I recently posted an issue I was having with a share after replacing a failed drive.  Attempting to fix the issue, I discovered the disks were in read-only, and then learned that the parity drive was failing.  I have now replaced the parity drive, but I am still seeing errors on the original replaced drive.  I ran a resierfsck and then a --fix-fixable, and received the following at the end of the report:

 

Comparing bitmaps..Fatal corruptions were found, Semantic pass skipped

338 found corruptions can be fixed only when running with --rebuild-tree

 

I am assuming that this occured because the drive was rebuilt while the parity drive was failing (?).  I am assuming that my only option here is to go through with the --rebuild-tree and likely lose data.  This is OK, if it has to be that way, but since the documentation said do not go ahead with this without first posting to the forums, I am looking for confirmation.

 

With thanks,

 

Keith

I recently posted an issue I was having with a share after replacing a failed drive.  Attempting to fix the issue, I discovered the disks were in read-only, and then learned that the parity drive was failing.  I have now replaced the parity drive, but I am still seeing errors on the original replaced drive.  I ran a resierfsck and then a --fix-fixable, and received the following at the end of the report:

 

Comparing bitmaps..Fatal corruptions were found, Semantic pass skipped

338 found corruptions can be fixed only when running with --rebuild-tree

 

I am assuming that this occured because the drive was rebuilt while the parity drive was failing (?).  I am assuming that my only option here is to go through with the --rebuild-tree and likely lose data.  This is OK, if it has to be that way, but since the documentation said do not go ahead with this without first posting to the forums, I am looking for confirmation.

 

With thanks,

 

Keith

I would go ahead with the --rebuild-tree run.    In my experience the vast majority of the time that does an excellent job of recovering all your data (with the last files written most likely to be the ones lost if anything).  That seems to be one of the strengths of ReaserFS and reiserfsck - it is so good at recovering from corruption on disks.  Having said that assuming you have backups it is a good idea to check the recovery against them after it completes.

Thanks for the confirmation.

 

Seemed to work great, though for what it's worth, the rebuild would not run from the GUI, it said it would run and output to stdout, but the "DONE" button appeared immediately.  Not sure if this is a bug, feature, or something related to my issues.  Ran from SSH without issue though.

 

Thanks again,

 

Keith