3 ms·

This is a hilariously small deployment. Anybody that assumes bare metal is easy at scale has not had to build out and cool data centers. Heck even designing and

This is a hilariously small deployment. Anybody that assumes bare metal is easy at scale has not had to build out and cool data centers. Heck even designing and keeping the hvac running properly requires expertise. Same with powering all this stuff. One electrical fried rat in the right place can really ruin your weekend. I still believe running your own servers is cheaper but there is a tun of logistics you don’t want to really deal with.

hanikesn 2y ago
Did you ever hear of colocation, root servers and VPSs?
jgb1984 2y ago
It's weird how the same straw man argument comes back when this topic is discussed. As if the only two options in the world are cloud and on-premises bare metal. While the best solution is often to simply rent a bare metal server from any of the big hosting providers like Hetzner. They take care of cooling, physical security, hardware failures, network infrastructure etc. All you have to do is manage the server. Which isn't more or less work than managing the horror that is AWS (just 1000x more enjoyable in my opinion) while getting 10 times more bang for your buck.
dijit 2y ago
I ran 1% of Web Traffic in 2011 without running my own datacenter. Our "platform" team was 6 people, two network engineers and four "devops" (Sysadmins that could write code, but were still called SysAdmins then). We had two datacenters, one was SAVVIS in Winnersh Triangle, the other was in Santa-Clara USA. We used remote hands- rarely- in the event that a cable died or a harddisk needed to be replaced, though such events are exceptionally rare even at the scale we had. Mostly we would collect up failures and visit one day every 3 months.
orev 2y ago
The vast majority of applications are “hilariously small deployments” and only use the cloud or k8s because the engineer wants that on their resume. And there’s a huge range of sizes between “hilariously small” and “needs a whole data center”. This conversation always seems to assume everything is and needs to be Google-scale.

hanikesn 2y ago

Did you ever hear of colocation, root servers and VPSs?

jgb1984 2y ago

It's weird how the same straw man argument comes back when this topic is discussed. As if the only two options in the world are cloud and on-premises bare metal. While the best solution is often to simply rent a bare metal server from any of the big hosting providers like Hetzner. They take care of cooling, physical security, hardware failures, network infrastructure etc. All you have to do is manage the server. Which isn't more or less work than managing the horror that is AWS (just 1000x more enjoyable in my opinion) while getting 10 times more bang for your buck.

dijit 2y ago

I ran 1% of Web Traffic in 2011 without running my own datacenter. Our "platform" team was 6 people, two network engineers and four "devops" (Sysadmins that could write code, but were still called SysAdmins then). We had two datacenters, one was SAVVIS in Winnersh Triangle, the other was in Santa-Clara USA. We used remote hands- rarely- in the event that a cable died or a harddisk needed to be replaced, though such events are exceptionally rare even at the scale we had. Mostly we would collect up failures and visit one day every 3 months.

orev 2y ago

The vast majority of applications are “hilariously small deployments” and only use the cloud or k8s because the engineer wants that on their resume. And there’s a huge range of sizes between “hilariously small” and “needs a whole data center”. This conversation always seems to assume everything is and needs to be Google-scale.