krisha Posted February 21, 2012 Share Posted February 21, 2012 I'm new to unRAID - i like the way it is working, but I saw that parity is not checked on read operations. Maybe it's better to check parity also on read operations? This might prevent the manual 'check parity' process. I think ReiserFS does not provide any checksums, so how unRAID is able to detect an error in communication or a wrong bit on HDD? (Maybe this is a special case, it's also unsure how to handle such a error then, since the source of the error could be parity or the data drive; double read might get any hint). On the other hand the parity drive needs to run also on read operations and might slow down transfer. Quote Link to comment
S80_UK Posted February 21, 2012 Share Posted February 21, 2012 I don't mean to be unkind, but I think you need to learn a little about how drives work. Every block of data that is read also has local CRC checking within the drive. And the way that unRAID works, you cannot check parity by just reading one data drive and the matching parity - you would need to read all data drives. This would negate one of the key benefits of unRAID - that to read a file you only need to access the drive containing the wanted data. Quote Link to comment
krisha Posted February 21, 2012 Author Share Posted February 21, 2012 ok, thanks for clarification on that. So URE's are only a problem when rebuilding an array ;-) [edit: URE - unrecoverable read error :] Quote Link to comment
S80_UK Posted February 21, 2012 Share Posted February 21, 2012 No - TLA's are a problem when used without explanation URE - Unrecoverable Read Error...? They can happen, of course, and the parity system is there to allow data recovery after such an error, but they are rare. In the event that there is such an error, then unRAID will spin up the other drives and calculate the missing data from the rest of the data and the parity information. That of course takes a little time. But I must emphasize that such errors are rare. Most users will never be aware that such an error has occurred, and when they do, they usually indicate a bigger problem such as a drive dying, or a faulty cable, power connection, controller, etc. When that happens, the fact that access to the array is slowed down is a minor inconvenience compared to the alternative of lost data. Quote Link to comment
krisha Posted February 21, 2012 Author Share Posted February 21, 2012 Yes, when writing the initial post I had in mind the following articles I somewhen read. Just I did not know (or remember) that the disk will report this error. http://www.zdnet.com/blog/storage/why-raid-5-stops-working-in-2009/162 http://www.tomshardware.com/news/RAID-5-Doomed-2009,6525.html 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.