My experience with ~100 drives is that about 5% failure rate at 3 years, and 2% more at 5 years. Mostly Seagate.
Hard Drive Failure Rate Study
- Thread starter fortkentdad
- Start date
for any good tech that plans for hard drive failures in advance, Seagate has an EXCELLENT RMA process. Coincidence? heh
Exactly but the writes these drives see in 1 month would equate to what 3, 4, 5, 10, 15 years of normal usage?
I'm trying to understand the justification of "consumer drives are so inexpensive, you are better off buying them and replacing more often".
From a system administrator perspective, there's so much more to the equation than the initial drive cost. For example, there are man hours lost ordering a new drive, accessing and replacing the existing drive and checking that no data has been lost during the failure. There is also the recycling and environmental cost that comes with treating a computer component (even one that's RoHS compliant) as disposable. Plus, RAID arrays aren't infallible and the loss of ANY data could have incalculable financial repercussions.
From a home user perspective, the situation is magnified. Most folks don't have the money to buy multiple drives and if they do, they're typically placed in a simple spanned or striped array without any redundancy. While performance and capacity are increased, the risk of data loss increases exponentially. For everyone I know, loosing their personal HDD's info is catastrophic. Considering how much people will spend to get their data BACK after corruption, drives with better longevity should be a no brainer.
Hence, I never, ever run the lowest common denominator in my systems. Ever.
From a system administrator perspective, there's so much more to the equation than the initial drive cost. For example, there are man hours lost ordering a new drive, accessing and replacing the existing drive and checking that no data has been lost during the failure. There is also the recycling and environmental cost that comes with treating a computer component (even one that's RoHS compliant) as disposable. Plus, RAID arrays aren't infallible and the loss of ANY data could have incalculable financial repercussions.
From a home user perspective, the situation is magnified. Most folks don't have the money to buy multiple drives and if they do, they're typically placed in a simple spanned or striped array without any redundancy. While performance and capacity are increased, the risk of data loss increases exponentially. For everyone I know, loosing their personal HDD's info is catastrophic. Considering how much people will spend to get their data BACK after corruption, drives with better longevity should be a no brainer.
Hence, I never, ever run the lowest common denominator in my systems. Ever.
Because they simply don't realize that they could buy a good drive for a bit more money and have it last longer than buying a cheaper drive and replacing it 3 or 4 times for double or triple the cost.
All those costs are easy to add up when you're working on a large scale, and adding up all the costs for consumer drives is cheaper than using enterprise drives. (Other than environmental costs, which they're not paying - but if you want to deal with those, the best way is probably to mandate all companies to provide an X year warranty on their products, so it will become too expensive to make high-failure products.)
For home use, if people can't afford multiple drives they should subscribe to a cloud backup service (like backblaze).
For either enterprise or consumer use, any drive has a 100% chance of failure, so the way you plan for redundancies doesn't really change, you have to plan for it to fail regardless of expected lifetime.
If you read the article I linked, you'll see that doesn't happen, enterprise and consumer drives have very similar annualized failure rates.
...because they walk into a Futureshop or Best Buy, Walmart, etc. and buy what the 16 year old tells them to.
That is, if they even bother to go looking for a backup solution.
Most people I know only buy a backup solution AFTER they have had a loss of data. they rectify the situation buy buying the cheapest solution at a big-box store that an inexperienced salesperson tells em is popular.
Its scale. Because some of these cloud storage places already have redundancy across many systems. The systems themselves can be as cheap as possible. Google started the whole cheap server movement vs fewer higher grade systems. What they have found is the difference between enterprise and consumer drives are not worth it. The consumer drives perform and last better per $ then enterprise drives. This only really works in large scale where there is no downtime because of multiple server redundancies and high availability. If your company has less then a 1/2 dozen servers I don't see it scaling down that low and you would want to buy the better hardware because you probably are not running any sort of HA system.
Iv had 1Tb wb black hdd go bad in a nas in under a month. And it was being used maybe 5-6 hours a week.
xentr_legal_notice_title
xentr_legal_notice_description