Im looking at options for fast scaling the single server environment I have now
For cost reasons Im investigating Cloudmin to understand the brutal truth of its ability to handle multiple servers and containerization scaling of a Docker based application. AI suggests Cloudmin is acceptable for managing multiple servers but needs further help for from NGINX for load balancing and that it doesnt scale containerisation at all but I wanted to to talk to you guys to hear your thoughts.
I use a 12GB ram server that costs $14 a month, its a non profit so we like low costs. Alternatives like Google Cloud Compute are $140 for the same spec and Im sure Amazon is even higher.
Cheap is good but we cant sacrifice performance too much. Interested to hear experiences and thoughts. Im not a network engineer and have limited knowledge about this but still need to make the right call for the non profit, advice will be received gratefully
Cloudmin is not very strong with regard to containers. It’s just not a very good fit. If containers are how you want to do things, Cloudmin is not a good solution.
Cloudmin is stronger when it comes to managing virtual machines using KVM.
So, if you have one or more big servers you want to split up into a bunch of virtual servers, Cloudmin can do that, and it integrates well with Virtualmin to automate a pretty broad bunch of web hosting and management tasks.
If you want a container orchestration system (i.e. Kubernetes), Cloudmin is simply not the right tool for the job. Kubernetes is a very expensive and complicated way to manage compute resources, and not appropriate for a lot of things people munge into it (I work with Kubernetes in my other job)…most of the time when I see folks using it, it’s because they just saw that some big megacorp was using it, and assumed it must be the best thing. But, you probably don’t have big megacorp problems to solve and you probably don’t have big megacorp budgets.
But, it sounds kinda like you’re thinking of containers as just a way to divide up a big server. Which is not really how the container ecosystem (Docker, Podman, Kubernetes, etc.) thinks about it. You may find you’re trying to fit a square peg into a round hole if what you want is a bunch of little Linux systems running on a big Linux system (that’s a job for virtual machines, IMHO).
Thanks, the app we use is Docker deployed and so far as I understand it we don’t have a choice, we need to be able to scale containerization. We aren’t doing it because the megacorps do you can rest assured of that. From what you are saying Ive decoded that as meaning that we either use Kubernetes or we use Amazon/ GCP infrastructure. I’m not so much thinking about how to divide up a big server but more how to piece smaller ones together but scaling containerisation. I don’t personally have enough systems engineering knowledge to understand and actually do this, so it could be that we have to pay 10 times the price and run with GCP
Using containers to deploy applications is orthogonal to something like Kubernetes or container orchestration. If you’re using it as a (big bulky kinda uncomfortable) package manager, which lots of people do (including me, sometimes, this forum is running a Docker container built by the Discourse folks), it does not follow that you must use something like Kubernetes (whether Google-hosted or otherwise…there are many ways to host Kubernetes, all are expensive).
I think you’d be best served by taking some time to understand the lay of the land and what problems you actually need to solve before making decisions. I’m not saying this because I’m trying to convince you to use Cloudmin. I don’t think Cloudmin is a good fit for your needs. I’m saying this because I think you’re about to waste a lot of money and time on suboptimal solutions because you’re thinking the problem you need to solve is mostly about containers, and my view from afar doesn’t seem like that’s true.