What's the correct way to put a new 4TB NTFS USB3.0 disk through its paces to ensure it's good?

Toggle sidebar Toggle sidebar
T

Timstertimster

Distinguished
Apr 3, 2013
93
0
18,640
After recently living through the nightmare of losing a brand new backup drive to hardware failure, im apprehensive (it literally failed completely the second time I plugged it in when I was planning to check data integrity after copying files).

I'd like to put the replacement drive I got through its paces, like a professional would.

I plan to format it with a 64kb cluster size as I assume it will help speed it up a wee bit.

How do I best ensure this drive is good, before I load it with 34,556,809 files worth 3.2TB ?

Thanks.
 
Solution
USAFRet
All drives die. Some within a couple of weeks, some after a couple of decades.
But they ALL die eventually.

My most recent death was a WD Green 3TB. Died at the 5 week point. Went from perfect to dead in about 36 hours.

A dead drive, within the warranty period, should never be more than a few emails back and forth for a replacement. Out of warranty, buy a new one.

Your data should never be affected. This is what backups are for.
It should not be a 'nightmare'.

How to test? This is what the SMART values are for. The drive reports errors.
But you can test all you want. Doesn't mean it won't die tomorrow.
Sort by date Sort by votes
A

Andy_K

Prominent
Jun 1, 2017
163
0
760
Having also lost a huge amount of data in the past, I've ended up going the raid route, 3 x drives that offer redundancy for if one dies.

Sure, it's not particularly cheap, but you don't have to do backups, and the ease of simply replacing the broken disk and letting it rebuild the array outweighs the possible loss.

Worth a look at.
 
Upvote 0 Downvote
T

Timstertimster

Distinguished
Apr 3, 2013
93
0
18,640
@13th monkey, you're assuming that all my files ar the same size... So, let me qualify a little more:

Tens of thousands of large files in the 50-400 MB range, a few hundred in the 0.5-2 GB, followed by millions of small files ranging from 0.4kb to 1.5kb

The small files are pretty much write once and forget. Likelihood of needing to retrieve them is extremely small.

The larger ones are archives that I do occasionally have to access in migration scenarios.

Should I just stick with 4096 or does a larger cluster actually make a difference?
 
Upvote 0 Downvote
T

Timstertimster

Distinguished
Apr 3, 2013
93
0
18,640
Based on these answers, it appears that it's always a lottery every time? I'm having trouble with that philosophy. There ought to be a way to test a drive to see if it's holding reasonably well for consumer use.

RAID is overkill, I've got no budget for it, and the drive is a backup. Over the past twenty years I've had no problems using a USB drive and collecting data when everyday drives failed.

Have we come to a place in computers where a consumer can have no expectation that a drive they purchase will last more than 2x plugging in?
 
Upvote 0 Downvote
USAFRet

USAFRet

Titan
Moderator
Mar 16, 2013
171,456
17,912
184,590
All drives die. Some within a couple of weeks, some after a couple of decades.
But they ALL die eventually.

My most recent death was a WD Green 3TB. Died at the 5 week point. Went from perfect to dead in about 36 hours.

A dead drive, within the warranty period, should never be more than a few emails back and forth for a replacement. Out of warranty, buy a new one.

Your data should never be affected. This is what backups are for.
It should not be a 'nightmare'.

How to test? This is what the SMART values are for. The drive reports errors.
But you can test all you want. Doesn't mean it won't die tomorrow.
 
Upvote 0 Downvote
Solution
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom