Reading between lines of host status updates
After three years of watching Hetzner and similar hosts, here is my translation layer for status posts.
"Scheduled maintenance" = kernel upgrade they deferred six months.
"Network optimizations" = someone misconfigured a switch and they are fixing it live.
"Improved stability" = they finally moved off OpenVZ 6 and paid the virtualization tax for proper KVM.
"Resource rebalancing" = a node hit cgroup limits and they are migrating the loudest complainers.
The worst: "We are investigating reports of intermittent connectivity." This means they do not know yet. Could be upstream, could be their router, could be a rat in the rack.
Watch for weasel words around SLAs. "Best effort" and "up to" are doing heavy lifting. If a post mentions "proactive measures" without specifics, check your backups.
I will update this as patterns emerge.