Two disks with error 187 reported uncorrect

[xxredxpandaxx]

One has 7 the other 8. No reallocated sectors. Both are under warranty still and I was wondering if they would replace the drives with this error. From what I have read this error doesn't mean it's bad but they will likely go bad soon? Oh also they are seagate drives.

 

 

Sent from my iPhone using Tapatalk

[Frank1940]

You don't have reallocated sectors only because those sectors have not had an attempt made to write data to them.  If that write should fail then the sectors would be reassigned with sectors from the reallocation pool.  You are living on the edge of disaster in that if you have two drives that can't be read, neither disk can be rebuilt.  You could easily lose one or more files from each disk.  You have to proceed carefully at this point!

 

I would be looking at trying to copy all of the files off of those two disks. 

[xxredxpandaxx]

26 minutes ago, Frank1940 said:

I would be looking at trying to copy all of the files off of those two disks. 

 

Sadly they are both the biggest drives in my array, both are 8TB drives and both are 100% full (awesome luck). I am in the process of backing up everything to google drive but my upload speed is maxed out at 10 Mb/s so it will take 40 years to finish. If I replace one of those drives will I lose only the data that it can't read from the other disk? I guess my only other option is to buy a new 8tb drive to copy all the data to, then get a replacement for that drive and copy all the data from the other bad drive to the new replacement? I would like to not have to pay $250 for a drive I don't really need at the moment.

[Frank1940]

If you had had only one drive with issues, you could have tried a rebuilt on that drive but with two the likelihood of data loss is large.  If you can get all the data off of one, you would  have more options to try and savage one or both drives.  Did this just happen within a short period of time (like a good parity check on the first of this month) or is a case of benign neglect?

[xxredxpandaxx]

Both showed up in my last parity check at the beginning of this month. The month before I just got don't replacing two 3tb drives that were going bad. one went bad then during the rebuild the other started to go. I had a small amount of data loss but it was all stuff I could replace. That is when I started backing up to the cloud. I guess this is what happens when you buy a few drives at the same time, multiples go bad at the same time.

[SSD]

I would not freak with this situation. It is not optimal to have smart errors like these, but the chances of true data corruption or loss are astronomically low. If a read error occurs, unRaid will attempt a reconstructed read (from all other disks) to determine the proper contents, and use that to update the disk to allow it to remap the sector. The 2 disks would have to have the exact same bad sectors for this to fail! With upwards of 10 billion sectors, your chances of having issues on corresponding sectors are infinitesimal.

 

If it were me, I would run 3 parity checks. If you see this attribute growing over the course of these checks, I would look to replace the drive. If they stay constant, I'd leave them in but continue to monitor with each monthly parity check, at least for a few months. If you see it getting worse, I would pull the trigger.

 

The problem with warranty replacement if you often get a refurbished disk. They might be fine but could also have nasty problems that can be hard to detect. In other words, could be worse than what you have.

 

If losing these 16T would have you considering suicide, because you would have no reasonable way to substantially recreate that or similar content by re-ripping, re-downloading, or re-recording, then you need a backup, regardless of the condition of these 2 disks. All of your eggs in one basket is a bad idea.

 

Suggest posting a diagnostics file which will contain the smart reports and allow forum members to have a look. If nothing else, would be nice to know what model disks and see if other users are seeing something similar.

 

Good luck! 

[xxredxpandaxx]

Thanks for the info. I will probably run a couple parity checks to see if they get worse or not. The only problem I have with leaving them is that all of my disks will have to spin up to be able to read from any of the bad spots correct? This seems like a lot more wear and tear on my other disks if I could just get them replaced to avoid that. Also as long as I go through the stress testing on the replacement drives they shouldn't they be ok? I haven't had any problems with my 3tb replacement drives I got like 2 years ago. Also I want to add a second parity drive soon so I kind of need all my drives to be good before I do that.

[SSD]

45 minutes ago, xxredxpandaxx said:

Thanks for the info. I will probably run a couple parity checks to see if they get worse or not. The only problem I have with leaving them is that all of my disks will have to spin up to be able to read from any of the bad spots correct? This seems like a lot more wear and tear on my other disks if I could just get them replaced to avoid that. Also as long as I go through the stress testing on the replacement drives they shouldn't they be ok? I haven't had any problems with my 3tb replacement drives I got like 2 years ago. Also I want to add a second parity drive soon so I kind of need all my drives to be good before I do that.

 

A few reconstructed reads will have the tiniest of impact. But if you see a pattern of read errors, you might consider stopping it, knowing the disks are failing. You are trying to decide if there is a serious problem or not. Normally escalating reallocated and pending sectors are the "gold standard" of problem disks. These other attributes are a little less definitive.

 

I found the following about issues with this attribute in Acronis website:

 

"Although this parameter is not considered critical by the most hardware vendors, degradation of this parameter may indicate electromechanical problems of the disk. Regular backup is recommended. If no other (critical) parameters report a problem, hardware replacement is recommended on mission critical systems only."

 

 

[Frank1940]

53 minutes ago, bjp999 said:

Although this parameter is not considered critical by the most hardware vendors, degradation of this parameter may indicate electromechanical problems of the disk. Regular backup is recommended. If no other (critical) parameters report a problem, hardware replacement is recommended on mission critical systems only."

 

 

I don't know about other folks on this forum but I consider my servers to "mission critical"  because I have to be able to read EVERY sector on EVERY disk except the one that I am trying to rebuilt!  Sure, I can use my backups to recover my irreplaceable data but the media stuff is a long drawn out process that takes a lot of time...

[SSD]

I consider the words 'mission critical' to be related to core operations of a business entity. With mission critical systems, time is money, uptime is everything, and even a short downtime is unacceptable.

 

Although my data is important to me, I would not think it meets that standard. If I had to power it down for a week awaiting a new motherboard, there would be no huge consequence. My mission critical systems are all biologic.

[xxredxpandaxx]

Well I finally got around to doing two parity checks, and both disks still report the same amount. so it seems to be stable correct? My concern is that if I leave it how it is now both disks will fail later on down the line and I could loose lots of data. So is it better to replace them now while there are only a couple sectors acting up so that I can minimize the data loss? My other option is to wait until I finish my backup of all of my data which will take around 6 months. Then replace my drives so that I can download any files that get corrupted do to the bad sectors. Do you think the disks will be fine for another 6 months / would you risk it?

Edit: That would be 6 months of 24/7 reading/uploading.

[Frank1940]

Depends on your tolerance to risk.  I googled   SMART attribute 187   and there are a lot of hits.  Read some of them and form your own opinion.  It looks like the situation is currently stable which is a good thing.  The fact that you run two parity checks without error indicates that you could have successfully replaced both of these disks had you been doing that rather than running those checks!  Whatever you decide to do, you definitely want to continue to monitor these two disks closely!  You might also look at how old the disks are.  (I would be more  likely to do a four+ year old disk that an eighteen month one.) 

 

If you decide to replace those two disks, be sure to remove one and set it  aside and rebuild it on a new one.  (I, personally, would have run three preclear cycles on the replacement first!)  That way if something does go badly, you should be able to read most of the files off of that old disk. 

Then do the second disk.