect locks
 CPU: load, IO wait, user, system  Be informed if system is overloaded
 Memory, Swap  Be able to know if we increase workload,
 File system free space parallelization.
 Number of users, number of transactions, average runtime

Database
 CPU load on DB server  Detect bad sql statement / execution plan
 Memory on DB server  Detect issues on DB
 Archive log on / off
 Row scan

Exceptions
 Dumps, Syslog, Lock  Alert of system wide issue or specific issue
 Benchmarking (a specif batch, process)  Allow Root cause analysis
 Updates
 RFC queues

BW-BWA
 Trex Memory, CPU, unloads, threads, inactive index  Alert if reporting is impacted
 Number of packets in an infopackage  Alert if bw load are impacted

12 © 2008

Flexibility is paramount – Easy as 1-2-3

 Quickly and
easily create
your own KPI’s

13 © 2008

Strengths of CanaryCode

 Global view of all your systems
– COCKPIT with all KPI, real time trending
 Sophisticated / Business related INDICATORS
 FEW alerts and near zero false alert
 ROOT CAUSE analysis
– logs (WP, memory, trace of exceptions, of alerts)
– to prevent reoccurrence
 UP and running in one day
 Lean and LIGHT like a canary
 Customer indicators can be created in few minutes

© 2008

Keeping Your System Fit:
BW Fitness Check

BW Fitness Check – Tool-based analysis of
system performance 1

5 2
Data volume

Recommendations
Data quality
(Workshop)

Performance
System
robustness
3

4

16 © 2008

David A. Fox
Managing Principal
Consultant

T: 888-325-6496 x511
E: dfox@claricent.com

20 www.claricent.com  twitter/claricent  facebook/claricent
© 2008

Safeguarding Your SAP System Availability And Performance

More Related Content

What's hot (19)

Similar to Safeguarding Your SAP System Availability And Performance (20)

Safeguarding Your SAP System Availability And Performance