HAMR HDD capacities to scale from 4GB in 2016 to 100TB in 2025

[crossy]

having 100TB discs is all well and good, but you need some sort of protection for power interruptions when a lighting storm passes by at the exact moment you are doing your monthly backups to external drives. [snipped...]

You've either been very unlucky (in which case I can empathise - really lousy day tech-wise yesterday) or you're doing backups wrong. Thought the whole purpose of backups was that the source was effectively read-only, in which case I'm a wee bitty confused how it would get corrupted too.
You've kind of got a point though - UPS's are getting cheaper and more readily available - heck, my local Tesco's sells basic APC ones.

Maybe, by 2025, Microsoft will have got around to providing a file system with proper data integrity and recovery built into it.

I thought NTFS had a reasonable reputation in that area? Personally, I'd prefer it if they just stopped messing around and adopted ext4, although I've got a soft spot for JFS too.

So Seagate is working on larger than 3.5 inch HDD's to get past 10TB while Samsung has 16TB 2.5 inch SSD's sampling now. Seagate should focus more on it's SSD's the HDD's time is numbered.

Sigh Seagate doesn't have any real NAND tech, which is why it's folks like Toshiba, Micron/Crucial, Samsung and Sandisk who are making all the SSD headlines. On the other hand they do know (apologies shaithis!) about HDD's. So it makes sense to go down those roads you know, as the old cliché says.
And SSD's are - if you believe the pundits - only a halfway house themselves. We're all going to be using some kind of atom scale or holographic storage in 25 years. (apparently)

No what I'm interested in is if HAMR tech is going to make our current "large drives" a lot cheaper to get. At the moment a 4TB desktop drive is about £120 (Seagate or WD-Red), but if that's the bottom rung of HAMR tech then maybe (here's hoping) that the price of those drives will fall to the levels of the 1TB or 2TB drives, (and those smaller drives will disappear).

[mikerr]

having 100TB discs is all well and good, but you need some sort of protection for power interruptions when a lighting storm passes by at the exact moment you are doing your monthly backups to external drives.
So as a consequence of that power interruption at the moment your computer is displaying the message 'please dont turn off your computer'. your computer loses power and all the files that are open in memory and the ones you are writing to your backup drive that are still in the write buffer all get heavily corrupted and you therefore lose everything on the windows drive and everything on the drive you were backing up to.

That's irrelevent to drive size, and just down to backup regime.

Tao3: Separation http://www.taobackup.com/

[shaithis]

There, fixed that for you. (BTW, I've had most problems with WD, and least with Seagate - but ymmv)

I used to build a lot of NASes for customers....used to go through hundreds of HDDs per year. The failure rate on seagates (since around the 2TB size) is unbelievable.

[stevie lee]

You've either been very unlucky (in which case I can empathise - really lousy day tech-wise yesterday) or you're doing backups wrong. Thought the whole purpose of backups was that the source was effectively read-only, in which case I'm a wee bitty confused how it would get corrupted too.
You've kind of got a point though - UPS's are getting cheaper and more readily available - heck, my local Tesco's sells basic APC ones.

was just doing a clone of C drive, to save reinstalling everything. going to an external USB3 2.5" drive. usb drive got corrupted because its usb and hadn't finished writing everything. no idea how the source drive got corrupted, could be because its the windows drive and it was doing other stuff in background, who knows, its certainly a mystery to me.
important other files go to USB drive, another computer, online and dvd. hadn't done backup of those yet this month


I suppose most people who have need of large scale storage will have actually thought about UPS, backup regimes and other assorted stuff beforehand. only use I have is for backing up steam/origin/gog to save downloading the games again.

[crossy]

I used to build a lot of NASes for customers....used to go through hundreds of HDDs per year. The failure rate on seagates (since around the 2TB size) is unbelievable.

Interesting, thanks for sharing that info. I'm always being told (by "experts") that I "can't go wrong with WD", yet this year alone I've had one fail completely and a second one perform like a dog and get binned.

Joking aside, the point I was trying to make was that all storage devices are eminently faliable.

was just doing a clone of C drive, to save reinstalling everything. going to an external USB3 2.5" drive. usb drive got corrupted because its usb and hadn't finished writing everything. no idea how the source drive got corrupted, could be because its the windows drive and it was doing other stuff in background, who knows, its certainly a mystery to me.

Wow, very unlucky! As you said in your first post in this thread, if you get a lightning zap then it's going to be a lottery what damage gets done. That's one reason why I make sure that there's a surge/spike suppressor on all my data-sensitive kit. But as you said, a UPS would be a far better (although more expensive) option.

[DanceswithUnix]

It's Seagate. All I see is a way to lose more data, faster.

Many many years ago, probably on the CiX bulletin board or Usenet, someone said "I would rather write my data onto burning paper". Some things never change

I thought NTFS had a reasonable reputation in that area? Personally, I'd prefer it if they just stopped messing around and adopted ext4, although I've got a soft spot for JFS too.

NTFS is a basic journaling filesystem, all the rage 20 years ago. The world has moved on, with the liked of ZFS and btrfs.

Apparently Facebook are moving their storage to btrfs, so even that seems to have finally come of age.

[Marcvs]

Sounds cool but still has the problem of being Seagate...

[Saracen]

....

is blanker than our current chancellor.

....

Which, even if true, is better than having an egomaniac in power for years labouring (or Labouring) under the misapprehension that he was God's gift, when really, he made a complete Horlick's of about 90% of it.

There's is a point of view that says that, with limited exceptions, the best thing politicians can do is stay the bleep out of the way of the economy, be they ideologically left, or ideologically right.

My biggest concern with our current Chancellor is that he isn't a complete blank. If he were, it might be a refreshing change.

But, that really is going off-topic. I just couldn't resist.

[Saracen]

Many many years ago, probably on the CiX bulletin board or Usenet, someone said "I would rather write my data onto burning paper. Some things never change.

That (CIX) takes me back.

Interestingly (or not) since the days of my first HD (20MB, IIRC, Seagate), and my days (years) on CIX, the ONLY Seagate drives I've had fail were three out of four 40GB drives on an Adaptec hardware RAID board, and that was due to a PSU failure. The board survived, though, which is more that I can say for a high-end pre-production graphics board in that machine. And about half of the entire system. And by failure, I mean it physically blew chunks out of chips on the HD IF board, and left my office smelling of that acrid electridal burning stench for about a month. I can't really lay even those failures on Seagate.

Admittedly, most (all bar two) are/were below 2TB in size.

[bledd]

RIP HDD, long live SSD

[DanceswithUnix]

That (CIX) takes me back.

Interestingly (or not) since the days of my first HD (20MB, IIRC, Seagate), and my days (years) on CIX, the ONLY Seagate drives I've had fail were three out of four 40GB drives on an Adaptec hardware RAID board, and that was due to a PSU failure. The board survived, though, which is more that I can say for a high-end pre-production graphics board in that machine. And about half of the entire system. And by failure, I mean it physically blew chunks out of chips on the HD IF board, and left my office smelling of that acrid electridal burning stench for about a month. I can't really lay even those failures on Seagate.

Admittedly, most (all bar two) are/were below 2TB in size.

My first drive was a 20MB Seagate, a SCSI drive connected to my Atari STE. It went back 3 times under warranty for replacement IIRC, as some days it just wouldn't spin up.

Then one day I bought a Quantum SCSI hard drive, I think it was an entire 40MB, and my expectations changed