際際滷

際際滷Share a Scribd company logo
Implementation of Lean Methods on
Management of IT Infrastructure
Monitoring System
By :
Dio Pratama | Rizal F. Aji | Setiadi Yazid
ICACSIS 2018
Faculty of Computer Science
Universitas Indonesia
Brief Description
Currently IT manage so many Infrastructure devices :
Faculty of Computer Science
Universitas Indonesia
Problem Analysis
There is an incident alarm which
completeion duration takes hundreds
to thousands of hours
Too many alarm notifications,
difficult to distinguish which really
shows the problem
There are many redundant
alarms
There are many red nodes, but in
reality there are no disrupted
services
1
2
3
4
Faculty of Computer Science
Universitas Indonesia
Problem Analysis (Contd)
11367
1735
58
13050
281
2540
0
2000
4000
6000
8000
10000
12000
14000
IT Support ERP Internal Apps IT
Infrastructure
IT
Collaboration
IT Security
Total Incident Each Departement - 2017
Faculty of Computer Science
Universitas Indonesia
Problem
The monitoring system cannot provide
accurate information regarding the
condition of the IT infrastructure
Faculty of Computer Science
Universitas Indonesia
Root Cause Analysis
Faculty of Computer Science
Universitas Indonesia
06
01
02
03
04
05
 Lean method is used to manage
and improve business processes
efficiently (Pillai, Pundir, &
Ganapathy, 2014)
 IT infrastructure is one of
the key operational services
of IT (Peppard & Ward,
2016)
 Business is rely on IT, systems must
be monitored 24x7 (Hernantes,
Gallardo, & Serrano, 2015)
 The monitoring aspect becomes
more complicated, notifications
accumulate, which makes
difficult to track problems
(Turnbull, 2016)
Literature Review
 Need a method to manage
monitoring system properly
 IT become more complex, services and
tools that need to be monitored are
increasingly and varied (Turnbull,
2016)
 One of popular process
management methode is lean
(Vieira, 2015)
 The management philosophy is
to emphasize the elimination of
waste (Plenert, 2012)
Faculty of Computer Science
Universitas Indonesia
Literature Review (Lean Implementation)
Loader & Janssen (2015) in his book Lean IT
Foundation Increasing Value Of IT explains that the
principle of implementing lean are through five stages
as shown by the following figure:
Transportation
Inventory
Motion
Waiting
Over Production
Over Processing
Defect
The management philosophy is to emphasize the
elimination of waste (Plenert, 2012), and the waste
is categorized as TIMWOOD (Bell & Orzen, 2011).
Faculty of Computer Science
Universitas Indonesia
Literature Review (Lean Toos)
Steven C. Bell and Michael A. Orzan (2011) explained that lean offers a variety of tools to eliminate process waste,
including:
Value Stream Mapping (VSM)
SIPOC (Supplier/Input/Process/Output/Customer)Gemba Walk (Go and see analysis)
Voice of Customer (VoC)
Work Standarization
5S (Seiri/Seiton/Seisho/Seiketsu/Shitsuke)
Pull Logistic
Continues Improvement (Kaizen)
Faculty of Computer Science
Universitas Indonesia
Research Methodology
 Research Design: Qualitative
 Research Methode: Case Study & Action Research
 Data Collection : Interview, Observation, FGD
1
Problem Identification
3 5
2 4
Literature Review
Observation Phase
Design Phase
Implementation Phase
Faculty of Computer Science
Universitas Indonesia
Lean Implementation (Value Identification)
Value Identification  Voice Of Customer (VoC) :
No Identified Values Achievment
Yes No
1 Technicians able to quickly see and responds to every occured
problems
O
2 Management able to quickly find out which services are
interrupted
O
3 Executives can quickly learn which business processes are
disrupted
O
4 Able to provide accurate alarm information O
5 Able to monitor the use of resources to maximize the use of the
service O
6 There are rules in the use of each feature on the monitoring
system
O
7 There is a clear workflow in handling every alarm O
8 Self service monitoring for IT service owners O
9 Simple display that can be directly understood by ordinary users O
Faculty of Computer Science
Universitas Indonesia
Lean Implementation (Value Stream Mapping)
Faculty of Computer Science
Universitas Indonesia
1
3
2
Waste Identification (Information Collection)
 There are resources that
are not being used and are
still monitored by the
system
 There is a function of resources
that are not owned by the
device, but monitored by the
system
 There are resources that are not
enabled by the device, but
monitored by the system
Waste  Inventory
Faculty of Computer Science
Universitas Indonesia
Waste Identification (Information Processing)
Alerting Waste Process :
1. Event Trigerred  Over Processing
2. Alert Logged  Over Production
3. Predefined Rules  Defect
4. Review The Alert  Over Production
Source : (Nand & Chaganty, 2008)
Faculty of Computer Science
Universitas Indonesia
Lean Implementation (Create New Flow by Elemeninate Waste)
Remove Unecessary Thing
 Stop monitoring of devices that are not in use.
