Is WD really that much better than Seagate?

[dfsdfgfkjsefoiqzemnd]

I've seen my fair share of dead 3TB Seagates too.  I'll never use a Seagate in my PC or NAS ... but I did buy a pair of 5TB Seagates for backup purposes and will be buying another pair once these are nearly full.

[comicsansms]

My PC bought in 2004 had a maxtor HDD (which is still working to this day) It has a bunch of bad sectors now but damn, the 80GB thing managed to boot up Windows XP on a fresh install in less than 10 seconds - on a Celeron 2.6GHz (single core) w/1GB RAM. It's a shame Maxtor HDDs don't exist anymore, that thing was reliable AF.

Good to hear.

[Captain_WD]

~snip~

 

Hey there FoxxyRin,

 

HDDs in general are mechanical units and as such failure do happen for one reason or another. Sometimes this happens well before the warranty period expires. This is why it's recommended to have backups of your important data on multiple storage drives, regardless of the brand. 

WD does make good and reliable drives and if they are used according to their design, they should serve you long and well. 

SSHD drives as a concept involve a bit more risk compared to regular HDDs since they have two points of failure compared to one in HDDs. In case the SSD portion of the drive fails, the whole drive becomes inoperable and having in mind that it's significantly larger compared to the small regular cache in HDD, I'd say better go with two separate drives (SSD+HDD) if you have the space in your PC. 

WD Blue SSHD is a good option to check out, if you are going for a hybrid solution: http://products.wdc.com/support/kb.ashx?id=THK86C

I could also suggest checking out WD Black2 which is a dual drive and works a bit different compared to SSHDs - it has two separate drives (120GB SSD and 1TB HDD) that work independently all in the same 2.5" form factor: http://products.wdc.com/support/kb.ashx?id=P5YMhD

 

It really depends on what you need the drives for. You may be better off with either a SSHD, or a dual-drive, pure SSD or even just a regular HDD. 

 

Feel free to ask if you happen to have questions

 

Captain_WD.

[ictdude1]

a wd blue cost about the same as a seagate barracuda, get what you like they will both fail at some point, just remember to always keep a backup

[Blake]

Seagates are fine and WDs are fine.

I wouldn't get a Toshiba, Samsung, Hitachi, Maxtor or Quantum, just because I have never experienced them.

I'd buy Samsung all day every day. Seriously, their Spinpoint drives, better read/write then WD Blacks, very reliable, cheap as WD Green. Unfortunately samsung sold off this department to Seagate. It was all down hill from there.

 

Seagate should have just re-branded the samsung drives... unfortunately they did not do this.

[Kevguy]

It has been acknowledged that there were some problems with the 3 TB Seagate drives. From what I've seen and heard lately, they seem to have sorted the issue(s) out.

I run 2 x 1 TB Seagate barracudas in raid1 mirror on my system and there has been no issues for the past 1.5 years.

[dalekphalm]

Already stated by Backblaze, but then again, people who would rather jump up and complain won't do any reading. How exactly does one discredit a 24/7 server environment that no typical user will be near doing? No matter what your bias, you can get a look at how different drives hold up in such an environment.

@sobe the problem is that their methodology is incredibly flawed, and that they lack a lot of crucial information. Blackblaze themselves probably have that extra info, but they have yet to publish it.

 

Take for example, they list average age of the HDD, yes, but they do not list the age of each HDD as it dies. Nor do they list the average age of a dead HDD. That can skew the data into very misleading numbers.

 

It doesn't even necessarily tell you how reliable each drive model is on average, since without those crucial missing data points, we can't form a proper conclusion.

 

And dismissing intended use-case scenario is also not good. A consumer hard drive could have a 99% reliability and last for 10 years under "consumer desktop" usage, but could have a terrible reliability under enterprise datacenter usage.

 

Does that mean said HDD is unreliable and more likely to die? No - not necessarily. In fact, it could have a higher reliability under consumer desktop scenarios compared to other similar HDD's, yet a lower reliability under datacenter usage. That's entirely possible, since it depends on how the drive is used, and what that particular model is susceptible to.

 

24/7 server environment does not reflect the intended use scenario of a consumer-based desktop drive. If anything the conditions are incredibly unrealistic and unfair due to the significant increase in heat and vibrations from the drives being rack-mounted so closely together. 

 

If anything it just shows that Seagate desktop drives do not perform good in a server environment which is, once again, not the environment intended for such drives.

