Post on 09-Nov-2018
transcript
Success Factors in APM Adoption
By: Larry Dragich Director, Enterprise Application Services
February 2013
Auto Club Group
Enterprise Systems Management:
Provide proactive system monitoring, maximizing system utilization to support the business needs.
Vision
Roles
Enterprise Systems Management will be the focal point for IT performance monitoring and capacity planning activities; achieved by partnering with the Subject Matter Experts (SME’s) within each of the technical domains and the application development areas.
Performance Monitoring (ESM)
ESM will be the focal point for performance monitoring activities, (i.e. Data Center alerts, Trouble Ticket Interface, Event Correlation) for events that occur within the Infrastructure and it’s components
Performance Tuning (SME’s)
Other technical domains, (e.g. Network, Server, Application Development), are responsible for tuning activities to make efficient use of resources, (i.e. defining thresholds, transaction timings, instruction text)
Continual Service Improvement (CSI)
Using the CSI concept (ITIL v3.0), our team drives the Application Performance Management (APM) meetings that convene bi-weekly to review and improve critical business application performance.
Larry Dragich, Director EAS
The Auto Club Group – April 2012
End User
Experience
Top Down
Monitoring
Reporting
(Metrics)
Bottom Up
Monitoring
Incident
Management
(ITIL)
End User
Experience
The EUE provides one of the highest values within the five dimensions of APM as defined by Gartner, in terms of application visibility for the business.
APM is the translation of IT metrics into business meaning (value). This is accomplished through multiple technologies and interlinking processes.
The success factors in APM adoption center around the EUE and the integration touch points with the Incident Management process.
Top Down
Monitoring
This is also referred to as Real-time Application Monitoring which is the cornerstone that gives the EUE its tangible value. It has two has two components:
Passive monitoring is usually an agentless appliance which leverages network port mirroring. Also referred to as Real User Monitoring (RUM) technology.
Active monitoring consists of synthetic probes and web robots which help report on system availability and predefined business transactions.
Bottom Up
Monitoring
This is also referred to as Infrastructure Monitoring which usually ties into an operations manager tool.
The Manager of Managers (MoM) becomes the central collection point where event correlation happens. System automation is the key component to the timeliness and accuracy of incidents being created.
Reporting
(Metrics)
Capturing the raw data for analysis is essential for an APM strategy to be successful.
These are key reporting metrics. Use 5 minute averages for real-time performance alerting and use percentiles for overall application profiling and Service Level Management
It is important to arrive at a common set of metrics that you will collect and then standardize on a common view on how to present the real-time performance data.
Incident Management
(ITIL)
The Incident Management Process as defined in ITIL is a foundational pillar to support Application Performance Management (APM).
This is a key component to the timeliness and accuracy of incidents being created through the Event Management process.
APM supports the CSI model and ties together specific processes in Service Design, Service Transition, and Service Operation.
Top Down Monitoring
Passive Monitoring (Port Mirroring)
Active Monitoring (Robots / Probes)
Bottom Up Monitoring
Reporting
(Metrics)
Incident Management
(ITIL)
Reporting – Service Level Management (SLM)
End User Experience
Events Incidents
TTI Engine
Data Center
Operations Manager
Event Correlation
Incident Management Service Desk
Metr
ics
Metr
ics
Metr
ics
Application Env. End-User-Experience
Incident Management Service Desk
Data Center
Operations Manager
Event Correlation
Passive Monitoring (Port Mirroring)
Active Monitoring (Robots / Probes)
Application Env. End-User-Experience
Reporting – Service Level Management (SLM)
Events Incidents
TTI Engine
Metr
ics
Metr
ics
Metr
ics
Enterprise Mgmt
Tools Device / App Agnostic
Feeder Systems
Other App Monitors Device / App Specific
Feeder Systems
Bottom Up Instrumentation
Infrastructure Monitoring
Infrastructure Agent Monitoring
SNMP Trap Receiving
Process Monitoring / Ping Scripts / Perl Scripts
Top Down Instrumentation
Application (Users Perspective)
Real User Monitoring (RUM) – Agentless
Synthetic Transactions (Probes Robots)
User Experience Mgmt. (UEM) Script Injection
Incident
Manager
Audible
Alerts
Ops
Console 3rd Party
Connectors
Trap
Listener Enterprise
Managers
SNMP
BSM
Web
Probes Service
Probes
Analysis
Engine
J2EE / .NET
Agents
RUM
Agentless
Incident MoM
Enterprise
Agents
3rd Party
Alarms
Ops
Agents
Front
Door (Custom)
Incident
Output Logical
Connection
ESM
System Monitor
Larry Dragich – EAS
Chris McDevitt, IT Architect
The Auto Club Group – May 2012
Data Center -------------------------- Operations Manager
-----------------------------
SNMP Listener
Larry Dragich, Director EAS
The Auto Club Group – May 2012
Real User
Monitoring
Routers /
Switches
UPS
Devices
PBX
Switches
Synthetic
Probes
Network
Sniffers
Web
Robots
Virtual
Servers
Encryption
Devices
WAN Optimization
Firewalls Servers
Bridge
Connector
Agent
Protocols SNMP
Traps
SNMP
Traps
Service Design
Service Level
Management
Availability
Management
Capacity Management
Service Transition
Change Management
Release Management
Service Operation
Event Management
Incident Management
Problem Management
Continual Service Improvement
Application Performance Management
Larry Dragich, Director EAS, The Auto Club Group – March 2012