 Stop monitoring of resources that are not being used
by the device.
 Stop monitoring of the resource function that the
device does not have.
 Stop monitoring of resources that are not enabled
by the device.
 Removing unnecessary alarm configuration.
Arrange Placements and Mark / Label on All Things
 Organize categorization and mark devices
based on area, city, region, service name,
and service owner.
 Organize grouping and marking devices
based on service function.
 Self service access to monitoring system
based on ownership of device or service.
 Remove the unnecessary dashboard
view and add another more informative
view by creating a single dashboad
view(Next 際際滷)
Cleaning and Tidying Up Display
Create Working Standard
 Making standard operational procedure (SOP)
Make sure SOP is running
 A reward system is applied for operators who perform every
month, with the assessment category being compliance with
SOPs.
 It has been proposed that the activity in the SOP be the
engineer performance assessment criteria which listed in the
Individual Work Plan (RKI).
 It has been proposed that adherence to this SOP be the point
of disciplinary Final Assessment (PA) of the engineer.
Faculty of Computer Science
Universitas Indonesia
Create Pull System
Aiming for the system to be only produce information when it is needed by customers at the right time, right amount and
right destination.
 Set the sending of alarms
according to working hours on
certain devices or services
 Make maintenance
scheduling arrangements if
there is downtime scheduled
 Send an alarm to a person
who is responsible for a
service
 Determine the parent and child of
the devices so that, if there are
problems with the parent device,
the child device will not send an
alarm
Faculty of Computer Science
Universitas Indonesia
Make Perfection
Continues Improvement (Kaizen)
Make procedures for using the
monitoring system
Implement best practice management
monitoring system (next slide)
Faculty of Computer Science
Universitas Indonesia
Make Perfection (Device Inventory)
Faculty of Computer Science
Universitas Indonesia
Make Perfection (Business View)
Faculty of Computer Science
Universitas Indonesia
Implementation Result
 Quantitatively after implementation, the
number of devices monitored increases while
alarms and events decreases
 Qualitatively, remains expected value
has been achieved
No Identified Values Caused Waste Activty Achievement
Yes No
1 Technicians able to quickly
see and respond to
problems that occur
 Inventory
 Motion
 Waiting
 Defect
 Short
 Shine
 Create pull
system
O
2 Management able to
quickly find out which
services are interrupted
 Motion
 Waiting
 Set in order
 Perfection
(Create
business
view)
O
3 Executives able to quickly
learn which business
processes are disrupted
 Motion
 Waiting
 Set in order
 Perfection
(Create
business
view)
O
4 Able to provide accurate
alarm information
 Inventory
 Defect
 Over
Production
 Over
Processing
 Short
 Create pull
system
O
5 Able to monitor the use of
resources to maximize the
use of the service - - O
6 There are rules in the use of
each feature on the
monitoring system
 Not directly
impact by
waste
 Standarize
O
7 There is a clear workflow in
handling every alarm
 Not directly
impact by
waste
 Perfection
(Create SOP) O
8 Self service monitoring for
IT service owners
 Not directly
impact by
waste
 Set in order
O
9 Simple display that can be
directly understood by
ordinary users
 Motion
 Waiting
 Shine
 Make
Perfection
(Create
business
view)
O
Network
Elements
Nodes Interfaces Alerts Events
Amount
Before
Lean
7683 639 6613 111 348172
Amount
After Lean
10705 1308 6683 26 177343
Percentage
Ratio
39% 105% 1% -77% -49%
Faculty of Computer Science
Universitas Indonesia
Summary
 After the lean method is implemented,
the system can convey the expected value
of the customer
 Waste in the monitoring system can be
reduced so that there is an increase in
accuracy related to information on the
condition of the IT infrastructure
Lean method can be applied to manage monitoring system through the steps of applying lean principle as follows:
VoC (Voice of Customer)
SIPOC of Monitoring
System
Remove Monitoring
Waste
Send alert on the right
time, right amount, right
destination
Implement Kaizen
Faculty of Computer Science
Universitas Indonesia
Source: xxx
Thank You

More Related Content

Lean Implementation on IT Infrastructure Monitoring System

  • 1. Implementation of Lean Methods on Management of IT Infrastructure Monitoring System By : Dio Pratama | Rizal F. Aji | Setiadi Yazid ICACSIS 2018
  • 2. Faculty of Computer Science Universitas Indonesia Brief Description Currently IT manage so many Infrastructure devices :
  • 3. Faculty of Computer Science Universitas Indonesia Problem Analysis There is an incident alarm which completeion duration takes hundreds to thousands of hours Too many alarm notifications, difficult to distinguish which really shows the problem There are many redundant alarms There are many red nodes, but in reality there are no disrupted services 1 2 3 4
  • 4. Faculty of Computer Science Universitas Indonesia Problem Analysis (Contd) 11367 1735 58 13050 281 2540 0 2000 4000 6000 8000 10000 12000 14000 IT Support ERP Internal Apps IT Infrastructure IT Collaboration IT Security Total Incident Each Departement - 2017
