INTERACT FORUM

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1]   Go Down

Author Topic: Is it dying?  (Read 9438 times)

marko

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8941
Is it dying?
« on: August 13, 2015, 12:38:22 am »

So, I've had this 3 Tb drive just shy of two years. It's the drive that takes any 'incoming' stuff so it can be pretty active, but even so...

HD Sentinel has been throwing alerts for it, predicting total failure by the end of August, but, despite having a replacement drive in the wings, I'm not convinced...

09/08/2015 16:17:46,#187   Reported Uncorrectable Errors  335 -> 724
09/08/2015 03:09:49,#187   Reported Uncorrectable Errors  320 -> 335
08/08/2015 20:13:01,#187   Reported Uncorrectable Errors  260 -> 320
08/08/2015 15:59:03,#187   Reported Uncorrectable Errors  115 -> 260
08/08/2015 02:40:44,#187   Reported Uncorrectable Errors  95 -> 115
04/08/2015 09:29:20,#198   Off-Line Uncorrectable Sector Count  16 -> 96
04/08/2015 09:29:20,#197   Current Pending Sector Count  16 -> 96
03/08/2015 17:11:42,#198   Off-Line Uncorrectable Sector Count  0 -> 16
03/08/2015 17:11:42,#197   Current Pending Sector Count  0 -> 16
03/08/2015 17:11:42,#187   Reported Uncorrectable Errors  10 -> 95

Nothing for four days now. I found four files that were unreadable or corrupt. If I tried to do anything at all with them, even delete them, the whole PC would slowly become unusable. I managed to get rid of them via command line eventually. The entries in the log from the 8th and 9th coincide with my efforts to first, identify the rogue files, and then delete them.

So, now they're gone, is the log quiet because nothing else has been written there yet?
Doesn't the HDD know not to write there in the future, I forget the tech, something to do with spares, perhaps?
Is the dire warning because the drive is maybe out of spares?

I have what I need of the drive backed up, and a spare ready to go in if needed, so no big freak out going on, just curious as to whether or not the software is on point?

-marko

jmone

  • Administrator
  • Citizen of the Universe
  • *****
  • Posts: 14267
  • I won! I won!
Re: Is it dying?
« Reply #1 on: August 13, 2015, 12:53:01 am »

The Magic 8 Ball says your drive is going to Die.... horribly and soon.  Here is an interesting read on some stats from 40,000 drives which indicated that the "Big 5" are:
SMART ID 5 (0x05): Relocated Sectors Count
SMART ID 187 (0xBB): Reported Uncorrectable Errors
SMART ID 188 (0xBC): Command Timeout
SMART ID 197 (0xC5): Current Pending Sector Count
SMART ID 198 (0xC6): Uncorrectable Sector Count

If you want an expert opinion then you can send a report via HDS and the dev has always replied to me. 

As long as you are not relying on it then it does not really matter but I'd be plugging the new drive in now.  Once any data is transferred you could then always do a "Reinitialise disk surface" in HDS.... but I've found that these sort of errors just keep growing.
Logged
JRiver CEO Elect

RoderickGI

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8186
Re: Is it dying?
« Reply #2 on: August 13, 2015, 06:34:21 am »

HD Sentinel is a very good tool at predicting failure. Based on my observations of hard drive failures, in addition to all those statistics. The drive is going to fail soon. The problem will escalate, even if you don't know about it because you haven't accessed affected files.

I also would replace the drive now. Then I would subject the drive to some performance tests outside my usual drive set, to see how bad it is. I suspect you will see cascading failures under load. Which if nothing else, will push the drive over the top to be a warranty replacement.

