August 13, 201213 yr Author Thanks jonathanm but it seems complicated. Is there a way to compare hdds by putting them into an external enclosure and then plugging them into a Windows machine?
August 13, 201213 yr Author On a regular basis I output the directory of the films that are on my server to a TXT file. The TXT file contains the exact bytes size of each film. I did a before and after TXT file. I then used WinMerge which compares and finds the differences in the files and there were none. Is this enough for an integrity check?
August 13, 201213 yr Thanks jonathanm but it seems complicated. Is there a way to compare hdds by putting them into an external enclosure and then plugging them into a Windows machine? Yes, but it ends up being more complicated, because you have to install a program into windows that will allow it to see the reiserfs formatted hard drives. I have no personal experience with any of the available windows packages to do this, there are several that show up with a google search, along with a fair number of people having issues with them. Google "windows reiserfs" and see what you come up with. Your file size comparison is a good start, but doesn't verify anything except allocated size. The contents could have several flipped bits and you would never know. The only 100% comparison is a binary compare. Any other comparison is less reliable, to one degree or another. Md5 is 99.999999% or better, it's 100% in this context, I can't remember the statistical chance of a random error causing a md5 checksum collision. At this point in the game, unless you have life changing documents at stake, I'd call it good and move forward. Chances are, everything transferred fine.
August 13, 201213 yr Author At this point in the game, unless you have life changing documents at stake, I'd call it good and move forward. Chances are, everything transferred fine. Exactly what I thinking! Let me clarify this up: If the parity rebuild was a bit dodgy then would the disk1 rebuild be a bit dodgy? Would subsequent rebuilds of other disks be dodgy too?
August 13, 201213 yr If the parity rebuild was a bit dodgy then would the disk1 rebuild be a bit dodgy? Would subsequent rebuilds of other disks be dodgy too? Until you have a clean parity CHECK, all rebuilds are possibly suspect. As soon as you trust all the drives to read properly by completing an error free parity check, all subsequent rebuilds are as good as you can statistically get.
August 13, 201213 yr Author Ok. Thanks again. I am currently 50% through the parity check and zero errors so far. Will update at 100%!!
August 13, 201213 yr At this point in the game, unless you have life changing documents at stake, I'd call it good and move forward. Chances are, everything transferred fine. Exactly what I thinking! Let me clarify this up: If the parity rebuild was a bit dodgy then would the disk1 rebuild be a bit dodgy? Would subsequent rebuilds of other disks be dodgy too? Yes. A successful parity check is required after each rebuild.
August 14, 201213 yr Author I am now back to RC5 and did a parity check and the speed is back to 75MB/s! I am also doing a rebuild and the write speed is at 83MB/s. It was 50MB/s with test2. I had a look at the issues with RC5 and the mvsas drivers as I have 2 of them. I have never had a problem with my server locking up on spinning up/down issues. Is there anything to worry about?
Archived
This topic is now archived and is closed to further replies.