Open Source Approach in Orchestration & Monitoring
Ufuk Tan Baler, MSc.
3 February 2026
[EN] This work cannot be used for commercial purposes and/or advertisement, and does not involve any commercial purpose, financial interest, or claim of official partnership or endorsement. This work is a technical presentation which has been prepared in the process of my learning and professional development. The content in this presentation includes container based systems and monitoring architectures. The author assumes no liability for any damages arising from the use of this material.
[TR] Bu çalışma ticari ve/veya reklam amaçlı kullanılamaz ve herhangi bir ticari amaç, gelir beklentisi veya resmi iş ortaklığı iddiası veya desteği içermemektedir. Bu çalışma kişisel öğrenme ve mesleki gelişim sürecimde hazırlanmış teknik bir sunumdur. Sunum içeriği container tabanlı sistemler ve monitoring mimarilerini içermektedir. Sunum içeriğinin kullanımından doğabilecek doğrudan veya dolaylı zararlardan sunumu hazırlayan kişi sorumlu değildir.
development and deployments
\(\Rightarrow\) compatibility issues
further issues in cloud paradigm
- fluctuating request numbers
- immediate provisioning of computing resources
- latency
- service monitoring
state of the art
open source solutions
containers
a platform demanding
- lightweight
- OS agnostic
- easily deployable
Dockerfile: docker build -t demo/myimage .
FROM busybox
COPY ./hello.sh ./
CMD ["sh" , "./hello.sh"]
run the image,
sudo docker container run --rm demo/myimage:latest
hello from this presentation!
in practice
kubernetes (k8s)
levels of the infrastructure:
| 6 | application |
| 5 | container |
| 4 | services |
| 3 | pods |
| 2 | nodes |
| 1 | cluster |
- horizontal scaling
- distributing workloads
- actual \(\rightarrow\) desired
- services enable access
deployment of infrastructures
| pod.yaml | replicaset.yaml |
| service.yaml | deployment.yaml |
kubectl apply -f pod-nginx.yaml
monitoring
- templates: storage, network, etc.
- hosts: observed devices
- triggers: thresholds
- graphs: transient states
- dashboard: coherent view
zabbix helm repository
proof of concept
