Is K8S ever suitable for small business?
K8S is only supported on expensive hosts (AWS, GCE, Azure), is it possible to spin it up on hetzner/ovh or another provider that is cheaper than AWS, possibly even on my own hardware? How hard it would be to make the simplest reliable prod environment on bare metal with K8S?
I'm at the moment running 1 (tech) man SaaS with minimum profits, here are my goals (1,2,3 are essential):
1. Be cheap - It's a side project, I don't need to sacrifice money for speed as I enjoy what I'm doing.
2. Automatic failover (This should cover majority of downtime imho)
3. Backups
4. Automatic scalability (CPU over 95% for last 2 minutes or smth like that? add new containers pls)
5. Load balancing (using erlang so this can be handled on application level if nodes can see eachother)
6. If gitlab autodevops can work out of the box that would be fantastic
Unrelated to above:
- How are K8S backups handled? I spoke with a friend (K8S fanboy) and he said you have several containers spawned of the DB and they are all interconnected. If one of them fails another one has the data. This seems stupid to me, what if the whole zone fails and all DB containers are shut off?
- Same friend asked me: "What restarts your services if they fail?". "init.d". "And what if the machine fails?" Well I gotta restart it manually. "You see, you should use K8S orchestrator". "What restarts K8S orchestrator if it fails?" <Silence>
^ I have a feeling big part of community is like that, "super cool feature of K8s you should use it" but there is no connection to the rest of the context.
Ps. conversation was in another language so I'm not sure I used the right terms eg. "K8S orchestrator"
I'm at the moment running 1 (tech) man SaaS with minimum profits, here are my goals (1,2,3 are essential):
1. Be cheap - It's a side project, I don't need to sacrifice money for speed as I enjoy what I'm doing.
2. Automatic failover (This should cover majority of downtime imho)
3. Backups
4. Automatic scalability (CPU over 95% for last 2 minutes or smth like that? add new containers pls)
5. Load balancing (using erlang so this can be handled on application level if nodes can see eachother)
6. If gitlab autodevops can work out of the box that would be fantastic
Unrelated to above: - How are K8S backups handled? I spoke with a friend (K8S fanboy) and he said you have several containers spawned of the DB and they are all interconnected. If one of them fails another one has the data. This seems stupid to me, what if the whole zone fails and all DB containers are shut off? - Same friend asked me: "What restarts your services if they fail?". "init.d". "And what if the machine fails?" Well I gotta restart it manually. "You see, you should use K8S orchestrator". "What restarts K8S orchestrator if it fails?" <Silence> ^ I have a feeling big part of community is like that, "super cool feature of K8s you should use it" but there is no connection to the rest of the context.
Ps. conversation was in another language so I'm not sure I used the right terms eg. "K8S orchestrator"