Want to become an AppDynamics expert? In this essential session, you’ll learn about best practices for configuring Business Transactions, role based access control and other top tips for APM success.
ACKNOWLEDGE: Welcome to this AppDynamics, thanks for being here.
DANCE: Put some music and make people dance.
BRIEFING: Dancing is something different, unexpected. Explain that disruptive changes in the industry.
ETR: Earn The Right. Why I’m here talking.
I have 10 best practices that could really help.
Challenge the audience to consider and evaluate how these best practices, both individually and as a cohesive system, can enhance the effectiveness of the APM solution.
3 segments
You can choose this tips and see what works for you
This best practices, comes from the experience of years working with customers.
People are essential and the foundation
Addresses directly and indirectly many roles
Goal is enablement and adoption of different users across various teams and disciplines
Technical problems can be solved, its mostly the human dimension which gives the biggest headache
Customer have incomplete strategy and just wants to roll out agent. Its NOT just turn on and the rest is done automagically
Franklin Covey: For every 1 min in planning, you get 5min back in the execution of the plan.
APM with AppD must save money or give a competitive advantage. It is an investment and not a technical playground. Hence, the solution must be aligned with business initiatives and Management and engineering must be aligned.
Strategy: Analyze current problems, define plan, set goals, set key initiatives, trade-offs among activities,
It is NOT, lets install and see later, its NOT lets turn on as many feature as AppD has and hope for the best, hope is NOT a strategy
When we define the strategy we make sure AppD is a solution not just a tool!!!
This is the WHY, HOW and WHAT
Customer have incomplete strategy and just wants to roll out agent. Its NOT just turn on and the rest is done automagically
Franklin Covey: For every 1 min in planning, you get 5min back in the execution of the plan.
APM with AppD must save money or give a competitive advantage. It is an investment and not a technical playground. Hence, the solution must be aligned with business initiatives and Management and engineering must be aligned.
Strategy: Analyze current problems, define plan, set goals, set key initiatives, trade-offs among activities,
It is NOT, lets install and see later, its NOT lets turn on as many feature as AppD has and hope for the best, hope is NOT a strategy
When we define the strategy we make sure AppD is a solution not just a tool!!!
This is the WHY, HOW and WHAT
WIKI
links, education, onboarding, naming conventions, change management, internal playbooks
Playbooks to document common processes
If you don’t measure you don’t know your status. And only worse than that is not knowing what to measure.
Track progress, leading indicators
Know why you measure -> noise
Development
Number of backlog Items
Project backlog
Number of warranty repairs
Time spent on production issues
Project throughput
Number of developers
Hours on project
IT OPS
MTTR and MTTI
Availability and service response time
Service availability
Service up-time
Transaction throughput
Number of issues
Cost of issues
Business
Conversion rate
Completion rate
% of online revenue
Key customer journey response time
Interaction response times
Usage by channel
Revenue impact
New principles how AppD solve APM problems: Business IQ
There’s a secret sauce. Make sure you understand our basic architecture in order to use it to its full advantage
DON’T try to avoid large flow maps on application level
Your flowmaps are a result how you define node, tiers and applications
hierarchical structure of node, tiers and applications.
Its nothing else than the true and actual architectural picture of your environment.
Leave together what belongs together
AppD provides means to slice and dice, not done by ripping apart
A app is a tier, not an application…
For large corps, create apps e.g. by support group
While it looks noisy to the eye, in fact it is a strict order, but there are places we may encounter noise
Where is the needle
Noise is created by ”Trace On”
We look for information
The main config artifact is the BT
See lots of customer tending to create to much noise here
What do you want to measure? It is NOT to measure ALL transaction individually!!
Technically it’s a filter
Be sure you are within limits
Rule of thump: If you have BTs which are not used for Alerts or dashboards they are not real BTs….
As I just talked about Dashboard and Alerts, lets see if there is a Best Practice around it
Configure Instrumentation
App agents ship with default instrumentation settings that cover the most common types of application frameworks and programming patterns. If your requirements are more complex, you can customize the instrumentation behavior to suit your environment.
Transaction detection rules
Backend detection rules
Error detection
Service endpoints
Diagnostic data collectors
Call Graph Settings
Configure JMX Metrics from MBeans.
Memory monitoring settings
Object Instance Tracking for Java
Object Instance Tracking for .NET
Object Instance Tracking for Node.js
Asynchronous transaction demarcators
Lots of control and apis to configure and extend your solution.
One example I want to pick out is the Health Rules and I want to show you a way to make more sophisticate health rules
Transaction Performance
Load
Response time
Slow calls
Stalls
Node and Infra
CPU
Heap usage
Disk I/O
Connection pools (JMX)
Thread pools (JMX)
Response time, load, or errors
User Experience
DOM building time
JavaScript errors
Requests per minute
Ajax callback execution time
End user response time
HTML download time
App crashes, starts, server calls, network requests and errors
People and Product must be supported by aligned Processes
Many customers take privacy and security very serious and want to understand the implications of using appd.
Out of the box we do NOT collect any PCI relevant data
Power comes with responsibility
-> MIDC
-> Raw SQL
Solution: RBAC for these features
Devops Culture
Automate everything that you could do twice or more
Collaboration mindset
Better: We don't settle for the status quo. We challenge ourselves to do better every day, every moment. We do it ethically and with integrity. Where greatness is the destination, better is the journey.
Together: We are one team. Diversity makes us stronger. We invest in our people and have each other's back. Together, we take pride in the phenomenal work we do.
Transformative: We are Agents of Transformation. We are the vanguard bringing industry from siloed, big data, backward-looking analysis into a real-time contextual, predictive, actionable world.
Champions: We are champions of our customers. We are dedicated to solving our customers' hardest problems. Their success is our success. Together we dream big.
Positive: We want more than just a job. Every moment we have a choice. We choose positivity. We choose to be a force for good in the way we show up, work with our teams, and care for the communities we live in.
Fun: We love what we do. We're serious about not taking ourselves too seriously. We are passionate about creating an environment where people love coming to work.