OVH datacenter fire

Appreciate your answer. However this:

has not been addressed (yet?).

1 Like

We are planning to move repair servers out of Hetzner into more regional datacenters. Even so, the repair servers are not a problem in the case of a datacenter outage. The repair servers are stateless, and do not need to be running at all times. If the data center running our repair servers has an unfortunate issue, we can spin up new repair servers elsewhere later with no durability penalty. We keep a healthy durability buffer to allow us to make changes like this.

13 Likes

Hah, maybe I’ll get some more ingress and repair egress

1 Like

May be it is time to moove some port of repear protsess and audit to nodes, or some spesial nodes, and will be just controled bystorj, and just do processing.

Hopefully we see it on our nodes…

3 Likes

From what I understand they used shipping containers. And the HVAC for the sites were not independent in terms of infrastructure.
So big no no’s.

Haha, no, actually, this is a French expression. When they say its “made of wood”, it means they made it “cheap” (in a poor quality) actually. Another material is used to mean the same thing when the French tell “its made of moss”. Like if you tell “I made a foam/wooden configuration for my node, I can’t garantee it will be reliable over the time.” :wink:

OVH is known to do “low cost” datacenters, and yes, it was one of the oldest infrastructures.

5 Likes

Just look what Filebase did:

3 Likes