Given today’s outage, their data centre fire and some other scary stories, are there any plans in the roadmap to move from OVH? Perhaps there is a particular technical reason for needing to use OVH rather than e.g Packet (Equinix Metal, as they now seem to be). It would be interesting to know a little more if possible about that.
Thanks!
Funnily enough we were literally about 2 days out from finishing a migration of our consul and related infra. We have new consul servers elsewhere and hadn’t quite got to vault and nomad (vault is now moved as well). Aside from some of these big, but pretty rare, issues, general network consistency has been getting somewhat worse on OVH as well so it was in our pipeline to move critical infra out of there.
Thanks Steve
@steveberryman can you also confirm that the future infra will have HA for your consul / vault and nomad servers?
Not saying that these incidents are happen often, but if there was a fallback server this wouldn’t have so long outage.
We actually run everything with HA, even across two separate datacenters.
Multi region Nomad / Vault / Consul is not quite a thing, though, which is part of why we’re working to replace Nomad entirely. It doesn’t make much sense to have centralized “servers” for the type of scheduling we need to do.