  • 5. Faculty of Computer Science Universitas Indonesia Problem The monitoring system cannot provide accurate information regarding the condition of the IT infrastructure
  • 6. Faculty of Computer Science Universitas Indonesia Root Cause Analysis
  • 7. Faculty of Computer Science Universitas Indonesia 06 01 02 03 04 05 Lean method is used to manage and improve business processes efficiently (Pillai, Pundir, & Ganapathy, 2014) IT infrastructure is one of the key operational services of IT (Peppard & Ward, 2016) Business is rely on IT, systems must be monitored 24x7 (Hernantes, Gallardo, & Serrano, 2015) The monitoring aspect becomes more complicated, notifications accumulate, which makes difficult to track problems (Turnbull, 2016) Literature Review Need a method to manage monitoring system properly IT become more complex, services and tools that need to be monitored are increasingly and varied (Turnbull, 2016) One of popular process management methode is lean (Vieira, 2015) The management philosophy is to emphasize the elimination of waste (Plenert, 2012)
  • 8. Faculty of Computer Science Universitas Indonesia Literature Review (Lean Implementation) Loader & Janssen (2015) in his book Lean IT Foundation Increasing Value Of IT explains that the principle of implementing lean are through five stages as shown by the following figure: Transportation Inventory Motion Waiting Over Production Over Processing Defect The management philosophy is to emphasize the elimination of waste (Plenert, 2012), and the waste is categorized as TIMWOOD (Bell & Orzen, 2011).
  • 9. Faculty of Computer Science Universitas Indonesia Literature Review (Lean Toos) Steven C. Bell and Michael A. Orzan (2011) explained that lean offers a variety of tools to eliminate process waste, including: Value Stream Mapping (VSM) SIPOC (Supplier/Input/Process/Output/Customer)Gemba Walk (Go and see analysis) Voice of Customer (VoC) Work Standarization 5S (Seiri/Seiton/Seisho/Seiketsu/Shitsuke) Pull Logistic Continues Improvement (Kaizen)
  • 10. Faculty of Computer Science Universitas Indonesia Research Methodology Research Design: Qualitative Research Methode: Case Study & Action Research Data Collection : Interview, Observation, FGD 1 Problem Identification 3 5 2 4 Literature Review Observation Phase Design Phase Implementation Phase
  • 11. Faculty of Computer Science Universitas Indonesia Lean Implementation (Value Identification) Value Identification Voice Of Customer (VoC) : No Identified Values Achievment Yes No 1 Technicians able to quickly see and responds to every occured problems O 2 Management able to quickly find out which services are interrupted O 3 Executives can quickly learn which business processes are disrupted O 4 Able to provide accurate alarm information O 5 Able to monitor the use of resources to maximize the use of the service O 6 There are rules in the use of each feature on the monitoring system O 7 There is a clear workflow in handling every alarm O 8 Self service monitoring for IT service owners O 9 Simple display that can be directly understood by ordinary users O
  • 12. Faculty of Computer Science Universitas Indonesia Lean Implementation (Value Stream Mapping)
  • 13. Faculty of Computer Science Universitas Indonesia 1 3 2 Waste Identification (Information Collection) There are resources that are not being used and are still monitored by the system There is a function of resources that are not owned by the device, but monitored by the system There are resources that are not enabled by the device, but monitored by the system Waste Inventory
  • 14. Faculty of Computer Science Universitas Indonesia Waste Identification (Information Processing) Alerting Waste Process : 1. Event Trigerred Over Processing 2. Alert Logged Over Production 3. Predefined Rules Defect 4. Review The Alert Over Production Source : (Nand & Chaganty, 2008)
  • 15. Faculty of Computer Science Universitas Indonesia Lean Implementation (Create New Flow by Elemeninate Waste) Remove Unecessary Thing Stop monitoring of devices that are not in use. Stop monitoring of resources that are not being used by the device. Stop monitoring of the resource function that the device does not have. Stop monitoring of resources that are not enabled by the device. Removing unnecessary alarm configuration. Arrange Placements and Mark / Label on All Things Organize categorization and mark devices based on area, city, region, service name, and service owner. Organize grouping and marking devices based on service function. Self service access to monitoring system based on ownership of device or service. Remove the unnecessary dashboard view and add another more informative view by creating a single dashboad view(Next 際際滷) Cleaning and Tidying Up Display Create Working Standard Making standard operational procedure (SOP) Make sure SOP is running A reward system is applied for operators who perform every month, with the assessment category being compliance with SOPs. It has been proposed that the activity in the SOP be the engineer performance assessment criteria which listed in the Individual Work Plan (RKI). It has been proposed that adherence to this SOP be the point of disciplinary Final Assessment (PA) of the engineer.
