Sign in with
Sign up | Sign in
Your question

Will RE3 likely outlast Caviar Black?

Last response: in Storage
Share
June 28, 2010 3:51:54 AM

Please compare long-term reliability of a Western Digital RE3 against that of a WD Caviar Black. Which one is likely to last longer in a desktop, and by what ratio? Can you support your opinion with convincing facts, or is it more of a guess or hunch?

WD rates the RE3's MTBF at 1.2 million hours. Does anybody know MTBF for the Caviar Black, and if so, how do you know?

Please assume that performance is not an issue. Reliability and longevity are the issue. Please assume no use of RAID. Please assume that the RE3's TLER feature, which limits its normal error recovery functions to accomodate RAID, is disabled so that the drive will recover from errors appropriately in a non-RAID environment. Please assume the system is to be operated about 12 hours each day, with two or three cold boots and power downs per day. Note that an RE3 is designed for 24/7 use. Will having 12 hours off each day extend its life? OR will two or three cold boots and power downs each day inflict thermal stress which will shorten its life? Will this usage pattern help shift the balance in favor of RE3 and against Caviar Black, or vice versa?

The system will operate 12 hours per day, but the RE3 or WD Caviar Black disk will usually receive only very limited use, as a secondary disk to perform certain backups, except when the primary disk fails and then the secondary will be used to run the system until the primary is replaced. So what if the secondary disk is placed into a separate external hard disk enclosure, to be powered up only during the few minutes each day when it is time to make a backup - will an RE3 last longer doing this than a Caviar Black, or vice versa? I would like to know the answer both using the secondary as an internal drive, and also as an external drive in an enclosure.
a b G Storage
June 28, 2010 7:33:12 AM

Quote from the WD site:
"We no longer measure the reliability of our hard drives using Mean Time Between Failure (MTBF). Our current drive reliability is measured using Component Design Life (CDL) and Annualized Failure Rate (AFR). The Component Design Life of the drive is 5 years and the Annualized Failure Rate is less than 0.8%."

Link at, http://wdc.custhelp.com/cgi-bin/wdc.cfg/php/enduser/std...
m
0
l
a b G Storage
June 28, 2010 2:13:02 PM

About time they started doing something like that.

A 1,200,000 hour mtbf is about 135 years of continuous operation. I am willing to bet that WD did not design it in the 1870's, build ten of them, run them continuously, have five fail by this year, and so declare a 1.2 million hour mtbf.

But, yes, the RE3 should outlast the Black. OTOH, if you were to get a random sample of each, you could possibly get a golden Black and a lemon RE3.
m
0
l
Related resources
June 28, 2010 3:19:34 PM

Jonmor68 said:
Quote from the WD site:
"We no longer measure the reliability of our hard drives using Mean Time Between Failure (MTBF). Our current drive reliability is measured using Component Design Life (CDL) and Annualized Failure Rate (AFR). The Component Design Life of the drive is 5 years and the Annualized Failure Rate is less than 0.8%."

Link at, http://wdc.custhelp.com/cgi-bin/wdc.cfg/php/enduser/std...



This is a strange and incomprehensible statement. It purports to offer the CDL and AFR to measure reliability of "the drive", but nowhere specifies WHICH drive. WD's website and marketing materials clearly and repeatedly emphasize that some of its drives are more reliable than others. This gives us nothing to answer any of the questions I have posed, which revolve around trying to compare reliability of different WD drive models.
m
0
l
June 28, 2010 3:25:17 PM

jsc said:
About time they started doing something like that.

A 1,200,000 hour mtbf is about 135 years of continuous operation. I am willing to bet that WD did not design it in the 1870's, build ten of them, run them continuously, have five fail by this year, and so declare a 1.2 million hour mtbf.

But, yes, the RE3 should outlast the Black. OTOH, if you were to get a random sample of each, you could possibly get a golden Black and a lemon RE3.


I think you misunderstand the concept of MTBF. It is the mean time between failures, assuming that the drive is regularly swapped out and replaced with a new drive whenever it reaches the end of its expected life cycle (apparently five years). WD does not assert that any single drive will last 135 years. They are saying that if you replace the drive every five (or so) years, then the average time between failures will be 1.2 million hours. Such a failure rate can be predictively estimated (or guessed) with limited accuracy by design engineers, and then it can be subsequently verified and corrected by running a large sample of drives for one useful life cycle. Such empirical verification would require only the apparent life cycle of five years, rather than the ridiculous 135 year interval you have suggested.