At two years old, it should still be within warranty? Most warranties are 3 years for good drives. I think HD Sentinel reports also highlight which failure modes should trigger a warranty replacement. There is some good information on their forum as well.
Logged
What specific version of MC you are running:MC27.0.27 @ Oct 27, 2020 and updating regularly Jim!                        MC Release Notes: https://wiki.jriver.com/index.php/Release_Notes
What OS(s) and Version you are running:     Windows 10 Pro 64bit Version 2004 (OS Build 19041.572).
The JRMark score of the PC with an issue:    JRMark (version 26.0.52 64 bit): 3419
Important relevant info about your environment:     
  Using the HTPC as a MC Server & a Workstation as a MC Client plus some DLNA clients.
  Running JRiver for Android, JRemote2, Gizmo, & MO 4Media on a Sony Xperia XZ Premium Android 9.
  Playing video out to a Sony 65" TV connected via HDMI, playing digital audio out via motherboard sound card, PCIe TV tuner

jmone

  • Administrator
  • Citizen of the Universe
  • *****
  • Posts: 14267
  • I won! I won!
Re: Is it dying?
« Reply #3 on: August 13, 2015, 06:37:23 am »

Good point.  I've returned drives with a print out from HDS for WTY replacement.
Logged
JRiver CEO Elect

mwillems

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 5175
  • "Linux Merit Badge" Recipient
Re: Is it dying?
« Reply #4 on: August 13, 2015, 08:31:31 am »

That many uncorrectable errors means impending doom.  Once you've gotten your data off of it safely (assuming you can get your data off it safely at this point), run a surface test in HDSentinel.  It will find scads of bad sectors and confirm the drive failure.  I have a three year old seagate that's circling the drain right now with climbing uncorrectable errors.

Admittedly, I have a small sample size, but I've never seen a drive with more than 3 or 4 uncorrectable errors that didn't fail* within 6 months.

*Fail in the sense that data on the drive became corrupted and inaccessible, not necessarily that the drive refused to mount although that usually followed soon after.
Logged

RoderickGI

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8186
Re: Is it dying?
« Reply #5 on: August 13, 2015, 06:43:20 pm »

Yep, and if you wait the six months without a good, reliable backup, by the time it fails lots of data is irretrievable.

Worse, if you have been running backups and only keep two months worth, all your backups are useless as well.

It just isn't worth holding out "just a little bit longer" for a drive that has started showing serious errors. Even if you clean up the drive and fix all errors, the drive has to be under a cloud of suspicion.

Fortunately you already have a safe backup of all data Marko.
Logged
What specific version of MC you are running:MC27.0.27 @ Oct 27, 2020 and updating regularly Jim!                        MC Release Notes: https://wiki.jriver.com/index.php/Release_Notes
What OS(s) and Version you are running:     Windows 10 Pro 64bit Version 2004 (OS Build 19041.572).
The JRMark score of the PC with an issue:    JRMark (version 26.0.52 64 bit): 3419
Important relevant info about your environment:     
  Using the HTPC as a MC Server & a Workstation as a MC Client plus some DLNA clients.
  Running JRiver for Android, JRemote2, Gizmo, & MO 4Media on a Sony Xperia XZ Premium Android 9.
  Playing video out to a Sony 65" TV connected via HDMI, playing digital audio out via motherboard sound card, PCIe TV tuner

marko

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8941
Re: Is it dying?
« Reply #6 on: August 14, 2015, 02:08:37 am »

Brilliant. Thanks for the feedback folks, it's appreciated, as always.

I'll try and find time to swap it out over the weekend. That'll make two Seagate and one Samsung drive gone west in the last eight weeks or so. Not impressed, I've been replacing with Western Digital ones. The brand wars have raged for years, and I have always been of the opinion that a consumer level HDD is just that, whoever makes it. Is one really better than the others?

Another thought occurred to me this morning...

My backups from this failing drive are the 'file/folder sync' variety. I can just reverse the sync on the new drive and put it all back. What if I had a drive image though? Would restoring from that copy the faults onto the new drive, or is that just not how that works?

jmone

  • Administrator
  • Citizen of the Universe
  • *****
  • Posts: 14267
  • I won! I won!
Re: Is it dying?
« Reply #7 on: August 14, 2015, 03:06:05 am »

They will all fail ... eventually.  I've gone through 100+ HDD over the years in my pools, but mostly through upgrading to larger sizes.  I have found it is specific models that are good or bad more than brands.  For example of the the WD 2TB HDD the EARS models (first of the 4K sectors) all failed horribly and quickly yet their but their EADS and EARX were fine.  I love my current 4TB Hitachis but they are slowly being replaced by bigger drives as I run out of space (was Seagate 6tb and now the new SMR 8TB ones).  Time will tell on these SMR Drives.
Logged
JRiver CEO Elect

astromo

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 2239
Re: Is it dying?
« Reply #8 on: August 14, 2015, 03:48:24 am »

