mrbens Posted August 24, 2013 Share Posted August 24, 2013 Hi guys, Quick question: Today I replaced a 250GB drive with a 1TB that has failed with many errors within a few minutes of rebuilding. Am I safe to remove the 1TB, reinsert the 250GB, do initconfig and go back to how it was a few hours ago? I'm assuming this is the best course of action but thought I'd double check first. Thanks. syslog attached. Pro version: 5.0-rc16c ----------------- More info: The 1TB has had 212 reallocated sectors for a few years but the number hasn't risen. Strangely while preclearing it said 212 before and 211 after. -1? Tried unassigning it, starting array, stopping, reasigning it but after a few minutes rebuilding it had 288 errors again and red ball. I don't mind chucking it away, it has had a good innings over the years but time to retire! syslog-2013-08-24.zip Quote Link to comment
mrbens Posted August 24, 2013 Author Share Posted August 24, 2013 No replies yet so I'm just going to go ahead and revert the 1TB back to the 250GB, do initconfig, reassign all the drives the same way and rebuild parity. Quote Link to comment
garycase Posted August 24, 2013 Share Posted August 24, 2013 No, you can't do that. Now that the system has a 1TB drive in that slot, you'll need to replace it with at least a 1TB drive. Alternatively, you could replace the drive with your old 250GB drive and do a new configuration -- but this will lose all parity protection until you do an initial parity sync, so unless you're SURE the data is good, I wouldn't do that. Quote Link to comment
mrbens Posted August 24, 2013 Author Share Posted August 24, 2013 Alternatively, you could replace the drive with your old 250GB drive and do a new configuration -- but this will lose all parity protection until you do an initial parity sync, so unless you're SURE the data is good, I wouldn't do that. Hi Gary, that's what initconfig does. The parity is rebuilding now. The data on the 250GB will be exactly the same as when I unassigned it from the array. Basically once the parity has rebuilt and then been checked I'll be in exactly the same config as before I replaced the 250GB with the faulty 1TB. Incidentally this is pretty much the same way as permanently removing a drive from the array that you won't be replacing which I've done a few times to move drives to a second server. I've created instructions to do that as follows based on the info in the link: http://lime-technology.com/wiki/index.php/FAQ_remove_drive How to permanantly remove a disk from the array and rebuild parity: Copy everything off the disk that is to be removed. Take a screenshot of the server drive configuration. Backup flash. Stop the array and un-assign the drive on the Devices page. Log in on the system console or via telnet and type: initconfig This will create a new system.dat file and reset the array configuration data. unRAID will ask you to confirm you wish to set a new disk configuration. You must respond with Yes (capital Y and lower case es). When the initconfig command is invoked, old parity data will be immediately discarded. At this point, your array will not be protected from a disk failure until the system can complete the process of generating new parity information. Click 'Main' in the GUI and re-assign the disks as per the screenshot minus the removed one. Start the array and let the parity re-built itself. The only difference being changing a blank 1TB with my full 250GB drive instead of leaving a slot unassigned where the drive I removed was. Hope this info can be useful to others Quote Link to comment
garycase Posted August 24, 2013 Share Posted August 24, 2013 I knew that -- I missed that you mentioned "initconfig" -- was thinking about the data on the old disk that you had rebuilt. If you were interested in preserving the parity info to reconstruct that data, clearly you wouldn't want to do the initconfig. Quote Link to comment
mrbens Posted August 24, 2013 Author Share Posted August 24, 2013 No worries, thanks for the reply anyway The 1TB failed both times just a few minutes into the rebuild which will have messed up the parity. Quote Link to comment
garycase Posted August 25, 2013 Share Posted August 25, 2013 The 1TB failed both times just a few minutes into the rebuild which will have messed up the parity. Actually that's not correct. If you had good parity in the array, and started a rebuild, the parity will remain good even if you have to abort the rebuild and do it on a different disk. You didn't lose parity until you did the "initconfig" -- THAT does indeed reset the configuration and invalidates the existing parity. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.