unRaid Killed my ssd


Recommended Posts

Hi, I've just recently installed the 30 trial of unRaid onto my server. It is an dell precision T7500 dual Xeon X5670 24 gigs ecc triple channel memory not bad for an older chip set. I also boot with an USB directly on the motherboard. I have (3) 4TB HDD. And I just put in an older OCZ SSD I was going to run as cache drive or as a vm, i just never decided. Never had any problems out of my SSD it was mainly used for game drive where I stored all my games. But didn't really need it anymore in my main computer. Well this is where my problem started. I reformatted it and now its gone.  Please tell me this is not true and there is away to bring it back? I've tried everything from installing back into my main computer and downloading the OCZ utility but i don't see nothing. Nothing in Bios that i can recall. And with my luck my warranty is out. So I hope Someone can help. This will be very sad if i lost my SSD due to unRaid.:$

Edited by M4ST3R-OWNER
Link to comment

Sucks when that happens. We've got so much old parts that it's easy to take for granted that so many things last as long as they do. Out of the many OCZ drives I've owned, not a single one still works today.

 

In my experience this isn't the type of hobby in which one spills tears over empty drives that died. Save those tears for the day the day one dies with data that you don't have backed up.

Edited by Lev
  • Upvote 1
Link to comment

Whilst I'm sorry for your loss, if the drive isn't showing up in another computer then that signifies to me a hardware failure of some description, not convinced you can attribute that to Unraid.

If Unraid were a SSD Serial Killer then I have absolutely no doubt we'd have known about it quite some time ago....

Sent from my LG-H815 using Tapatalk

Link to comment

Can't speak to the drive dying, but will mention that using an SSD for caching writes and DL/repairing every DL, can put a lot of miles on the SSD in a hurry.

 

A fast unassigned device for DLing and allowing your writes to go straight to the array may be a better use, reserving the SSD for Docker and VM use. And on the VM side, you might consider disabling virtual memory and moving your TEMP directories to spinners to save wear and tear.

 

Make sure you are trimming the SSD from time to time. And checked the SMART attribute that indicates the amount of wear.

Link to comment
2 hours ago, bjp999 said:

Can't speak to the drive dying, but will mention that using an SSD for caching writes and DL/repairing every DL, can put a lot of miles on the SSD in a hurry.

 

A fast unassigned device for DLing and allowing your writes to go straight to the array may be a better use, reserving the SSD for Docker and VM use. And on the VM side, you might consider disabling virtual memory and moving your TEMP directories to spinners to save wear and tear.

 

Make sure you are trimming the SSD from time to time. And checked the SMART attribute that indicates the amount of wear.

 

I have a lot of experience over the years with this, but not enough data to support my theory which is quite similar to yours. Need more data.

 

 I've lost 4 SSDs over the years, each having 80-100TB written. A wide range of drives, see list below. All died far before their rated lifetime. Have you seen any other posts that discuss this further? I came to the same conclusion as you and stopped using a SDD for Downloads, Repairs and Caching, basically anything that uses it as a temporary scratchpad. All those operations have been moved to a unassigned spinning disk.

 

Docker allowed me to stop using VMs. In my experience I found containers that utilize a large number of small files, seem to also be contributors to problems on SSDs but I don't have enough examples yet. I've moved Docker and and all containers now to a spinning unassigned disk. Other devices on my network only perform read operations from UnRaid's array, so further reduces the need for any cache drive and mover. Life without a cache disk or SSD seems to be working out just fine for me, I don't perceive any performance difference. No longer having to deal with the mover has been a nice bonus.

 

120GB OCZ, 2x 256 740 evo and last month my prized 2TB Crucial MX300

  • Upvote 1
Link to comment
20 hours ago, M4ST3R-OWNER said:

I reformatted it and now its gone.  Please tell me this is not true and there is away to bring it back?

 

Apparently, I have a different take on this than everyone else. If you formatted the drive, then yes, all the data will be gone. I'm not reading into that, or anything that follows, that the drive itself is no longer appearing to the OS/BIOS.

 

Maybe you guys are just better at inferring intent than I am...

