SOLVED HDD is degraded with no SMART Errors

[joeschmuck]

Your drive is an SMR drive, that is why this is happening. ZFS doesn't work well with SMR drives and it's well documented in the forum. Your signature line is wrong and does not list this drive model otherwise I would have caught this when I first started this troubleshooting with you because I looked up the drives you had in your signature, those are all CMR drives.

Replace the SMR drive with a CMR drive, that is the fix. You should verify all your drives are CMR.

 

[Stux]

Well, once you've replaced the SMR drive with a CMR drive, you still need to transfer your data to a new pool made with some redundancy before a drive fails again and you lose everything.

 

[joeschmuck]

They need to add a MIRROR CMR drive to the SMR drive, once all the data had been resilvered, remove the SMR drive. It's a nice little trick. I have never done that myself, Hum... If I only had a 6TB HDD that I could test that with, then I'd feel better about offering it up as advice.

 

[Stux]

They need to add a MIRROR CMR drive to the SMR drive, once all the data had been resilvered, remove the SMR drive. It's a nice little trick. I have never done that myself, Hum... If I only had a 6TB HDD that I could test that with, then I'd feel better about offering it up as advice.

whoops. I forgot about that.

And I have done it. Not sure if it can be done via the UI though.

 

[joeschmuck]

Not sure if it can be done via the UI though.

That is a good reason for me to try it out. Hum... I could make a VM with some VM drives and test it that way. Much less risk as well. Then delete the VM and back to normal.

 

[joeschmuck]

whoops. I forgot about that.

And I have done it. Not sure if it can be done via the UI though.

So I tested it this morning on a VM with three drives (da1, da2, da3) as a STRIPE, put some data on it, then did a 'REPLACE' (da1 with da4) via the GUI. The screen flashed and automatically mirrored the old drive, then it removed it (da1) and left the new drive (da4) intact. I conducted a scrub both before and after, all is good.

I didn't think that would happen to be honest. And this was on CORE 13.0-U4 (that is what I had downloaded at the time). So you can replace a disk in a Stripe and not lose data, provided the drive you are replacing isn't dead or corrupt. Interesting. There is hope for someone to replace an SMR drive in a Stripe.

 

[Ashvath Sridharan]

Ok so how do I find out if my other drives are SMR? And how does having a SMR drive make a difference, sure the drive being SMR is giving lot's of issues but was there any sort of data corruption or loss at all? From what I have been seeing, I don't see any corrupted files at all. It would be helpful if you could tell me the capacity of the drive or the serial number of the drive that is SMR so I can replace it.

 

[joeschmuck]

Ok so how do I find out if my other drives are SMR?

Do an internet search for the drive model and SMR CMR. For the drive above I simply went to Google and entered "ST500DM002 SMR CMR" and looked at what came up. Pretty simple.

And how does having a SMR drive make a difference, sure the drive being SMR is giving lot's of issues but was there any sort of data corruption or loss at all? From what I have been seeing, I don't see any corrupted files at all.

I will not get into the topic of how ZFS does not work well with SMR drives, that is all over these forums and even the internet. Search google for "truenas smr" and you will find a lot to read.

It would be helpful if you could tell me the capacity of the drive or the serial number of the drive that is SMR so I can replace it.

I don't understand your question. You have only listed one drive on this forum (ST500DM002), it is the only drive I have identified as SMR. The serial number you posted as well.

I'm not trying to be a jerk but you need to put forth some effort yourself if you plan to run TrueNAS, or any home made NAS. You need to read the documentation, read the data you are posting, do searches for stuff before just asking someone to give you the answer. I have given you enough information that you can identify which of your drives are SMR or CMR, you do that by the model number, which you can read on the drive label or from the 'smartctl' output, including the serial number.

If you still need additional help, please ask but try to find the answer yourself first. I am more than willing to help but if you learn how to do this stuff, then you can be self-sufficient. The words "Teach an man to fish" come to mind.

 

[Ashvath Sridharan]

Sorry about the silly questions I asked, the reason why asked those questions is because I am now having 2 errors from 2 disks, one error is that the 500GB HDD is SMR and the other error I found was when I looked into the notifications pane in the Dashboard and it says that my other 1TB HDD is failing with read and checksum errors and on top of that reading it says that due to the failing 1TB HDD, the pool is degraded and applications may be affected. Is that the reason why my pool is degraded? And for the SMR drive, I am ready to replace that with a 500GB SSD, is that fine? I looked arround in Google but I am not getting an answer whether my SSD is SMR or CMR. I am sorry if this is very time consuming for and you are wasting your time on someone who has a server with lots of errors but I really do need some help, am I not framing my questions correctly? Is that the reason why I am unable to get the answer I am looking for? I was using these HDD for almost 2 years and I've never had such issues with them until now.

