College Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
Meetup Devops April 2023
1. Jérémy Bivaud & Olivier Eeckhoutte
DevOps at Tag Heuer
Feedback of our key success factors
2. About us
Devops/ Tech team at TAG Heuer
• Little and dedicated Tech team of 4 (dev. background) created ~ 4 years ago
• Today speakers
• Bivaud Jeremy
• Tech Lead @Tag Heuer
• MACH evangelist
• Devops addict
• Olivier Eeckhoutte
• Started Freelancing 7 years ago
• Web3/Crypto trader ~ 3 years ago
• DevOps Engineer Mandate at Tag Heuer
#curious #hyperactive
https://www.linkedin.com/in/oliviereeckhoutte/
@oliviereeckhout
https://github.com/oeeckhoutte
Blog: https://leandeep.com/
https://www.linkedin.com/in/jeremy-bivaud-2786a636
https://github.com/jeremybiv
jeremy.bivaud@tagheuer.com
J+O
3. What is DevOps ?
Little reminder
• A philosophy not only job position
—> cultural change (break barrier between IT Operations and
Developers)
—> Everyone should embrace this change
• Improvement in apps deployments
• Apply software development best practices in infrastructure
• More reliability (—> SRE)
O
4. Agenda
• What is DevOps ?
• Project context presentation
• Our Devops Culture
• Automation
• Monitoring
What is working very well (for us) and what we would recommend for new
projects
O
5. Our mission
What we do on a daily basis ?
• We create the glue between the di
ff
erent systems
• Agile developments:
• Homemade ESB (Enterprise Service Bus) / Middleware platform
• REST APIs
• Projects industrialisation on Azure
• Active Monitoring and RUN
• Devops ratio: Dev (75%) & Ops (25%)
J
6. Our project
Software Architecture (over simpli ed)
• Event driven Architecture
• Light CQRS (Command and Query Responsibility
Segregation) for:
• Fast feedback
• Do not loose customers actions (queues &
topics with retry mechanisms leveraging Azure
Service Bus…)
J
7. Our project
Technical Architecture (concept)
• Hub & spoke (Pros: Centralisation, Security, Scalability,
Costs reduction and Perf)
• Hub with Software-De
fi
ned Network Firewall (Pull
Requests + CI/CD)
• Spoke with AKS (Azure Kubernetes Service)
• Highly secured (Bastion + VPN + MFA + OTP)
References on this Architecture:
• https://cloud.google.com/architecture/deploy-hub-spoke-vpc-network-topology
• https://learn.microsoft.com/en-us/azure/architecture/reference-architectures/hybrid-networking/hub-
spoke
• https://docs.aws.amazon.com/wellarchitected/latest/reliability-pillar/
rel_planning_network_topology_prefer_hub_and_spoke.html
O
9. DevOps culture
• Automate as much as possible (today we are able to Release/con g change in 5” and do a
cluster migrations with 0 downtime e
ff
ortlessly)
• Continuous Improvement (i.e. Devops meetings in addition to Agile retro.)
• Sharing (Be totally replaceable—> Your colleagues will appreciate 😀)
• (Absolutely) Everything Code (Dev, Con
fi
g, Secrets, CI/CD pipelines con gs, Infra…) with
automatic CI validations and Pair Reviews
• KISS principle is well know by developers
-> Apply it also for Ops tasks (and do the same with all other development best practices
and others. Ex: 12 factors https://12factor.net/ or summary: https://bit.ly/3nmKK4D )
O
10. Automation
Infrastructure As Code
• 95% of our infra stack can be rebuilt using code (the missing 5% are due to the tools limitations or
priority)
• Stack:
• Security should is a priority (Zero trust and do not trust yourself)
• Give yourself the ability to test any technical brick individually to quickly troubleshoot issues (Example
CDN and Load balancer)
• Do not try to host everything yourself (DB, heavy duty external services…) —> Lean: Business value 1st
• Be a not a
J
12. Automation
Continuous Integration &
Continuous Delivery
• AKS CI validation (Automatically validate YAMLs)
• Terraform CI validation
• Automate end2end tests
• Code validation and Docker build via CI (and even locally before push)
• Unit tests (Test Driven Development)
• Functional tests (Behaviour Driven Development)
• Python tooling stack: pre-commit +
fl
ake8 + Black + mypy + sonar + pytest…
O
13. Monitoring
Track, Monitor everything and add alerting
• Stack:
+ Heartbeat (for SLA/SLO) + Alert Manager + JIRA (communication +
traceability) + Automated End 2 end tests (usable in ≠ environments)
• Spend time on your logs (format them to troubleshoot issues e ciently)
J
15. Our project
Next (short term)
• Continue to improve and learn
• Proactive: detect issues before anyone detect them
-> Business Activity Monitoring (I.E: verify integrity in all systems)
O