Hard drives

ShiftZZ

ShiftZZ

LIFE MEMBER
Joined
Feb 19, 2008
Posts
21,457
Likes collected
88,581
Location
Dark Side of the Moon
Funster No
1,546
MH
A class
Exp
Since 2007
I have a couple of NAS and over the last 12 months 3 Western Digital Drives have failed!

Are WD drives that unreliable? Seagate any better?
 
How many hours do you think they have run? I have had seagate fail, but don't think there is much in it. Check the ratings are okay for NAS drives before you buy!!
 
Could be all sorts of reasons for failure but that seems a high rate, especially if they aren't that old.

Most manufacturer's produce NAS versions of their drives. Theoretically they will be more resilient to the repeated long term use in a NAS, which could be getting abuse from a number of different appliances 24/7, rather than the average computer that is only doing one thing at a time for short periods.

But with that failure rate I'd be wondering about other causes such as a dodgy power source creating spikes, faults in the controller on the NAS housing, temperature, and similar.
 
A conventional hard drive will run for many years if left alone and not moved around when it is spinning. Laptop hard drives suffer if you do not put it on a table, they wear quickly if you balance it on your knees. The problem is the disc spin at around 5400rpm, if you tilt it the read head can physically crash onto the rotating disc. For mobile use, such as a motorhome, a solid state drive is better. No moving parts so much more tolerant of movement. They also start much faster and will shorten the boot up time. Known as SSD they are available in usable sizes now for a reasonable price.
 
We used to quote 5 years 'mean time between failure' for disks. Some lasted longer, some not so much.

Subscribers  do not see these advertisements

 
ShiftZZ said:
I have a couple of NAS and over the last 12 months 3 Western Digital Drives have failed!

Are WD drives that unreliable? Seagate any better?
Click to expand...
Do a Google search for NAS HDDs. Most manufacturers do drives specially for the job that cost a little more but are happy to run constantly.
 
are you using em for cctv
if its for continuous CCTV recording then they wont last long unless you get a hard drive that is designated for continuous over recording.
 
The last of the batch of 4 hard drives has failed , so buying 4 WD hard drives, nothing heavy , just storing photogra, music and documents, ALL have failed., WD said they are not offering further support, time to change manufacturer..
 
Take a look at synolgy NAS mine has lasted 3 years and still going strong, fingers crossed

Subscribers  do not see these advertisements

 
SSDs tend to fail without any warning. One minute fine, next dead. As much a likelihood its the controller chip as the actual storage, much the same as USB stick or photo flash card dies. I've lost a couple of the original (first generation) type OCZ units.

HDDs tend to gradually fail, unless they've been dropped or thumped somehow. You'll hear them getting noisier. Always worth periodic checks with the Smart readings for signs of too many bad sectors and stuff.

For large capacity storage of a NAS system, HDDs are still the cost efficient approach. Even a 2TB SSD is eyewatering, and rarely are you needing the SSD speed in a NAS.

Realistically, manufacturers offer warranties. and 5 years is now possible. Like any warranty, outside of that and all you have is a brick.

If buying I would use the lead manufacturers generally WD and Seagate for HDDs adding Samsung for SSDs.
 
I have 5 SSD in this computer 2x2 raids and a single backup memory. All Samsung.
Are SSD's not big enough to be cost effective? You can buy 1Tb SSD now for around £70.
 
ShiftZZ said:
I use them in a NAS 4 x TB the cosyt woud be massive.. I use RAID also
Click to expand...
3 times the cost. A lot more than 3 times the reliability. Especially for your usage patterns, which almost certainly involve very few writes and long uptimes.

Subscribers  do not see these advertisements

 
Out of 3 HD external failures doing the same duty as backup drives, 2 were WD and 1 was a Seagate, so IMO they're around equals.
 
When I worked for a computer component wholesaler (tech manager and purchasing manager) I saw a lot of failures to our returns dept.
There is no one manufacturer that stood out as being particularly bad. BUT they all went through phases of producing really bad batches apart from a few manufacturers.
I recall multiple seagate and western digital bad batches.

That said, I started following BackBlaze's reports many years ago and based on personal experience and their experience it is worth spending a little more and buying HGST.

Check their latest report out here.

2020 Hard Drive Reliability Report by Make and Model

The AFR for 2020 dropped below 1% down to 0.93%. In 2019, it stood at 1.89%. That’s over a 50% drop year over year.
www.backblaze.com www.backblaze.com

I am pretty much all in on HGST 4TB disks and haven't had a single failure. I replace them every 5 years to be safe though.

It is worth noting that raid in of itself is not a backup mechanism. Raid is great for bulk storage for your primary system. But for backup seperate drives backing up the raid array is needed and they ideally should be stored at a different location to avoid the risk of flood, fire or theft if your data is valuable to you.
 
Just checked and the model I use is no longer produced sadly.

The other one I have seen recommended is the Seagate Iron Wolf Pro as it has data recovery services within the price and an enterprise warranty.
If you have the drives in a raid system (next to each other) always buy the NAS version. Drives mounted next to each other have interactions due to vibrations which can shorten life. NAS drives have added features to protect against this;

Even shouting at your hard drives makes them underperform :Eeek: :rofl:

 
Not sure SSD are enterprise ready - but maybe they’re evolving tech.
One manufacturer I know put a “time to failure” type display in a product GUI for the SSD system drive.
Not had that definitive display for a HDD though they do have S.M.A.R.T. variables that can be tracked.