[sobe]

@sobe the problem is that their methodology is incredibly flawed, and that they lack a lot of crucial information. Blackblaze themselves probably have that extra info, but they have yet to publish it.

 

Take for example, they list average age of the HDD, yes, but they do not list the age of each HDD as it dies. Nor do they list the average age of a dead HDD. That can skew the data into very misleading numbers.

 

It doesn't even necessarily tell you how reliable each drive model is on average, since without those crucial missing data points, we can't form a proper conclusion.

 

And dismissing intended use-case scenario is also not good. A consumer hard drive could have a 99% reliability and last for 10 years under "consumer desktop" usage, but could have a terrible reliability under enterprise datacenter usage.

 

Does that mean said HDD is unreliable and more likely to die? No - not necessarily. In fact, it could have a higher reliability under consumer desktop scenarios compared to other similar HDD's, yet a lower reliability under datacenter usage. That's entirely possible, since it depends on how the drive is used, and what that particular model is susceptible to.

 

For the most part yes, and I agree, as I placed in my statement "that no typical user will be near doing". Whether anyone likes it or not though, the drives are used 24/7 in a server rack, think of it like the SSD torture test(s) that were done... Typical users won't be throwing that much data at an SSD within a month's time, but it still gives us an idea of how long they last. Otherwise, we have to discredit ANY testing methodologies for any hardware reliability testing including those done by Linus or any other in the industry, no matter how accurate or scientific, because "no user will ever do that".

 

This does not mean I am saying Backblaze's data is reflective of what end-users would face, that would be flawed thinking on my part as I am quite sure if I bought a Seagate drive, chances are it would last me a good several years. Hell, before my house fire, I had a SCSI IBM Deathstar that still ran fine, and those things were notorious for failure back in the day.

 

Completely agree on the missing info though, but even without it we can still get a broad picture. But yes, individual drive or perhaps batch listings would be informative alongside install date. Hell, I'm sure they have that info actually.

 

I just think that people hear Backblaze and go into a stupidity frenzy with complete bias simply because it is in a server environment. I dunno, I've seen both sides on various forums and reddit and the entirety just seems dumb, just take it for what it is, information, if you don't agree with it, so be it, no need to seek attention. Btw, it is nice to see someone with information in a post other than the typical "its a server environment zomg that means it means nothing".

[EmeraldFlame]

Seagate simply tend to have a higher failure rate than comparable HGST or WD drives, but your chance of failure when buying a drive is really low. If going for reliability HGST/Hitachi, the WD sub-brand, seems to have the highest reliability according to BackBlaze and Seagate had the highest failure rate.

https://www.backblaze.com/blog/hard-drive-reliability-stats-for-q2-2015/

 

I'm partial to WD myself, but I've picked up some HGST drives as of late and have had no issues. For me personally, a long time ago I purchased some Seagate drives and while the RMA process was relatively painless, they kept dying on me, regardless of Seagate's willingness to keep replacing drives. It was only for 2 of the 4 Seagate drives that I had to RMA several times, but it left a sour taste in my mouth and I've just had great luck with WD and Hitachi over the years.

 

 

My God, people need to stop referencing this so-called "study". It barely gives any details on the methodology used for testing and has been discredited so many times. Linus even crapped all over it on the WAN show many months back. 

I just want to reiterate what @Mujteba said about this because way to many people link to this thing. BackBlaze's ongoing 'study' is so amazingly bad as a measurement of drive longevity it isn't even funny. They tend to use consumer grade drives in an enterprise environment and report their findings without actually handing out the real drive usage. The drives are obviously ran outside of their suggested operating environment and far beyond the suggested workload level for that kind of drive. This is like taking a Toyota Prius to a race track then being surprised that you broke some stuff. 

The drives they are testing often weren't made for their usage scenario and aren't designed to handle the environmental factors or the load they are throwing at them.

[sobe]

I just want to reiterate what @Mujteba said about this because way to many people link to this thing. BackBlaze's ongoing 'study' is so amazingly bad as a measurement of drive longevity it isn't even funny. They tend to use consumer grade drives in an enterprise environment and report their findings without actually handing out the real drive usage. The drives are obviously ran outside of their suggested operating environment and far beyond the suggested workload level for that kind of drive. This is like taking a Toyota Prius to a race track then being surprised that you broke some stuff. 

The drives they are testing often weren't made for their usage scenario and aren't designed to handle the environmental factors or the load they are throwing at them.

 

Look up one post