Link to comment
17 hours ago, 3605481mah said:

My first SSD was OCZ. It rests in the garbage after not a long time of use. Just for storage it was ok-ish, but it could not hold an OS.

Sent from my SM-G920F using Tapatalk
 

 

22 hours ago, Lev said:

Sucks when that happens. We've got so much old parts that it's easy to take for granted that so many things last as long as they do. Out of the many OCZ drives I've owned, not a single one still works today.

 

In my experience this isn't the type of hobby in which one spills tears over empty drives that died. Save those tears for the day the day one dies with data that you don't have backed up.

 

This. My first SSDs were OCZs, all three died within about 6 months of each other after maybe a year of use, one of the ones that died was actually the replacement for one that had died originally! Purchased a 60 GB for my netbook (yes, this was like 8 years ago hahaha) and found it worked out well and bought a 120 GB one, a few months later the 60 GB randomly died, I RMA'd it and then a few months later the 120 GB died I forgot about it and then was past the warranty for it or I decided to trash it and not bother with an RMA (I forget which), then a few months later the replacement 60 GB died so I just trashed it and went with Samsung and Crucial SSDs. I have had one or two Crucial M4s give up the ghost, but that was due to excessive writes, well it never technically died so I couldn't RMA it, it just kept warning that it's write cycles were like 96% used. I still have it in my parts bin, even though I'll never use it again hahaha

Link to comment
22 minutes ago, brando56894 said:

 

 

This. My first SSDs were OCZs, all three died within about 6 months of each other after maybe a year of use, one of the ones that died was actually the replacement for one that had died originally! Purchased a 60 GB for my netbook (yes, this was like 8 years ago hahaha) and found it worked out well and bought a 120 GB one, a few months later the 60 GB randomly died, I RMA'd it and then a few months later the 120 GB died I forgot about it and then was past the warranty for it or I decided to trash it and not bother with an RMA (I forget which), then a few months later the replacement 60 GB died so I just trashed it and went with Samsung and Crucial SSDs. I have had one or two Crucial M4s give up the ghost, but that was due to excessive writes, well it never technically died so I couldn't RMA it, it just kept warning that it's write cycles were like 96% used. I still have it in my parts bin, even though I'll never use it again hahaha

Brother... you and I have lived the same life. Wow I know exactly your struggle cause I lived it too. Thanks for reminding me of my past, but ugh is it painful. As we both know.

  • Upvote 1
Link to comment

Hahaha the struggle is real! Ever since then I've only ever purchased one other OCZ product and it was one of the first PCI-E SSDs (not NVMe), which had two SATA SSDs RAID0'd together via a controller built into the board. The site said that it supports Linux and Windows, which was great because I planned on using it in Linux....but they neglected to mention that it only worked in Ubuntu and RHEL/CentOS because it required binary blobs for the drivers and I was using Arch. Some Canadian guy made a hacky driver for it, but that made it show up as 2 single SSDs in Linux, while Windows and the BIOS still saw it as one device. After about 1.5 months I sold it on Ebay for like half the price because I couldn't return it to NewEgg. What a waste of $450! That was years ago and I have never bought a single OCZ product since then.

 

I've largely stuck with Samsung for my SSDs and WD for my HDDs.

Edited by brando56894
Link to comment
24 minutes ago, brando56894 said:

Yep, my Samsung 840 PRO has been powered on for a total of 1.3 years (9865 hours) and I've written a total of 48.27 TB to it, and there is not a single smart error B|

Question: How can you check the TBW of 48TB?

My SSD shows me TLC writes 32mib: 1046176 - what does this mean?

Edited by Zonediver
Link to comment

Hello thanks for everyone reply. Hopefully this will give more information. After looking all over the internet I saw a few people talk about a boot-lock. And from what I read if you do a reboot every so often it could unlock it i guess. Well after doing more research i decided to keep looking at not just bios but my device manager and what you know i have an unknown device. so i open disk manager and i see disk 2 with a red upside down arrow saying Unknown disk not initialized. well if you right click it ask to initialize but it just gives me error. so im still working to bring my ssd back to life. I see none of you never had this problem before but hopefully with my help i will help someone else down the road. 