I hope that you won't be offended that I can't rely on your answer, since you do not appear to understand the concepts involved. But I appreciate your effort to assist.
m
0
l
a c 103 G Storage
June 28, 2010 4:02:11 PM

Here is the deal with hard drives. Expect them to fail at any time, no matter what the drive. Backup accordingly. Unless there is some hugely known issue such as with the old IBM Deskstar (known well as the DEATHstars) don't worry too much if one drive is rated to run 1.2 million hours or 1.5 million hours.
m
0
l
June 28, 2010 6:45:43 PM

hang-the-9 said:
Here is the deal with hard drives. Expect them to fail at any time, no matter what the drive. Backup accordingly. Unless there is some hugely known issue such as with the old IBM Deskstar (known well as the DEATHstars) don't worry too much if one drive is rated to run 1.2 million hours or 1.5 million hours.


I am very worried about it, and that is why I started this thread. I do not understand any reason why I should not worry about it. So I will continue to worry about it. I hope that somebody will be able to offer some facts to help me choose the disk which will be more reliable for my situation, rather than less reliable.
m
0
l
a c 103 G Storage
June 28, 2010 8:15:36 PM

tomsusername said:
I am very worried about it, and that is why I started this thread. I do not understand any reason why I should not worry about it. So I will continue to worry about it. I hope that somebody will be able to offer some facts to help me choose the disk which will be more reliable for my situation, rather than less reliable.


You are looking for enterprise class drives then,

Hitachi Ultrastar http://www.hitachigst.com/internal-drives/enterprise/

Seagate Contallation or Savivio or Cheetah http://www.seagate.com/www/en-us/products/enterprise/

You probably won't like their prices too much though :-)
m
0
l
June 28, 2010 9:13:39 PM

hang-the-9 said:
You are looking for enterprise class drives then,

Hitachi Ultrastar http://www.hitachigst.com/internal-drives/enterprise/

Seagate Contallation or Savivio or Cheetah http://www.seagate.com/www/en-us/products/enterprise/

You probably won't like their prices too much though :-)


No, at this point, I am not looking for enterprise class drives. I am instead looking for specific facts and specific numbers and statistics, so that I can compare reliability of an enterprise class drive, the RE3, to that of a consumer class drive, the Caviar Black, so as to make an informed choice as to which is most appropriate for my situation. Does anybody have any relevant facts about measuring reliability of the RE3 or Caviar Black?
m
0
l
a c 82 G Storage
June 29, 2010 1:11:51 AM

None will necessarily be more reliable than the other. The RE3 and RE4 drives are designed to be used in Raid configurations and that's why they're referred to as enterprise class drives. This review explains the difference quite well: http://techreport.com/articles.x/15588

If you plan on using only one or two drives, then go with the Caviar Black. If you'll setup an 8-drive Raid, then use appropriate hard disks like the RE4.
m
0
l
June 29, 2010 4:03:55 AM

GhislainG said:
None will necessarily be more reliable than the other. The RE3 and RE4 drives are designed to be used in Raid configurations and that's why they're referred to as enterprise class drives. This review explains the difference quite well: http://techreport.com/articles.x/15588

If you plan on using only one or two drives, then go with the Caviar Black. If you'll setup an 8-drive Raid, then use appropriate hard disks like the RE4.


Please remember that I will NOT be using RAID, so I need to compare the alternatives based on no use of RAID.

Why do you believe that neither RE3 nor Caviar Black will have better reliability statistics than the other? The review you cited states that the RE3 do receive better quality control than the Caviar Black. WD's website markets the RE3, but not the Caviar Black, based on RE3's "enhanced reliability". Are you saying this is just BS? Why? or why not?
m
0
l
a c 82 G Storage
June 29, 2010 12:40:30 PM

It basically is the same hard disk with the same 5-year warranty, but with a different firmware. I have several Caviar Black at home and they are very good, but they are not used for Raid.

