guys im thinking suddenly today my files became somehow corrupt cause at first plex was showing errors, at first i thought maybe their something wrong with plex so then i tried watching them directly through file explorer and it failed then i tried copying those files which somehow now are unplayable and it failed to even copy those media files. file explorer is saying some unexpected network error occured when i try to copy them.

so what should i do, should trying running parity, please help

and just to make it clear not all my files are showing this strange problem some of them are still working fine.

unraid-diagnostics-20210315-2105.zip

Both disks are showing data corruption (disk1 also read and write errors):

 

Mar 14 13:59:36 Unraid kernel: BTRFS info (device md1): bdev /dev/md1 errs: wr 35, rd 175, flush 0, corrupt 57, gen 0
...
Mar 14 13:59:40 Unraid kernel: BTRFS info (device md2): bdev /dev/md2 errs: wr 0, rd 0, flush 0, corrupt 504, gen 0

This suggests a hardware problem, start by running memtest, then there are some recovery options here.

@JorgeBdoesn't parity helps with data restoration or parity does work that way or its cause both my disks are showing data corruption so my one parity  drive cant help me.

and what should be my next step if memtest can't find any errors cause so far it is showing zero errors.

and one more thing when i tried to launch memtest from unraid boot menu, it just rebooted my system so i downloaded passmark's memtest86 v9.0, i hope its good enough.

Edited March 15, 20215 yr by Max

6 minutes ago, Max said:

doesn't parity helps with data restoration or parity does work that way or its cause both my disks are showing data corruption so my one parity  drive cant help me.

Parity cant help with this since it will also be corrupt.

@JorgeB what should if memtest can't find any errors, cause now out of 4 it has already passed 2 without any error although now it is showing a note that ram may be vulnerable to high frequency row hammer bit flips, so this be relevant here.

Unfortunately memtest not finding errors doesn't guarantee that there's not a problem, only the opposite is true, data corruption like that is most often caused by RAM, so I would recommend using different one or at least running those an the standard JEDEC settings instead of the overclocked XMP profile.

@JorgeB sorry for disturbing you again and again but the wierdest thing just happened with me just a minute ago memtest86 finished without any error, so after that i checked and realized that my kingston hyperx ram does support xmp profiling so im running it on stock clock which is 1866 and then i booted unraid to try your data recovery suggestions but now all those files which were unreadable earlier, are now working fine.

so as i dont want to corrupt my data (specially if its not corrupt already) i immediately stopped the array.

so what should i do now ??

could i use my server normally thinking it was just fluck and it wont happen again or i should just stop thinking about using my server without new rams.??

Most likely some corruption is still there, but you can run a scrub to confirm.

@JorgeB disk1

Error summary:    csum=311
  Corrected:      0
  Uncorrectable:  0
  Unverified:     0

disk2

Error summary:    verify=27 csum=26789
  Corrected:      0
  Uncorrectable:  0
  Unverified:     0

 

Can scrub help me fixxing this @JorgeB

Nope, scrub on the array can only detect corruption, not fix it since there's no redundancy, there will be a list of corrupt files in the syslog, delete or replace them from backups.

5 hours ago, JorgeB said:

Nope, scrub on the array can only detect corruption, not fix it since there's no redundancy, there will be a list of corrupt files in the syslog, delete or replace them from backups.

okay thanks for all the help mate and i should definetly replace my ram right ??

1 minute ago, Max said:

and i should definetly replace my ram right ??

Probably a good place to start.

okay 

4 hours ago, JorgeB said:

Probably a good place to start.

okay i will get new rams tommorow but meanwhile i thought of replacing those corrupt files on disk as they were only two but now i can't write to disk1.

syslog shows this error

Mar 17 00:02:04 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:04 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:04 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:05 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:05 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:07 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:07 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:07 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:07 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:07 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:09 Unraid kernel: btrfs_validate_metadata_buffer: 3 callbacks suppressed
Mar 17 00:02:09 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:09 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:09 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:09 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:09 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:09 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:10 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:10 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:10 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:10 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:15 Unraid kernel: btrfs_validate_metadata_buffer: 21 callbacks suppressed
Mar 17 00:02:15 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:15 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:15 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:15 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:15 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:15 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:15 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:16 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:17 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:17 Unraid kernel: BTRFS error (device md1): bad tree block start, want 210731008 have 0
Mar 17 00:02:20 Unraid kernel: BTRFS: error (device md1) in btrfs_finish_ordered_io:2736: errno=-5 IO failure
Mar 17 00:02:20 Unraid kernel: BTRFS info (device md1): forced readonly

what can i do to fix this.

You should really avoid using the server until the data corruption issue is fixed, or will likely just corrupt more data, possibly an entire filesystem.

14 minutes ago, JorgeB said:

You should really avoid using the server until the data corruption issue is fixed, or will likely just corrupt more data, possibly an entire filesystem.

okay i stopped my array and i wont use it again and replace my ram first thing in the morning and then what would i need to do make my disk1 writeable again or its like if its ram then it will disk1 will become writeable again ??

Edited March 16, 20215 yr by Max

Boot and see how it goes, difficult to say for sure one way or the other, also not absolutely certain RAM is the problem, but it's the most common reason so it should be the first thing to try.

13 hours ago, JorgeB said:

Boot and see how it goes, difficult to say for sure one way or the other, also not absolutely certain RAM is the problem, but it's the most common reason so it should be the first thing to try.

Okay I have replaced my ram and I'm still getting that same error while writing to disk 1. Disk1 is still goes forced read only mode. What should I do next ?

There are some recovery options here:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=543490

 

8 minutes ago, JorgeB said:

There are some recovery options here:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=543490

 

Do I really need to use those, cause I still can normally browse/read any data of off it. I just can't right now. Can't just attach a external HDD and copy all my data through unassigned devices.

And again we still don't know if it's my ram or not cause we thought replacing might fix this but I'm still getting those errors.

Thing is ddr3 is bit more on expensive over here now. I paid like 70 USD for 8 gig stick and I will have to get one more to use my vm and I really don't wanna spend another 70 bucks on 4-5 year old system which I'm already planning on upgrading for a part that might be faulty.

@JorgeB hey another thing that i just noticed is that just now i created a new share and i set it specifically write to disk 1 and their in that share i was able to write data to disk1 but if did the same thing to my older share and tried to write to it, i get error stating that my disk is write protected.

Edited March 17, 20215 yr by Max

Besides some data corruption the filesystem on disk1 is also corrupt, so you should backup and re-format, if you still have read access no need to use the recovery options, unless you want to recover the corrupt files, since you might no be able to copy those.

39 minutes ago, JorgeB said:

Besides some data corruption the filesystem on disk1 is also corrupt, so you should backup and re-format, if you still have read access no need to use the recovery options, unless you want to recover the corrupt files, since you might no be able to copy those.

when i try to copy paste a file i get this error although i do end up with a copy of that file in my external hardrive.

Cannot chown target file "/mnt/disks/Elements/RARBG.txt" 
            Operation not permitted (1)

and when i try to copy a folder im getting this error again, it makes a folder in external hdd with same name but without its content.

Cannot chown target directory

 

That's likely the target filesystem no supporting permissions, what file system? Though data should still be copied correctly.

7 minutes ago, JorgeB said:

That's likely the target filesystem no supporting permissions, what file system? Though data should still be copied correctly.

its exfat and yeah data is copying normally but this way it will ages to copy all files.

Edit its exfat, i had another ssd under ud plugin with exfat i was getting same error then i tried with xfs and this time no error.

Edited March 17, 20215 yr by Max
figured it out.