This document outlines 5 key skills that enterprise administrators need to know to manage Atlassian products at scale. The 5 skills are: 1) Governance - centralizing setup, teams, and processes; 2) Performance - understanding factors that impact performance and implementing quick wins; 3) Monitoring - establishing monitoring strategies, metrics, and tools; 4) Upgrade Strategy - deciding on release timing, planning upgrades, and leveraging experts; 5) Human Side of Scaling - empowering users, implementing change control, and providing support. The document provides examples and recommendations for each skill area from Atlassian and Airbnb's experiences scaling their Atlassian instances.
21. Centralise setup
Two ways: Scale single instance or adopt
Datacenter
Centralise teams
Establish one team that administers Atlassian
products
Centralise processes
Make sure that there are processes set in place
across the Atlassian product suite
27. Understand access levels
Sys-admin vs admin vs project admin
Limit access
Give the right access to the right people
Regular audit of access
Continually reassess whether the access is still
required
32. Establish knowledge base
A single source of truth for users to understand
Atlassian products
Outline intended use
How does your company use Atlassian products?
33.
34.
35.
36. Establish knowledge base
A single source of truth for users to understand
Atlassian products
Outline intended use
How does your company use Atlassian products?
Create open channel of communication
Through communication mediums like Slack or
support desks
43. Number of issues ≠ performance
Don’t use only number of issues as an indicator of
performance
44. Number of issues ≠ performance
Don’t use only number of issues as an indicator of
performance
Understand other factors that impact
performance
E.g. Usage patterns, configuration, deployment
45. Number of issues ≠ performance
Don’t use only number of issues as an indicator of
performance
Understand other factors that impact
performance
E.g. Usage patterns, configuration, deployment
Understand your environment
Everyone’s environment is different, having a firm
understanding of yours is crucial
49. Hardware limits
It’s essential that your servers are correctly sized for
your usage level
Usage limits
Understand your usage patterns, especially when it
comes to integrations
50. Hardware limits
It’s essential that your servers are correctly sized for
your usage level
Usage limits
Understand your usage patterns, especially when it
comes to integrations
Configuration complexity
Products do not scale infinitely, ensure that your
environment can handle your level of customization
53. Understand user needs and goals
What are your user’s goals and what do they need
to be successful?
54. Understand user needs and goals
What are your user’s goals and what do they need
to be successful?
Explain “why”
Behind all your actions
55. Understand user needs and goals
What are your user’s goals and what do they need
to be successful?
Explain “why”
Behind all your actions
It’s not you, it’s them
Sometimes users screw things up
59. Custom field context
Get in the habit of setting this for all new custom
fields!
Plugin: Cleaner for Jira
This plugin can help clean up your instances as you
establish better habits
62. Turn off empty JQL Search
This is a big improvement in perceived performance,
and very easy to fix!
Custom field context
Get in the habit of setting this for all new custom
fields!
Plugin: Cleaner for Jira
This plugin can help clean up your instances as you
establish better habits
63.
64. Know what to
monitor
Establish a
monitoring
strategy
Think about
non-alerted
metrics
External
monitoring
tools
Skill #3: Monitoring
65. Know what to
monitor
Establish a
monitoring
strategy
Think about
non-alerted
metrics
External
monitoring
tools
Skill #3: Monitoring
66.
67. There are two types of metrics
Usage and Infrastructure
68. Hosting options
CPU/Disk speed
Available memory
JVM/JMX statistics
Database performance
Network speed
Number of issues
Number of custom fields
Number of concurrent users
Activity of major integrations
Available license seats
USAGE INFRASTRUCTURE
69. There are two types of metrics
Usage and Infrastructure
Establish alert levels
Behind all your actions
70. Pager
🚨 JIRA IS DOWN 🚨
Critical CPU/Disk consumption (>90%)
Manual incidents
Slack Channel
Mail hasn’t been processed for x minutes
Staging/Development node down
Early warnings for CPU/Disk (>65%)
71. There are two types of metrics
Usage and Infrastructure
Establish alert levels
Behind all your actions
Outline alert actions
What will you do once this alert is triggered?
72. Know what to
monitor
Establish a
monitoring
strategy
Think about
non-alerted
metrics
Consider other
tools
Skill #3: Monitoring
75. Set realistic targets
Understand your instance before setting these
targets
Key Performance Indicators
Identify KPIs that show you that your instance is
performing at an optimal level
76. Key Performance Indicators
Identify KPIs that show you that your instance is
performing at an optimal level
Set realistic targets
Understand your instance before setting these
targets
Service Level Agreements
Establish SLAs to react to alerts to ensure you’re
effective
77. Know what to
monitor
Establish a
monitoring
strategy
Think about
non-alerted
metrics
Consider other
tools
Skill #3: Monitoring
80. Number of users/fields in Jira
The rate of which this is growing
Number of active HTTP threads
Any easy metric to see how actively used the
product is.