Good convo guys. I'm filing this one upstairs for reference when I'll eventually need it.
Logged
MC31, Win10 x64, HD-Plex H5 Gen2 Case, HD-Plex 400W Hi-Fi DC-ATX / AC-DC PSU, Gigabyte Z370 ULTRA Gaming 2.0 MoBo, Intel Core i7 8700 CPU, 4x8GB GSkill DDR4 RAM, Schiit Modi Multibit DAC, Freya Pre, Nelson Pass Aleph J DIY Clone, Ascension Timberwolf 8893BSRTL Speakers, BJC 5T00UP cables, DVB-T Tuner HDHR5-4DT

RoderickGI

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8186
Re: Is it dying?
« Reply #9 on: August 14, 2015, 07:36:00 pm »

You had a Samsung fail? I thought those guys were doing pretty well with reliability.

I've had lots of brands fail over the years, and quite a while back just decided to standardise on Western Digital. Sometimes, better the devil you know. I've had few failures with WD, and I think all of them were EARS disks, which were good value and worked pretty well most of the time . . . until they failed. They failed quickly as well, so you couldn't "wait and see" how they went.

For the last couple of years or more, since they were introduced, I have been using WD Red disks. They seem to be that little bit more reliable, but time will tell.


Another thought occurred to me this morning...

My backups from this failing drive are the 'file/folder sync' variety. I can just reverse the sync on the new drive and put it all back. What if I had a drive image though? Would restoring from that copy the faults onto the new drive, or is that just not how that works?

Any file sync software, or even backup software, can backup corrupted files. As long as the first read and verify reads match. Sometimes a file can be corrupted but still look okay from the operating system point of view, so backup software won't pick up a problem. It is best to go back to a known good backup, before any issue became apparent, unless there have been a lot of changes since then. In which case, you may need to use a recent backup and then do some recovery work for broken files. A drive image could be in the same boat, but should be better as they tend to do more checks than sync software. A proper backup application with verify turned on will do more checks, and enable retention for a series backups, which is about as good as your can get.

I use EaseUS Todo Backup these days, after looking around for quite a while to replace my previous software, which let me down when I had my last drive failure, before I was using HD Sentinel. It has real backup features, and is priced much better than most professional software, which can be ridiculously priced. I don't use it for my media library though, just my system and data disks. I trust HDS for my media library, because it isn't critical. I do have backups of all my music though, just not TV, movies, etc.
Logged
What specific version of MC you are running:MC27.0.27 @ Oct 27, 2020 and updating regularly Jim!                        MC Release Notes: https://wiki.jriver.com/index.php/Release_Notes
What OS(s) and Version you are running:     Windows 10 Pro 64bit Version 2004 (OS Build 19041.572).
The JRMark score of the PC with an issue:    JRMark (version 26.0.52 64 bit): 3419
Important relevant info about your environment:     
  Using the HTPC as a MC Server & a Workstation as a MC Client plus some DLNA clients.
  Running JRiver for Android, JRemote2, Gizmo, & MO 4Media on a Sony Xperia XZ Premium Android 9.
  Playing video out to a Sony 65" TV connected via HDMI, playing digital audio out via motherboard sound card, PCIe TV tuner

marko

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8941
Re: Is it dying?
« Reply #10 on: August 17, 2015, 01:18:55 am »

Job's done. Thank you all. Gave everything a good clean while the power and the sides were off too, so all good now.

Quote
You had a Samsung fail? I thought those guys were doing pretty well with reliability.
I did. It just refused to mount. Windows would take an age to boot up, and you could hear it thrashing, then, eventually, Windows would just boot without it. I bought two of them, installed at the same time. HDS reports the other one with 100% health, no issues at all, with a power on time of 1615 days.

