November 7, 20169 yr Hi guys - my monthly parity check finished this morning with some warnings. **** Event: unRAID array errors Subject: Warning [MBFS01] - array has errors Description: Array has 2 disks with read errors Importance: warning Disk 21 - WDC_WD40EFRX-68WT0N0_WD-WCC4E1419057 (sdb) (errors 1772) Disk 22 - ST3000DM001-9YN166_W1F0ZLEC (sdf) (errors 114) and also Event: unRAID Disk 22 SMART health [187] Subject: Warning [MBFS01] - reported uncorrect is 1 Description: ST3000DM001-9YN166_W1F0ZLEC (sdf) Importance: warning ********* I don't want to replace those disks as I have plenty of room on the array on some newer drives. Am I correct in thinking I need to do the following. 1 - Move the content of those to drives to a new drive which has plenty of room. 2 - Stop the Array 3 - Do Tools new config. 4 - Remove the drives I am no longer using. 5 - Re configure the array. 6 - Start the array and let is build the new parity for the array which no longer has those drives. Thanks in advance! Stuart
November 7, 20169 yr Community Expert Read errors are often not due to a disk problem. Post complete diagnostics zip.
November 7, 20169 yr With a single parity system you definitely don't want to take a chance on having two disk failures ... so I would indeed be sure you have backups of the data on these drives before doing anything else. Errors without the disks being disabled means UnRAID was able to recover from the errors by emulating the data using the other disks, and then successfully wrote it back to the disk. HOWEVER ... the fact you're getting the errors is a bad sign => it could indeed mean you have a connection or cabling issue; but it could also mean the disk is having issues. Before "fiddling" with the connections/etc. I'd be sure the data is backed up. Copying it to a different disk in the array is a reasonable way to do this, as long as you avoid the "user share copy bug". The easiest way to ensure you don't take a change with that is to copy the data to a DIFFERENT share (You can later rename the root folder to match the share you actually want the data on after you've removed the data from the failing disks). If you simply don't need these disks anymore, the process you outlined is fine -- just be sure you understand the "user share copy bug" issue and absolutely avoid it.
November 8, 20169 yr Author Hi guys - thanks for info. The parity check completed successfully ******** Event: unRAID Parity check Subject: Notice [MBFS01] - Parity check finished (0 errors) Description: Duration: 21 hours, 28 minutes, 21 seconds. Average speed: 103,5 MB/s Importance: normal ****************** I got the emails in my original post about the warning during the parity check itself which then completed successfully. I am just worried that those disks might be getting ready to fail, especially with the reported uncorrected smart error. In terms of connections for the disks - they are all connected via the back plane so no individual leads go to disks. Have attached diagnostics. Thanks so much for advice. Stuart mbfs01-diagnostics-20161108-1040.zip
Archived
This topic is now archived and is closed to further replies.