 

[joeschmuck]

SSD's are not SMR, ever. SMR is a recording technology for hard drives.

I was using these HDD for almost 2 years and I've never had such issues with them until now.

SMR drives work fine until they need to rewrite data, and that problem occurs after the drive has written over itself at least once. You are not the first to have this problem a few years after setting up your NAS.

 

[Ashvath Sridharan]

Just found a new 1TB Surveillance HDD laying around, I changed the HDD using the replace option in the UI and now my data is back to normal, but one issue I came accross was that my SMB share configurations were messed up, I tried looking around in Google and eventually just redoed my ACLs and other SMB configs again and still the issue. Upon creating the share, I noticed only one of my pool Sridharan-Data was being shared and I can connect to it via file explorer in Windows but when I create a share for my other pool Media, the share just redirects me to my other Sridharan-Data pool and when I verify that I chose the Media pool, it shows the exact same data as from Sridharan-Data pool. I know for a fact that my data is still intact and I can still see and play my movies, TV shows and other media because I have 2 apps that are using my pool, Plex and a Download Manager and they seem to be able to access the media without any issues (I can play the media via plex and I can download media via my download manager). It is just my SMB share configuration that is messed up I guess.

 

[joeschmuck]

That is a very odd thing, to have your SMB messed up just for replacing a drive, and ACLs as well. They really do not seem to be related at all with a drive replacement. Next time you have something strange like this happening (hopefully never again) then just restore your backup TrueNAS configuration file (something you should backup). That should restore your configuration.

 

[Ashvath Sridharan]

That is a very odd thing, to have your SMB messed up just for replacing a drive, and ACLs as well. They really do not seem to be related at all with a drive replacement. Next time you have something strange like this happening (hopefully never again) then just restore your backup TrueNAS configuration file (something you should backup). That should restore your configuration.

I found a backup file for my NAS when I moved from the old PC to the new one, but now my storage is degraded again and it is unable to show the pool space it says unknown and under pool status it says UNAVAIL. I tried restarting several times but no luck.

 

[joeschmuck]

I found a backup file for my NAS when I moved from the old PC to the new one, but now my storage is degraded again and it is unable to show the pool space it says unknown and under pool status it says UNAVAIL. I tried restarting several times but no luck.

I suggest that you stop using old hardware and install new CMR drives, otherwise you are just looking to have bad experiences.

 

[Ashvath Sridharan]

Thanks for your explanation of how to check whether if my drives are also CMR, after checking them I found that 3 hard drives were CMR and I ordered 3 new CMR drives. Now the big issue occurs, after the replacement of the 500GB HDD, I still don't know whether if my data is still intact because of ZFS saying pool status and pool space is UNKNOWN and UNAVAILABLE, so I just want to verify whether if replacing a HDD via the GUI does work and keeps my data intact while doing this in a Striped RAID configuration. The other problem is that due to this whole ZFS status thing not giving any answer and my applications and SMB shares are not working, this basically means that this server is definitely broken. So I was just thinking whether if I could like start fresh with TrueNAS Scale and configure everything properly one by one while having my data still intact and if I can then can you send me the steps or a guide to do so? If it can't be done then I understand and I can just reset everything and start fresh but like I said from the start, I am not yet ready to lose my data.

 

[joeschmuck]

No one can tell you if your data is intact unless you run a Scrub, but you should not run a scrub until you have all CMR drives installed. Or you can copy all your data off to another hard drive/server. Reading an SMR drive is not an issue, only writing.

 

[Ashvath Sridharan]

No one can tell you if your data is intact unless you run a Scrub, but you should not run a scrub until you have all CMR drives installed. Or you can copy all your data off to another hard drive/server. Reading an SMR drive is not an issue, only writing.

How do I copy when my SMB shares isn't working?

 

[joeschmuck]

How do I copy when my SMB shares isn't working?

I honestly do not know why your SMB is not working. You are having some very strange problems and none seem to be related.

Here is something you can try. Install a large capacity hard drive and create a single stripe from it (do not add it to your current pool), then copy all the files over manually. This would by far be the fastest, however you need to know how to use the CLI and the copy commands. That is the only safe thing I can suggest.

 

[Ashvath Sridharan]

After digging through more articles about my issue, I tried zpool status again but the status shown for the pool is SUSPENDED and the reason being that one 500GB HDD, due to the HDD failing, the pool suspended itself if I am correct, I haven't considered buying a 6TB HDD just yet as I am hoping to just replace the failing HDD with a 500GB SSD and that brings me to this question, can I just replace the HDD with an SSD given the fact that the HDD is the only issue so far or do I have to get a 6TB HDD? The failing HDD has been kicked out of the pool so I don't have an option to replace it via the UI so I was hoping whether if you know any CLI command to make this work

 

[joeschmuck]

You must replace a drive (any drive) with one that has equal or greater capacity.