Quote
I've had lots of brands fail over the years, and quite a while back just decided to standardise on Western Digital. Sometimes, better the devil you know. I've had few failures with WD, and I think all of them were EARS disks, which were good value and worked pretty well most of the time . . . until they failed. They failed quickly as well, so you couldn't "wait and see" how they went.
I have one EARS and one EADS still going here, with 1800+ days of power on, in good health still, though, the pile of dead disks on my desk is five high, and one of those is an EARS!

I have been replacing drives with WD Red ones too. Having swapped out the dying drive at the weekend, I now have three of the 4Tb ones in action here. I am thinking of binning the desktop, and being undecided as to how best to keep the files online, figured these would be a good idea, thinking they would give me more options to choose from when the time comes.

When I was musing about disk images earlier, I was thinking more about the faults with the disk itself rather than the files. Is it possible to copy the faults from one drive to another like that, effectively harming the new drive, or are the bits and the hardware completely unrelated when restoring a disk image?

RoderickGI

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8186
Re: Is it dying?
« Reply #11 on: August 17, 2015, 02:41:42 am »

When I was musing about disk images earlier, I was thinking more about the faults with the disk itself rather than the files. Is it possible to copy the faults from one drive to another like that, effectively harming the new drive, or are the bits and the hardware completely unrelated when restoring a disk image?

The bits and the hardware an image is stored on are unrelated, in the sense you are asking. You can't copy hardware faults between disks. All those "Reported Uncorrectable Errors" though mean that the disk hasn't been able to read a section of the disk and write the files it contains safely to a new location, which is what it would normally do, and then lock out the bad sector that contained the problem. Disk drives fix errors all the time if they have trouble reading files, shown by SMART errors, a failed first read, or some other metric, by moving the files to good locations.

Real disk corruption, of the sort HDS reported to you, would be caused by small hardware failures, probably surface defects on the disk itself. These can be caused by minor bumps while the disk is working, causing the read heads to touch the disk, or particles inside the drive, or temperature issues, etc. Those won't be transferred to a backup disk or the disk you restore it to. But it is possible to copy the errors they create. So a bad sector that can't be read won't be backed up, and the backup software should either report it or fail and stop the backup. A restore of a disk image should also check the format and indexes and report or stop if there is a problem. In fact it should check the image as a first step in the restore process, or the default option should be to do so.

However if a disk has had a glitch at some time, but the sector and the index information about it is still good, then a corrupted file or disk sector can be backed up even to a disk image. Of course the problem will only show itself as file corruption.


On my last data disk failure, a lot of my specific problems were with JPG image files. Pictures I had taken with various cameras over time. I often sorted the directories I had them in by Date taken, which is not a well supported sort in Windows, so the disk got thrashed quite a bit when I sorted that way. At the time I was using the backup function in Norton 360 to do incremental file backups, rather than image backups. That software failed miserably when I tried to do a restore. For some reason the software had lost its record of the backup set, even though the index file containing the information was still there, and I was eventually able to restore most of the data. But it took a lot of intervention from me. Unfortunately Symantec were no help at all, and didn't know the trick I had to use to force the recovery to start, or wouldn't share it.

When I finally did have all but the small amount of lost data restored, I found that more than a few image files had corruption in them. They showed tearing, colour changes and so on in the image, or pieces of the image were in the wrong sequence. So it was file corruption, which had been read from the original disk and backed up to my backup disk. The backup software was verifying the files, so the file was corrupted on the original disk. That sort of corruption will also be backed up to an image, and restored to a new disk. It is the same as a Word document having a few characters changed because of a few bad bits, but the file size stays the same and is readable, just not human readable!

Anyway, good to hear you got it all fixed. Now get some of those drives replaced under warranty!
Logged
What specific version of MC you are running:MC27.0.27 @ Oct 27, 2020 and updating regularly Jim!                        MC Release Notes: https://wiki.jriver.com/index.php/Release_Notes
What OS(s) and Version you are running:     Windows 10 Pro 64bit Version 2004 (OS Build 19041.572).
The JRMark score of the PC with an issue:    JRMark (version 26.0.52 64 bit): 3419
Important relevant info about your environment:     
  Using the HTPC as a MC Server & a Workstation as a MC Client plus some DLNA clients.
  Running JRiver for Android, JRemote2, Gizmo, & MO 4Media on a Sony Xperia XZ Premium Android 9.
  Playing video out to a Sony 65" TV connected via HDMI, playing digital audio out via motherboard sound card, PCIe TV tuner

