summaryrefslogtreecommitdiff
path: root/spec.md
blob: f8a11a7b6565fa266328e1f62fa7b72e20dc9ec3 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
# Spawning a Kubernetes Cluster from SystemD

##### Overview:
    the idea of this program is to establish a linux operating system that basically boots into being one functioning node of a kubernetes cluster. it uses the configs in its disk storage to know its node label/category, if it's a master, a kafka coordinator, etc. 
    the user configures each system component, down the applications running in each OCI container, generates an iso, installs it to each node of the cluster, boots, ...??, profit!

## First-time boot: (install)

    * write filesystem (which daemons to run, given the host (cluster node). configs for the various daemons. etc

## Normal boot:

#### start the following systemd services
    - kubernetes
    - ceph (or other distributed storage system)
    - dns
    - kafka
    - mqtt
    -
    -
#### service-specific behavior
    - kubernetes
        - data-pipeline services (e.g. use kafka messaging, springboot, rabbitmq, nginx)
        - monitoring, logging, tracing, observability
    - ceph