HDD reliability

pgh13

Specialist Contributor
My Satellite Setup
XTrend ET8000, Dr HD 15, PC/twinhan combination. 80cm dish. GBPVR with 2 DTTV cards +hauppauge Media MVP +Raspberry Pi running XBMC
My Location
Midlands UK
I've built quite a few PCs over the last 20 years and had relatively few problems with HDDs. I've had several instances lately where HDDs that I've tried to access after periods of disuse have come up as not containing a recognised file system. In some cases I've managed to recover data from them using Testdisk.
They've all been SATA disks (3.5 and 2.5) from different manufacturers and used with a number of Windoze operating systems. Obviously over the years the data density has increased.... has this made them more 'fragile?'
Any thoughts?
 
My personal opinion is yes but from what I read, the reality isn't necessarily the case. I have a couple of 2 and 3Tb drives and they've been ok. With 6Tb drives out there now, I don't know if they are as reliable but time will tell.

Bear in mind that spinning disk technology is very old and a lot of work and progress has been made into making them more and more reliable. Think back to older SSD drive days - they used to fail quite frequently and without warning, with no chance of recovering your data (from what I know).

In the future, I'll buy some sort of NAS box or build one myself to create a RAID5 array. That way I would be protected against single drive failures.

Would be interesting to hear other people's experiences.
 
Did you try reading the discs with a Linux OS? I have seen Linux read discs, and rescue the data, that Windows refuses to recognise on a few occasions.
 
Did you try reading the discs with a Linux OS? I have seen Linux read discs, and rescue the data, that Windows refuses to recognise on a few occasions.
Good thought.I'll give that a try
 
A month ago, a strange fault occured on my main PC with 1 TB harddisk (WD10EADS-22M2B0, green version).
It was not crashed or damaged, all files were still accessible, but with a very slow read/write (a factor of 100 or so).
Starting up the PC took about an hour! And without any further error messages.

Solved meanwhile with a new harddisk (WD again...)
 
It could be the SATA cables themselves - dust or oxidation build up giving a poor connection.

I had a set of unbranded cables on my old computer that gave various errors. The hard disks would sometimes not show at the BIOS screen. Or they would, but the names/model number would be partly corrupted. Or everything would look fine at the BIOS, but Linux would flag up I/O errors when reading/writing files (not sure what the Windows equiv messages are). Or the system would freeze.

Blowing into the SATA connectors out with compressed air and plugging in 5 to 10 times was a temporary fix. I later replaced the cables which helped... but if the computer hasn't been on for a few months it sometimes starts throwing up the same errors.

The old cables never seemed a good fit. They'd wobble and move around at the motherboard end. I guess the tolerances of the SATA plugs and/or the mobo sockets was too loose...

My current computer with a Gigabyte mobo came with a bunch of cables which fitted far more snugly. I've never had any problems with it.
 
Thanks for the thoughts intracube. I recovered some data to my laptop using a SATA to eSATA cable. Very slow as well.
So probaly the cable is not the problem in my case.
 
I'm sure I've read that SATA plugs and sockets are only rated for a very limited number of connection and disconnections, perhaps as low as 10 or 20 times. Perhaps they have been used a bit too much?
 
I'm sure I've read that SATA plugs and sockets are only rated for a very limited number of connection and disconnections, perhaps as low as 10 or 20 times. Perhaps they have been used a bit too much?
That's interesting.

I only started unplugging the cables after they caused problems, and once or twice didn't seem enough. Oh well o_O

Will avoid doing that in future.
 
Interesting reading.
 
Seagate Barracuda 1.5Tb failure rate nearly 25%!!!! That's the drive I have in my PC.... aahhh... better back up the data, eh..
 
Western Digital 4Tb, looking ok for me! :)

A bit worrying for some though, gotta be a right pita when it all goes "tits up" :(
 
After reading this, I've decided to upgrade my current drives. My 3Tb Seagate drive is the one that has a failure rate of 43%! Aaaaaggghhhh
 
Getting a NAS-drive is one way of getting diagnostics and early warnings if the data
is of great value or irreplaceable.

I got a ReadyNAS NV+ a few years ago when they were on sale.
(actually I got two, I'm a coward, backup of the backup)
Each has three (place for four drives) WDC WD20EARS-00S8B1 which are not
considered to be best but the price was nice.

The SMART diagnostics in the NAS gives the following listing after clicking the
[Health]-button for each disc.
------------------
SMART Information for Disk 1
Model: WDC WD20EARS-00S8B1
Serial: WD-WCAVY3810542
Firmware: 80.00A80
SMART Attribute

Raw Read Error Rate 0
Spin Up Time 9216
Start Stop Count 1067
Reallocated Sector Count 1
Seek Error Rate 0
Power On Hours 10060
Spin Retry Count 0
Calibration Retry Count 0
Power Cycle Count 1065
Power-Off Retract Count 1064
Load Cycle Count 7
Temperature Celsius 39
Reallocated Event Count 1
Current Pending Sector 0
Offline Uncorrectable 0
UDMA CRC Error Count 0
Multi Zone Error Rate 0

ATA Error Count 18
Extended Attribute

Hot-add events 0
Hot-remove events 0
Lp stat events 8
Power glitches 0
Hard disk resets 0
Retries 0
Repaired sectors 0
------------------

I don't understand much of the stuff in the listing, I assume it will give me
a warning and the green "health" light will turn yellow (or red) and I will get replacement discs.
If software rot gets into my computers it's a nuisance but no catastrophy.

The bottom line - if SMART-diagnostics are available you know the state of the HDD's.
 
Welp, I just went ahead and blown a lot of money on a Synology DS414 4-bay NAS box along with 3x 3TB Western Digital Red drives. Look forward to it :D
 
Heard from inside of excavator:

- Whats that flashing in the hole?
- Probably the power line to the Ericsson Communications building.
It just went black and people are running in our direction.
- OK, if the next building to go black is Emerson Process I know exactly where
we are on this infernal map. Keep digging!

... that's why I got my NAS-storage.
 
I'm looking into nas too but Synology and QNAP is just ridiculously priced.
Anybody got experience with Buffalo? I would like raid5 but have no need for recoding video in real time.


Have a look on the Edugeek forums and search for Nasdrive. These are school and college IT guys who have seen most problems with IT kit over the years! Theres some interesting chat about NAS's on there (written in easy to understand language!)

Ive had a Netgear Duo NAS for a few years - simple mirror raid - and its one of the best things Ive ever bought. Only one problem with it over the years - a drive failed but that was after a dirty shutdown caused by a power cut. A drive swap out rebuilt the array perfectly though.

Apart from that it just keeps whirring away quite happily
 
_hard_drive+platter_table.webp
Western Digital 4Tb, looking ok for me! :)

A bit worrying for some though, gotta be a right pita when it all goes "tits up" :(


I heard (from a colleague) that Western Digital were the last company to produce platter hard drives, anything else on the market a few years ago was the same but out of another door of the production plant. I gave up on them after they brought out Passport portable drives - utter crap, moreso when their own encryption system - for no reason than the algorithm crashing - forgets the password you put in to protect your work/emails/tax records/livelyhood.

The small print - later - did say 'put your important data elsewhere' which sort of defeats the object of taking what you paid over the odds for in the first place.

Anyway I recall some horror stories regarding Quantum (and the Fireball)IDE drives in the early 00's. Total failure with no chance of recovery.
 
Back
Top