  • 16. Faculty of Computer Science Universitas Indonesia Create Pull System Aiming for the system to be only produce information when it is needed by customers at the right time, right amount and right destination. Set the sending of alarms according to working hours on certain devices or services Make maintenance scheduling arrangements if there is downtime scheduled Send an alarm to a person who is responsible for a service Determine the parent and child of the devices so that, if there are problems with the parent device, the child device will not send an alarm
  • 17. Faculty of Computer Science Universitas Indonesia Make Perfection Continues Improvement (Kaizen) Make procedures for using the monitoring system Implement best practice management monitoring system (next slide)
  • 18. Faculty of Computer Science Universitas Indonesia Make Perfection (Device Inventory)
  • 19. Faculty of Computer Science Universitas Indonesia Make Perfection (Business View)
  • 20. Faculty of Computer Science Universitas Indonesia Implementation Result Quantitatively after implementation, the number of devices monitored increases while alarms and events decreases Qualitatively, remains expected value has been achieved No Identified Values Caused Waste Activty Achievement Yes No 1 Technicians able to quickly see and respond to problems that occur Inventory Motion Waiting Defect Short Shine Create pull system O 2 Management able to quickly find out which services are interrupted Motion Waiting Set in order Perfection (Create business view) O 3 Executives able to quickly learn which business processes are disrupted Motion Waiting Set in order Perfection (Create business view) O 4 Able to provide accurate alarm information Inventory Defect Over Production Over Processing Short Create pull system O 5 Able to monitor the use of resources to maximize the use of the service - - O 6 There are rules in the use of each feature on the monitoring system Not directly impact by waste Standarize O 7 There is a clear workflow in handling every alarm Not directly impact by waste Perfection (Create SOP) O 8 Self service monitoring for IT service owners Not directly impact by waste Set in order O 9 Simple display that can be directly understood by ordinary users Motion Waiting Shine Make Perfection (Create business view) O Network Elements Nodes Interfaces Alerts Events Amount Before Lean 7683 639 6613 111 348172 Amount After Lean 10705 1308 6683 26 177343 Percentage Ratio 39% 105% 1% -77% -49%
  • 21. Faculty of Computer Science Universitas Indonesia Summary After the lean method is implemented, the system can convey the expected value of the customer Waste in the monitoring system can be reduced so that there is an increase in accuracy related to information on the condition of the IT infrastructure Lean method can be applied to manage monitoring system through the steps of applying lean principle as follows: VoC (Voice of Customer) SIPOC of Monitoring System Remove Monitoring Waste Send alert on the right time, right amount, right destination Implement Kaizen
  • 22. Faculty of Computer Science Universitas Indonesia Source: xxx Thank You

Editor's Notes

  • #2: Good morning everyone, my name is Dio Pratama from University of Indonesia. Im here today going to share and describe about my research, its about Tittle
  • #3: As a brief, Kalbe Farma is one of largest Health Care & Phrameutical company in South East Asia. We have about .
  • #4: Just explain
  • #5: It proven by auto generate ticketing system, that IT Infra department has produce the most total incident. And actually that incidents is mostly produced by our monitoring system.
  • #8: By literature review, everyone agree that IT Infrastructure is one of key operational services that need to be monitored in proper management. One of of popular process of management methode is lean. Lean