Link to comment
15 minutes ago, Zonediver said:

Question: How can you check the TBW of 48TB?

 

I'm not sure what you mean by "check the total bytes written of 48 TB" but I'm going to assume you mean "how do I find out how much data was written to me drive over its lifetime".

 

I had to look it up because I couldn't remember how I found it before, this isn't the method that I used initially but the one I just used for my above values: http://www.virten.net/2016/12/ssd-total-bytes-written-calculator/

 

smartctl -a `mount|grep cache|grep dev|awk '{print $1}'`|grep Total_LBAs_Written|awk '{print $10}' will give you the value (of the cache drive) to plug into the above calculator.

 

12 minutes ago, M4ST3R-OWNER said:

I see none of you never had this problem before but hopefully with my help i will help someone else down the road. 

 

Some of us have had this issue, and there is no way to fix it. 99% of the time the controller is dead on the SSD, and you said it was an OCZ SSD which three of us have confirmed are absolute garbage. Cut your losses, toss it in the garbage and get a new one that isn't made by OCZ for like $50-$200 haha

Edited by brando56894
Link to comment
10 minutes ago, brando56894 said:

Some of us have had this issue, and there is no way to fix it. 99% of the time the controller is dead on the SSD, and you said it was an OCZ SSD which three of us have confirmed are absolute garbage. Cut your losses, toss it in the garbage and get a new one that isn't made by OCZ for like $50-$200 haha

i very rarely give up just because it broke. and yes i have another on the way but i would appreciate if you have no help advice just move on Thanks. ;)  

Link to comment
19 minutes ago, M4ST3R-OWNER said:

i very rarely give up just because it broke. and yes i have another on the way but i would appreciate if you have no help advice just move on Thanks. ;)  

 

Just trying to save you from wasting time and beating your head against the wall wondering why you can't get it to work. I've seen that error many times in Linux (been using it for nearly a decade) and it's always an error in communication with the drive, if switching SATA ports and SATA cables don't work, there's pretty much no hope for it because the controller is dead. If you want to keep wasting time on a dead drive be my guest!

 

18 minutes ago, Zonediver said:

When i put this "strange" command in, i get this: >

Thats all...

 

Make sure you copied it correctly, that's the $PS2 prompt waiting for more input, and usually happens when you miss a closing quote/bracket/parenthesis. Otherwise, I guess your columns and stuff are setup differently than what I have, because I just pasted it into my terminal and it works as expected. That's just a bunch of command substitution so you don't have to type out various commands and copy the values, here it is in piecemeal form:

 

root@unRAID:~# mount|grep cache|grep dev
/dev/sdb1 on /mnt/cache type xfs (rw,noatime,nodiratime)
root@unRAID:~# smartctl -a /dev/sdb|grep Total_LBAs_Written
241 Total_LBAs_Written      0x0032   099   099   000    Old_age   Always       -       103659642080

 

You want the value in the last column on the last line (103659642080 in this case)

Edited by brando56894
Link to comment
5 minutes ago, brando56894 said:

 

Just trying to save you from wasting time and beating your head against the wall wondering why you can't get it to work. I've seen that error many times in Linux (been using it for nearly a decade) and it's always an error in communication with the drive, if switching SATA ports and SATA cables don't work, there's pretty much no hope for it because the controller is dead. If you want to keep wasting time on a dead drive be my guest!

 

 

I guess your columns and stuff are setup differently than what I have, because I just pasted it into my terminal and it works as expected. That's just a bunch of command substitution so you don't have to type out various commands and copy the values, here it is in piecemeal form:

 

 


root@unRAID:~# mount|grep cache|grep dev
/dev/sdb1 on /mnt/cache type xfs (rw,noatime,nodiratime)
root@unRAID:~# smartctl -a /dev/sdb|grep Total_LBAs_Written
241 Total_LBAs_Written      0x0032   099   099   000    Old_age   Always       -       103659642080

 

 

You want the value in the last column on the last line (103659642080 in this case)

 

I think, my SSD doesnt have this value - 241 is (in my case) Host writes 32mib

The first command is working and shows me the dev (sdp) but the second command gives me nothing - just an empty line

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.