New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Don't remember to be honest, I would have to check, but not sure in which garage I left it. Its a part of 2000's workstation pc.
A trip down memory lane... Friends and I once stumbled upon a company's dumpster full of old PC's. They were like the accounting job kind, not anything fun like graphics cards and big storage disks, but our eyes bulged at the trove of potential server parts. While many parts broke after being victims of aggressively thrown head-first into the metal dumpster by (I can only guess) people that hated PC's, we scrounged enough working dinosaur bits and threw them in two old tower chassis that we already had, which ran a production website in my home for 3 years. One still ran mostly fine for months after a drive blasted errors at the console. This was when datacenter rentals and co-lo were hundreds per month. Backups, downtime mitigation, all things we learned while hosting and saving money.
These old Kimsufi's may not be as satisfying & fun as dumpster-diving, but at least the datacenter will replace erroring drives for free and cost about the same as the electricity to power it from your home. Affordable VPSs are the same dollar or euro with similar CPU performance, better disk speed, and faster bandwidth with greater reliability, but if you want the dedicated resources and big disk storage, few deals rival the Kimsufi's with the old disks. It feels frustrating to use hardware so old/slow, but think of the savings.
If it were me, I would stress-test the drives. If they pass, you're probably good for a while. If they produce errors, ask OVH to swap with another disk... then if you're near OVH, go find the old drive in their dumpster 🙂 or rest assured you're old disk will go to the next tech-savvy dumpster-diving kid.
Oh, and encrypt your private data because when drives start to fail, they're difficult or impossible to wipe.
It's in their blog.
Yes, some drives are like that. You start to hit only those Spartan like HDDs after like 4-5 years, the failure rates just don't climb, or might even start to get lower.
not sure if bias since you will naturally have less of the oldest drives, or if that really happens. I think we should start tracking at the same level as backblaze does. Kinda wish we did that from get-go a decade ago. We only have partial at that level, and typical is sticker on the drive for the received date, and from where
Do it now.
got to pick "battles" carefully with limited resources
All dev resources right now are tied into the new datacenter + new dedi platform right now.
But even for dedi platform we should figure out some sort of tracking, but that's nigh impossible without resorting to OVH style "backdoor" (they actually call it a backdoor).
oops. delete this please.
I have 1 kimsufi more than this running time but running fine from last 2 years
What a historical hard drive its is ! It is so old that should be put into museum...
But now kimsurf dig from earth and put it in your server again,oh
Since webeast necro'd this thread:
I even have a ssd in a server w/ 120k+ power_on_hours that is still working. When I saw that number I first thought: did they built ssds at that time already (about ~ 2010)?
If it is a single drive, I'd worry. At home I rarely let single drives go longer than 30Khours if they contain anything that isn't entirely ephemeral.
If you are covered by RAID & SMART and have checksums noted for important files (and some routine for verifying said checksums), or if you just have nothing on there that isn't easy to reproduce from elsewhere (off-server backups perhaps), all good.
We regularly retire drives that have more than 100k hours on them and not a single SMART error or SMART attribute out of spec. Likely still lots of life left, assuming it is performing as expected and doesn't have any SMART indicators that are out of spec.
Unless it's a Seagate. If it is, they are famous for dying suddenly and spectacularly.
My disks are
Disk 0 HDD 70,574
Disk 1 HDD 64,681
Post all the SMART attributes and mention full HDD model.
While replacing a hard drive after about 5 years isn't the worst idea, this is quite correct as well:
And this (especially if you see DM001 or DM002 / DM003 in the model):
They are upselling you to get a backup storage.
its okay guys, i have a kissmysufi with 90k hrs
kissmysuffix = kiss my ass?
asking for @ehab
we all need love and fluffies