alt Hacker News>That CPU is ancient, though. Over a decade old.
Coincidentally so are the t3 / t3a instances on AWS that everyone loves to use especially for dev/staging environments
>That CPU is ancient, though. Over a decade old.
Coincidentally so are the t3 / t3a instances on AWS that everyone loves to use especially for dev/staging environments
10 years is where hardware failure rates start ramping quickly, in my experience.
Not necessarily obvious failures, but subtle errors, memory problems (like this case without an ECC capable CPU) and little instabilities.
With cloud instances I can migrate to a new instance with a couple clicks if I want.
Trying to save a couple hundred euros per month on hosting costs needs to be balanced against the risks and extra developer time.
For personal projects these old instances can be an excellent deal though.