Hacker News Reader

Comment by danwee | Hacker News Reader

> Then you contact support, appoint disk change, you first deactivate disk on raid (save geometry etc), they replace disk and then you rebuild raid in new disk. That's it.

I imagine this would take time, right? Like not 5 minutes, but maybe 3 hours top? So, if I pretend to run a saas (that shouldn't be down more than 1h/day), then renting only 1 dedicated server could be qualified as "risky"?

View on HN

RAID is not a replacement for backups anyway. There are many other ways to lose data besides physical disk failures. There are also ways to lose the RAID in face of physical disk failure. It's an availability solution (a box can probably keep running without an outage without resorting to backups).

You should be running at least three, preferrably four (3 + hot spare) mirrored instances if you require that level of uptime, regardless of provider or tech (bare metal/VM).

Well, yes, having a single server is risky in any production context if it's storing any sort of state that can't be easily brought back up on another server

I'm not sure what "deactivating RAID" means.

They will all be hot-swap disks. You remove the old disk and slide in the new one (or in this case, tell them to do it). The RAID system rebuilds the array in the background over the next few hours.

During that time you will lose data if it's RAID 5 and another disk fails.

> I'm not sure what "deactivating RAID" means.

mdadm --manage <array> --remove <failed disk>

so your machine doesn't have a fit when the disk is detached. Or equivalent.

Presumably he means detaching the disk from your RAID solution so it doesn't freak out when it's physically removed and replaced.

I am using software raid servers, as other commenters say is not a hot swapping disks operation, is cold swap :D

Rebuilding raid doesn't take the server down but disk performance i.e. IPOS will be reduced while it takes place Running service on 1 server always carries some risk whatever form that server takes

For example I have loads of stuff on Linode but always make sure I keep backups off-linode, incase I get a random TOS account shutdown and they stop speaking to me etc

Absolutely, some redundancy is necessary if uptime is critical (as well as backups).

Is risky, 3 hours top is a good estimation. You can appoint for an out of business hours change. But better choose another strategy start cloning disk at first failure notice and cancelling old server.

Reply on news.ycombinator.com