The WD1002FAEX costs $95 and the Raid Edition WD1002FBYS costs $140 on Newegg.com. Is the Raid Edition worth it in an enterprise environment? To me TLER certainly is worth it to keep the Raid working properly, but it doesn't make them better than the desktop version. If you believe that the RE3 is a better drive for you, then why not buy it? It only is $45 more.
m
0
l
June 29, 2010 4:04:42 PM

GhislainG said:
It basically is the same hard disk with the same 5-year warranty, but with a different firmware. I have several Caviar Black at home and they are very good, but they are not used for Raid.

The WD1002FAEX costs $95 and the Raid Edition WD1002FBYS costs $140 on Newegg.com. Is the Raid Edition worth it in an enterprise environment? To me TLER certainly is worth it to keep the Raid working properly, but it doesn't make them better than the desktop version. If you believe that the RE3 is a better drive for you, then why not buy it? It only is $45 more.


I haven't bought either disk, because I don't have the belief you attributed to me, and because I do not yet have enough information to make a decision.

How do you know the two disks are identical, apart from firmware? I have seen this asserted in reviews, and I have been told contradictory things by WD employees, but I haven't seen any written confirmation from WD, and WD's website at least tries to suggest the opposite. How do you know?

And what about the claim of superior quality control for the RE3, as compared to Caviar Black? Is this true? Does it matter, and if so, how much does it matter?
m
0
l
a c 103 G Storage
June 29, 2010 4:55:03 PM

I think you need to visit WD factory, and see how these things are made, buy 50 of each and run them for 5 years under various conditions, then see how they hold up.

That will be about the only answer that will be really good since you keep asking for some positive proof that one will outlast the other. I doubt anyone here has a full disk testing lab setup in their house to answer what you are looking for.
m
0
l
a c 82 G Storage
June 29, 2010 5:51:59 PM

Quote:
And what about the claim of superior quality control for the RE3, as compared to Caviar Black? Is this true? Does it matter, and if so, how much does it matter?
If enterprise drives were better, then I would expect to replace fewer of those than desktop units. Unfortunately that isn't always true. I have both enterprise and desktop drives at home. My comments are not based on home use only; they are based on what I also see at my customer sites. As hang-the-9 suggested, buy both and see which one survives the other. Also read the bad reviews about the WD1002FAEX on Newegg. Please read them as an IT person who understands that stuff and not as a home user who doesn't know the difference between Raid ready drives and their desktop counterpart. Ignore comments about the drive not working on an old PC, etc. Don't forget to also read the bad reviews about the WD1002FBYS. Again do it as an IT person, not as a home user.
m
0
l
June 29, 2010 8:45:38 PM

GhislainG said:
Quote:
And what about the claim of superior quality control for the RE3, as compared to Caviar Black? Is this true? Does it matter, and if so, how much does it matter?
If enterprise drives were better, then I would expect to replace fewer of those than desktop units. Unfortunately that isn't always true. I have both enterprise and desktop drives at home. My comments are not based on home use only; they are based on what I also see at my customer sites. As hang-the-9 suggested, buy both and see which one survives the other. Also read the bad reviews about the WD1002FAEX on Newegg. Please read them as an IT person who understands that stuff and not as a home user who doesn't know the difference between Raid ready drives and their desktop counterpart. Ignore comments about the drive not working on an old PC, etc. Don't forget to also read the bad reviews about the WD1002FBYS. Again do it as an IT person, not as a home user.


I don't want to spend time and money experimenting to assess hard drive reliability. A sample of two drives, one from the RE3, and one from the Caviar Black, would not be sufficiently large to provide the information I need. The information I need requires statistics from a large population of drives.

I can't understand why anybody would buy a hard drive from Newegg, since their practices with respect to poor packaging can only be described as reckless and foolish. I also can't understand why anybody would want a hard drive after it has been shipped via UPS. Earth to Newegg: hard drives are fragile! I pay close attention to the Newegg reviews, but their usefulness is limited by Newegg's reckless packaging and shipping practices for hard drives. I read all the one-egg and two-egg reviews for the two drive models you specified, but I was unable to draw any relevant conclusions from them. How about you?
m
0
l
June 29, 2010 8:57:15 PM

hang-the-9 said:
I think you need to visit WD factory, and see how these things are made, buy 50 of each and run them for 5 years under various conditions, then see how they hold up.