*- I remember now. Omnoen’s SystemManager. So that was >5yr ago.
 
kevenh said:
Not sure SSD are enterprise ready - but maybe they’re evolving tech.
One manufacturer I know put a “time to failure” type display in a product GUI for the SSD system drive.
Not had that definitive display for a HDD though they do have S.M.A.R.T. variables that can be tracked.

*- I remember now. Omnoen’s SystemManager. So that was >5yr ago.
Click to expand...

SSD's are getting there on price. The only issue with SSD is that you can choose lifespan or you can choose capacity. You give one up to gain the other.

Each cell of an SSD used to only store 1 bit of data. It was either on or off.
With MLC, TLC and QLC they store 2,3 or 4 bits of data. So each cell has multiple levels of voltage. As cells age it gets harder to store and recover the cell level so they wear out.

So you can buy an SLC nand chip and store a quarter of the data at very reliable and long lasting rates. Or you can buy a QLC nand chip and store 4 times the data but it will wear out sooner.

Spinning disk storage also has an issue at modern capacities. If you set up a raid array with 4TB disks for instance if one disk fails you swap it out and the raid array gets rebuilt over the next x hours.
If you set up a raid array with 8TB disk the time time to rebuild more than doubles. The rebuild time is a window during which another disk failure would cause a total failure (in some array configurations).
So the bigger the window the more chance there is of a second or third disk failing. With 20TB hard disks now available you are talking about a risk window in days not hours. In a busy array this could extend to weeks.

In the enterprise you used to carry one parity drive per array, this increased to 2 about a decade ago. There is now talk about 3 or 4 parity drives to allow for multiple failures during a rebuild. Imagine you swap out a 20TB disk and during the rebuild window another disk fails. You would be pooping your pants if you only had 2 parity drives. Also worth noting that during the rebuild process ALL other drives in the array get hammered due the read/write process required for rebuilding the failed disk.

Then there is the issue of bit rot. The more data on a drive the more chance of bit rot hitting you. So now with filesystems like ZFS you have parity and correction data stored for each block as well as at the file level.

Storage tech in the enterprise is getting ever more complicated and difficult. Datacentre space is getting harder to find and more expensive so fitting 60 drives into 4U rack server is becoming more attractive from a cost point of view, but then you have noise issues, cooling issues and the fear of drive failures causing a cascade failure.

This is a subject I could tell some stories on :eek:

Subscribers  do not see these advertisements

 
I've had a WD 2TB My Cloud drive running for at least 6 years without a problem. It holds all my books, films, photos, music and of course files. Can't get anywhere near that on my laptop, either with a 256 gb SSD or 500 gb HDD. However, it won't last for ever will it? I've been lucky for six years, lets not push it.
So rather than buy a couple of new ones, I've opted for a subscription to Family edition of MS OneDrive. That's 6 Tb for £79 pa including MS Office 360* vs a 6 TB My Cloud from Currys at £230, ... so over 3 years its a similar expense. For the peace of mind of not having all the eggs on one drive which could fail and the hassle of transfers, its all in the cloud now. Plus I have access to all my files on ten devices incl pcs, tablets, laptops and mobile phones. (Mine and Hers and shared areas, ) I reckon the futures in the Cloud anyhow.

* Needed to update MS Office anyhow as Student edition Office 2003 was a bit long in the tooth,
 
I have 8 thecus 8900 fully populated with WD black disks cannot remember the last fail, they each have their own UPS 2200VA I keep this in a cool dark room
 
After talking about 2 disks for parity and failure rates. This came up on my youtube subscriptions. Not had time for the last 2-3 days to have a look through so the timing was propitious...
 
Just watched (well... skimmed) the first five mins the above-mentioned youtube video while eating my cornflakes. Needless to say this is one path i wont be going down. Life is too short for both. Its either mending hard drive arrays in the garage or going out in the MH. I know where im going today while all my stuff is in the cloud, (but not Google's cloud. )
 
Kannon Fodda said:
SSDs tend to fail without any warning. One minute fine, next dead. As much a likelihood its the controller chip as the actual storage, much the same as USB stick or photo flash card dies. I've lost a couple of the original (first generation) type OCZ units.

HDDs tend to gradually fail, unless they've been dropped or thumped somehow. You'll hear them getting noisier. Always worth periodic checks with the Smart readings for signs of too many bad sectors and stuff.

For large capacity storage of a NAS system, HDDs are still the cost efficient approach. Even a 2TB SSD is eyewatering, and rarely are you needing the SSD speed in a NAS.

Realistically, manufacturers offer warranties. and 5 years is now possible. Like any warranty, outside of that and all you have is a brick.

If buying I would use the lead manufacturers generally WD and Seagate for HDDs adding Samsung for SSDs.
Click to expand...
I agree with all your points and this is my experience too

I would recommend Samsung for all types of drives. Get a traditional HDD though if your data is valuable. SSDs are getting better but as you say when they go your data is very difficult to retrieve whereas with an old fashion magnetic disc, its still there somewhere.

Subscribers  do not see these advertisements

 

Join us or log in to post a reply.

To join in you must be a member of MotorhomeFun

Join MotorhomeFun

Join us, it quick and easy!

Join us

Log in

Already a member? Log in here.

Latest journal entries

Back
Top