Not specifically related to Oracle Cloud itself, but I am curious how folks remain on the Free Tiers of these providers, such as Oracle Cloud, GCP, AWS, etc.
It seems, from a glancing review, all of these services structure the Free Tier to force an "on-demand" or "serverless" architecture, since the CPU-seconds and GB-seconds are always undersized for an always-on system (such as a traditional server or OCI container).
For hobby projects or book exercises, the Free Tiers can be enticing, but seem like a gateway into surprise billings. What do you do if you require a few OCI containers at the same time?
How do you folks do it? Is everyone just doing "serverless" these days and I'm old fashioned?
The ARM VM provided by Oracle is actually quite powerful (definitely much more so than anything free by GCP or AWS). Since I have no point of reference for that processor, I have no idea if you're getting four real cores, but it at least feels so. Compiling large C projects is faster than doing it on my machine (although the target architecture is different, so the comparison is a bit pointless).
The two x86 VMs are puny and can only be used as VPN gateways, or for static site hosting, or something like that.
Not using any of that newfangled "serverless" nonsense, and do not plan to. For work projects we rely on colocation with properly self-provisioned and fully controlled servers. It would be silly to use free tiers since you get absolutely zero uptime guarantees.
It seems, from a glancing review, all of these services structure the Free Tier to force an "on-demand" or "serverless" architecture, since the CPU-seconds and GB-seconds are always undersized for an always-on system (such as a traditional server or OCI container).
For hobby projects or book exercises, the Free Tiers can be enticing, but seem like a gateway into surprise billings. What do you do if you require a few OCI containers at the same time?
How do you folks do it? Is everyone just doing "serverless" these days and I'm old fashioned?