That will be about the only answer that will be really good since you keep asking for some positive proof that one will outlast the other. I doubt anyone here has a full disk testing lab setup in their house to answer what you are looking for.


No, really, all I need are reasonable statistics from WD, who actually DO have a full disk testing lab. I never asked for positive proof that an RE3 will outlast a Caviar Black, or vice versa. I asked only for some factual evidence, such as MTBF stats. I want stats because I don't want to waste time, money, and data by experimenting with inferior drives.
m
0
l
June 29, 2010 9:02:06 PM

A WD employee told me that the RE3's Annualized Failure Rate is 0.73%, and that of Caviar Black is 0.5%, within their 5 year design lives. I'm sure the RE3's failure rate was based on the assumption of 24/7 usage, but I still need to know the assumptions about daily operating hours made for the Caviar Black. I need to know these assumptions, so that I can compare their respective failure rates I would expect to see in my particular usage pattern. The employee will try to get me that info, but doesn't know if e can.
m
0
l
a c 82 G Storage
June 29, 2010 9:15:16 PM

My conclusion is that the desktop drive is as good as the enterprise drive. I ignore people who complain about it not being fit for RAID (that's why WD have the RE3 and RE4 drives) and other similar and useless complaints. I'm still happy with the Caviar Black drives that I use at home as well as the enterprise drives. Since I only experienced one enterprise class SCSI drive failure in over 10 years, I can't complain.

I bought a few hard drives from Newegg and I agree that their packaging isn't ideal (DirectCanada package them better), but I never had problems with them.

If you read the WD shock specs:
Operating Shock (Read) 30G, 2 ms
Non-operating Shock 300G, 2 ms

In other words, a drive in a carton box that's dropped a couple feet won't likely be damaged. I own a Samsung F1 that was dropped on a carpeted floor and it still works fine, but it wasn't spinning when it slipped from the desk.

The Samsung F3 (3 year warranty) also is an excellent drive.
m
0
l
April 21, 2011 1:19:53 PM

tomsusername said:
Please compare long-term reliability of a Western Digital RE3 against that of a WD Caviar Black. Which one is likely to last longer in a desktop, and by what ratio? Can you support your opinion with convincing facts, or is it more of a guess or hunch?

WD rates the RE3's MTBF at 1.2 million hours. Does anybody know MTBF for the Caviar Black, and if so, how do you know?

Please assume that performance is not an issue. Reliability and longevity are the issue. Please assume no use of RAID. Please assume that the RE3's TLER feature, which limits its normal error recovery functions to accomodate RAID, is disabled so that the drive will recover from errors appropriately in a non-RAID environment. Please assume the system is to be operated about 12 hours each day, with two or three cold boots and power downs per day. Note that an RE3 is designed for 24/7 use. Will having 12 hours off each day extend its life? OR will two or three cold boots and power downs each day inflict thermal stress which will shorten its life? Will this usage pattern help shift the balance in favor of RE3 and against Caviar Black, or vice versa?

The system will operate 12 hours per day, but the RE3 or WD Caviar Black disk will usually receive only very limited use, as a secondary disk to perform certain backups, except when the primary disk fails and then the secondary will be used to run the system until the primary is replaced. So what if the secondary disk is placed into a separate external hard disk enclosure, to be powered up only during the few minutes each day when it is time to make a backup - will an RE3 last longer doing this than a Caviar Black, or vice versa? I would like to know the answer both using the secondary as an internal drive, and also as an external drive in an enclosure.


If you are not going to use 24x7, why not go for a main Caviar Black and a Blue or Green drive as a mirror? That would be more foolproof because even if failure rate of RE3 is negligible, the probability of both black and blue drives failing at once would be even less. Black is pretty reliable itself and don't forget extra performance boost that you would enjoy with it.
[Why pay a dollar for a bookmark? Why not use the dollar for a bookmark?]
m
0
l
a c 289 G Storage
April 21, 2011 1:51:08 PM

tomsusername said:
I am very worried about it, and that is why I started this thread. I do not understand any reason why I should not worry about it. So I will continue to worry about it. I hope that somebody will be able to offer some facts to help me choose the disk which will be more reliable for my situation, rather than less reliable.

