What is true reliability of MBLD and the safety of your data?

I used My Book Life Duo for storing important professional and family data. To add extra security I had RAID 1 configuration. One morning, two months after warranty expiration, a message appeared on my computer screen saying that a network drive is not accessible. An attempt to open web access to the MBLD resulted in a screen filled with gibberish, a very small sample of which you can see here.

Config|a:5{s:9:“userAgent”;s:32:“009bc9236061daaf8f5d8ad5903162cc”;s:4:“time”;i:1399485264; s:7:“timeout”;i:10;s:13:“box_is_secure”;s:5:“false”;s:4:“lang”;s:5:“en_US”;}owner|a:1:{s:19:“owner_configuration”; a:4:{s:10:“owner_name”;s:5:“admin”;s:8:“fullname”;s:20:“System Administrator”;s:12:“empty_passwd”;s:4:“true”;s:12:“owner_passwd”;s:0:"";}}

One would assume that with RAID 1 system and one drive failing the data should not be lost. As it turns out that, besides the box being unresponsive, both drives, otherwise seemingly good, have corrupted file systems – critical data have been lost. This is very disconcerting as primary role of devices like MBLD is storing data safe, there were no indications about storage corruption coming, the equipment was kept in good environmental conditions, no firmware upgrade, etc.

  1. Can somebody competent tell me what might went wrong?
  2. How one can keep important data safe?
  3. What is the true reliability of hardware like MBLD? Being a customer for of WD for more than ten years (purchased more than twenty five drives), I have very unsettling impression that drives failing right after the warranty expiration date are happening with increasing frequency.

Qazed wrote:

  1. How one can keep important data safe?

There is only ONE answer to #2:    BACKUPS.   RAID is *NOT* a Backup.

Are you saying that, in practical terms, instead of having two drives in RAID 1, it is better for data safety to use them separately: one for work, one for backup? That is hard to believe.  Assuming I would do the latter and designate one drive from MY MBLD for work and the second for backup. How that could possibly save me in the described case? The outcome would be the same.

If you search WD site for “RAID 1”, you will find numerous statements like these from
    http://www.wdc.com/wdproducts/library/AAG/ENG/4178-705284.pdf
“… And with RAID1 technology, you have double-safe backup for your data. …” and
“… Half of the capacity is used to store your data and the other half is used for a duplicate copy. …”
The claim is, as I understand, that by mirroring with RAID 1 one makes an extra copy of the data for safety. Isn’t that in fact a backup?  On other hand, if RAID 1 (or presumably other versions like RAID 5) is not good for protection - what is the purpose?

The essence of my questions remains the same: (1) how it can possibly happen that entire system, designed for storing data and in RAID 1 configuration collapses in such spectacular way for no apparent reason, as described in my posting, and (2) what are truly reliable ways to store the data. How many redundant systems one needs to buy and operate?

Qazed wrote:

…Isn’t that in fact a backup?  On other hand, if RAID 1 (or presumably other versions like RAID 5) is not good for protection - what is the purpose?

No, RAID-1 is NOT a backup.

If you have a RAID1 disk volume, and you or your software somehow accidentally delete or screw up the data:  It’s gone.  

The purpose and benefit of RAID is indicated by its name:  Redundant Array of (inexpensive) Disks.    

RAID provides REDUNDANCY (or simple “disk backup”), which is not the same as a data backup.

If a member of a RAID1 (or higher) array fails, you replace the failed unit, and you’re back in business with no loss of data.

In a true backup, you have historical copies of data (usually more than one) and you can restore back to a point in time.

I back up my systems every night, and keep backups for 90 days.  If I screw up one of my files, I can go back in time to any of the 90 prior days and restore that version.

Re:

I back up my systems every night, and keep backups for 90 days.  If I screw up one of my files, I can go back in time to any of the 90 prior days and restore that version.

If you use MBLD like one I did, these 90 days can be one in a blink for no apparent reason. You seem to focus on largely semantic difference between data backup and disk backup in context of particular scenario. I am interested in very practical questions aimed at avoiding problems like this. Assume that was a backup device. Are you saying one backup is not enough? Should have two separate backup devices, three? If RAID 1 does not increase my protection why WD promotes it?

So far, nobody attempted to answer WHAT could possibly cause such disaster. If I don’t know the cause, HOW I can prevent the problem in future? Is the device even repairable? How to diagnose it? The web access is gone - what is the procedure of bringing it back to life. Should I use as an organ donor - reformat hard drives for reuse and trash the rest?

Qazed wrote:

If RAID 1 does not increase my protection why WD promotes it?

RAID1 *does* protect your data in the event of a disk failure – and ONLY in the event of a disk failure.   RAID doesn’t protect you from “pilot error.”

You also need to understand that this is not a function of how the WD works…  The limitations of RAID are universal.

As to how many separate backups you need?  No one here can tell you.   Everyone’s requirements are different, and there are entire branches of computer science that deal with attempts to answer that exact question.

Every additional backup target reduces your risk of data loss.   But at some point, there’s a diminishing return that varies.

You seem to be obsessed with what you describe as ‘pilot error’. I admit, like many others, I can and I do damage files occasionally. But these damages are relatively small in scope and easy to fix. You seem to have difficulty to understand that people might have data like documents, family photographs and such that are stored in fixed archived form. The risk of your exaggerated ‘pilot error’ is negligible. What happened in this case is massive - in my opinion - loss of 2.5TB mirrored data on RAID 1 system on WD’s flagship product completely spontaneously, FOR NO APPARENT REASON. No one, including you, seem have any clue WHY. No one knows how to bring MBLD back to life even without restoring the data. By design the access to MBDL is through web interface, which is now completely gone (!). Doesn’t that sound like a design flaw?

I can assess the level of risk I can take myself if there are metrics available such as what is an average time of life of MBLD, PoH, probabilities of failing, possible reasons for failures (WHY the failure happened here), etc.? You seem to be unable or unwilling to provide such information.

I would be interested to know, if such failures happened in past and how often - getting that seems likely hopeless. I would presume that if WD engineers are interested in improving quality of their products, one would get in touch with me to collect facts about such spectacular failure - unless that is not new for them.

Because beside hair splitting on minor semantic aspect and repeating commonly known basic information I am not offered any constructive technical guidance, I will now thank you very much for your time, wish you luck in future endeavors, and leave this exchange.

Oh, one more thing. I decided to give chance others and just ordered Synology DS414 loaded with four Seagate EST4000VN00. Accept this as Solution