81. Number of users/fields in Jira
The rate of which this is growing
Number of active HTTP threads
Any easy metric to see how actively used the
product is.
Mail sent by Jira
Jira is chatty, it’s good to keep an eye on it
82.
83. Know what to
monitor
Establish a
monitoring
strategy
Think about
non-alerted
metrics
Consider other
tools
Skill #3: Monitoring
86. Java Melody
Monitor the health of your JVM
New Relic
End-to-end transaction tracing, application
response times, JVM analysis, thread profiling
87. Java Melody
Monitor the health of your JVM
New Relic
End-to-end transaction tracing, application
response times, JVM analysis, thread profiling
Dynatrace
End-to-end tracing and monitoring.
88. Incident alerting based on
data from monitoring tools
Airbnb’s Monitoring Tools
Monitoring infrastructure and
usage metrics
Log aggregation and
monitoring
89. Decide what
release is right
for you
Plan your
upgrade
Upgrade your
instance
Lean on experts
Skill #4: Upgrade Strategy
90. Decide what
release is right
for you
Plan your
upgrade
Upgrade your
instance
Lean on experts
Skill #4: Upgrade Strategy
93. Platform release
(e.g. Jira 8.0) contains significant or breaking
changes
Feature release
(e.g. Jira 7.9) can contain new features,
improvements etc
94. Platform release
(e.g. Jira 8.0) contains significant or breaking
changes
Feature release
(e.g. Jira 7.9) can contain new features,
improvements etc
Enterprise release
(e.g. Jira 7.13) will receive bug fixes for a longer period
of time than a standard feature release
95. Decide what
release is right
for you
Plan your
upgrade
Upgrade your
instance
Lean on experts
Skill #4: Upgrade Strategy
98. Understand changes in supported
platforms
Dedicated “Supported Platforms” page
Run instance health checks
Ensure your instance is healthy before an upgrade
99.
100. Understand changes in supported
platforms
Dedicated “Supported Platforms” page
Run instance health checks
Ensure your instance is healthy before an upgrade
Pre-Upgrade Planning Page
Tailored upgrade path specific to your instance
101.
102. Decide what
release is right
for you
Plan your
upgrade
Upgrade your
instance
Lean on experts
Skill #4: Upgrade Strategy
103.
104. Always test your upgrade
And ensure you document all your learning
105. Always test your upgrade
And ensure you document all your learning
Utilize Data Center upgrade features
Jira zero downtime upgrades and Confluence read-
only mode
106. Always test your upgrade
And ensure you document all your learning
Utilize Data Center upgrade features
Jira zero downtime upgrades and Confluence read-
only mode
Automate (where possible)
Airbnb leverages Chef and Terraform to make
upgrades as painless as possible
107. Decide what
release is right
for you
Plan your
upgrade
Upgrade your
instance
Lean on experts
Skill #4: Upgrade Strategy
110. Priority & Premier Support
Coverage for Atlassian applications at scale
Technical Account Management
Strategic scaling of Atlassian products and
managing for growth
111. Priority & Premier Support
Coverage for Atlassian applications at scale
Solution Partners
For solutions, training and strategic services
Technical Account Management
Strategic scaling of Atlassian products and
managing for growth
116. Atlassian University and Certification
Online classes, live team training & tutorials
Power User Groups
Team of power users that learn from each other
117. Atlassian University and Certification
Online classes, live team training & tutorials
Power User Groups
Team of power users that learn from each other
Establish Subject Matter Experts
For particular use cases such as specific add-ons
121. Understand the “why”
What problem are they trying to solve?
Test all changes
In a non-production environment
122. Understand the “why”
What problem are they trying to solve?
Test all changes
In a non-production environment
Have a backup plan
In case things go wrong
127. Establish single source of support
E.g. Jira Project
Have a dedicated comms channel
E.g. Slack
128. Establish single source of support
E.g. Jira Project
Have a dedicated comms channel
E.g. Slack
Be available
Balance project work vs operational work
133. 1
4
3
2
Establish governance early on
Learn how your instance performs & set guidelines
Monitor changes in your instance and set alerts
Develop an upgrade strategy & plan to update regularly
Skills recap
5 Recognise that users are only human