Tomsusername

The fact is, disks fail. If you are concerned with reliability and the safety of your data, which you should be, you need to look at additional layers of protection.

As a mathematical note, the mean-time-to-failure or other statistics are not that useful for someone who owns ten drives. I've had multiple drives of the same model, where one fails in two months and one was thrown out after five years of use. You are dealing with a very small sample size, and outliers are entirely possible.

That said. Stop me if you already know this, but in addition to getting the most reliable drives that you can, you should
  • Backup early and backup often. If your data is critical to you, have offsite copies of the backups.
  • Do image backups of your OS drive. I've lost OS drives and been back up and running in a half-hour after pulling a spare and my backup from my drive case, plus the boot floppy for the restore program.
  • If your need for reliability includes high availability, so the above half-hour is unacceptable to you, either use RAID 1 (and have a spare controller of the same type on hand) or have a hot-standby system.

    No one drive is going to give you a guarantee of success. Layers of best practices will protect you against all but the most outlandish circumstances. Having a hot-standby system in a managed datacenter in another country is probably the only way to protect against everything except global plague or war.
    m
    0
    l
    May 5, 2011 9:34:13 PM

    I bought 4 150GB Raptors 5 years ago and ran them on a RAID 5 in my desktop. They started failing at 4 years. every two or three months another one would drop out and I RMA'd them. I kept running the system while I waited on the new drive to come in. The last one failed one month before the 5 year warranty was up. I think htey were all the same batch. No loss of data.
    m
    0
    l
    May 5, 2011 9:34:21 PM

    I bought 4 150GB Raptors 5 years ago and ran them on a RAID 5 in my desktop. They started failing at 4 years. every two or three months another one would drop out and I RMA'd them. I kept running the system while I waited on the new drive to come in. The last one failed one month before the 5 year warranty was up. I think htey were all the same batch. No loss of data.
    m
    0
    l
    May 10, 2011 2:45:20 PM

    tomsusername said:
    I am very worried about it, and that is why I started this thread. I do not understand any reason why I should not worry about it. So I will continue to worry about it. I hope that somebody will be able to offer some facts to help me choose the disk which will be more reliable for my situation, rather than less reliable.


    Hard Drives fail at any time for any reason. Each part is unique. You cannot know how many particles of dust made it into your particular unit no matter which model you chose. Your choices are to either a) stop worrying, or b) rethink your use of raid. The industry at large knows that drives can and *do* fail all the time, and the solution they came up with is raid. You can either use raid or take a valium, but comparing two very good drives by the same manufacturer isn't going to help you any.

    -Fluffy
    m
    0
    l
    a c 371 G Storage
    May 10, 2011 3:17:48 PM

    I agree with what some of the others here have said. The RE drives may have better quality control, so the oddsof them lasting longer is better than the blacks, but in the end, both of those drives will fail at some point in the future.

    Backups are the only solution to prevent data loss.
    m
    0
    l
    February 17, 2012 12:00:08 AM

    There are only two kinds of drives. . . . . Those that have failed and those that will fail.
    m
    0
    l
    January 13, 2013 1:01:09 AM

    Tom, I completely sympathize with your instance and frusration in getting the information you're seeking, as I am now hittng the sam roadblocks in planning my storage investment. The fact that drive failures are statistical rather than predictable doesn't mean that those statistics shouldn't be available and comprehensible.

    e.g., you asked 'if i cold boot a raid drive a few times a day, is it more destructive than leaving them on?" (i.e. are raid drives more reliable in all situations, or only in raid situations?). A very good general questions that no one even tried to answer, even if the answer is 'i don't know / the drive manufacturers won't tell us.'

    The fact is, when you raise good questions that people haven't thought of or don't know the answer to, they (at first) tend to push your questions aside or ignore them to maintain the illusion that they are experts on the topic. I guess their egos are too fragile somehow to work with you on finding the right answer. If we came together as a community and pushed manufacturers for answers to these kinds of questions, we would get a lot farther in increasing the reliability of our builds, and save a lot of headaches and money in the process.
    m
    0
    l
    !