Trumpetguy

  • Citizen of the Universe
  • *****
  • Posts: 974
Re:
« Reply #12 on: August 17, 2015, 08:01:11 am »

With WD reds, anyone experienced high hovering temperatures? I have one red at quite constant 45 to 50 degC (even when picked out of the cabinet with 23 degC room temp). Two WD greens working around 32 degC.
Logged

mattkhan

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 3961
Re: Is it dying?
« Reply #13 on: August 17, 2015, 08:16:51 am »

With WD reds, anyone experienced high hovering temperatures? I have one red at quite constant 45 to 50 degC (even when picked out of the cabinet with 23 degC room temp). Two WD greens working around 32 degC.
not something I've ever seen with mine
Logged

Trumpetguy

  • Citizen of the Universe
  • *****
  • Posts: 974
Re: Re: Is it dying?
« Reply #14 on: August 17, 2015, 09:01:39 am »

not something I've ever seen with mine
I'll try adding some cooling fan for now. And replace the drive in not too long.
Logged

mwillems

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 5175
  • "Linux Merit Badge" Recipient
Re:
« Reply #15 on: August 17, 2015, 09:05:42 am »

With WD reds, anyone experienced high hovering temperatures? I have one red at quite constant 45 to 50 degC (even when picked out of the cabinet with 23 degC room temp). Two WD greens working around 32 degC.

I've never seen that;  I run 5 reds (three 3TB, one 4TB, and one 6TB) and they typically run as cool/cooler than my greens.  Did you do any firmware tweaking maybe?  Any clues in the SMART data?

To give you a calibration, I monitor and collect realtime data on all of my reds, and over the past year my reds have averaged between 30C-34C (some of them are slightly better ventilated than others).  The max recorded temp on any of them is 41C, and that was achieved during the initial "write-in" when I added that drive to the box. Under normal circumstances they never get above 39C.

FWIW, most drives I've interacted with actually run hotter when run out of a cabinet, presumably because of the lack of dedicated airflow.
Logged

Trumpetguy

  • Citizen of the Universe
  • *****
  • Posts: 974
Re:
« Reply #16 on: August 17, 2015, 09:08:28 am »

Ok, now I started worrying... Good thing all data are mirrored on a backup drive. Thanks for the input!
Logged

jmone

  • Administrator
  • Citizen of the Universe
  • *****
  • Posts: 14267
  • I won! I won!
Re: Is it dying?
« Reply #17 on: August 17, 2015, 04:48:43 pm »

Don't worry.  Most of the large studies have shown that "high" temps are not correlated to drive failure (though "cold" drive are). 
Logged
JRiver CEO Elect

RoderickGI

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 8186
Re: Is it dying?
« Reply #18 on: August 17, 2015, 06:30:46 pm »

In Summer my office where my PC resides can get up to the mid to high 30's °C. Most unpleasant. The drives can run hot at these times. I just checked and was surprised to see the highest temperatures recorded at only 43°C. But the specification for the Reds says their operating temperature is up to 70°C so I don't worry about it.

I have the temperature warning set at 43°C, and the alert at 50°C. The drives are all reported as 100% perfect.

But I may be asking WD about a 20°C temperature differential to the environment, under no load.
Logged
What specific version of MC you are running:MC27.0.27 @ Oct 27, 2020 and updating regularly Jim!                        MC Release Notes: https://wiki.jriver.com/index.php/Release_Notes
What OS(s) and Version you are running:     Windows 10 Pro 64bit Version 2004 (OS Build 19041.572).
The JRMark score of the PC with an issue:    JRMark (version 26.0.52 64 bit): 3419
Important relevant info about your environment:     
  Using the HTPC as a MC Server & a Workstation as a MC Client plus some DLNA clients.
  Running JRiver for Android, JRemote2, Gizmo, & MO 4Media on a Sony Xperia XZ Premium Android 9.
  Playing video out to a Sony 65" TV connected via HDMI, playing digital audio out via motherboard sound card, PCIe TV tuner
Pages: [1]   Go Up