←back to thread

Use One Big Server (2022)

(specbranch.com)
343 points antov825 | 4 comments | 31 Aug 25 17:29 UTC | HN request time: 0.815s | source
Show context
garganzol ◴[31 Aug 25 19:13 UTC] No.45086114[source]▶
And then boom, all your services are gone due to a pesky capacitor on the motherboard. Also good luck trying to change even one software component of that monolith without disrupting and jeopardizing the whole operation.

While it is a useful advice to some people in certain conditions, it should be taken with a grain of salt.

replies(1): >>45086204 #
fragmede ◴[31 Aug 25 19:21 UTC] No.45086204[source]▶
That capacitor thing hasn't been true since the 90's.
replies(2): >>45086318 #>>45086354 #
1. garganzol ◴[31 Aug 25 19:38 UTC] No.45086354[source]▶
Hardware still fails. It isn't a question of "if", it's a question of "when". Nothing lasts forever, the naivety lasts only so long too.
replies(1): >>45090129 #
2. fragmede ◴[01 Sep 25 06:45 UTC] No.45090129[source]▶
Obviously. But you get duplicate hardware, set up HA, get vendor support contracts, use multiple colors in disparit location. Cloud providers have figured this out fairly well, as we all did in the aughts. (Well, some of us anyway.) You can definitely determine that's a bunch of really annoying work and just pay a cloud provider to deal with it, or not, and go your own way. But if you want to be credible when saying that hardware fails, maybe people shouldn't use a problem from three decades ago as their example and use anything more recent?
replies(1): >>45091600 #
3. garganzol ◴[01 Sep 25 10:55 UTC] No.45091600[source]▶
When you get duplicate hardware, it is not "One Big Server" anymore. "Two Big Servers" at least. In September 2015, the failure rate caused by capacitors was still around 30% [1].

[1] https://www.researchgate.net/figure/Failure-rates-of-differe...

replies(1): >>45092804 #
4. fragmede ◴[01 Sep 25 14:09 UTC] No.45092804{3}[source]▶
Respectfully, 2015 was still a decade ago.