January 22, 201016 yr Brand new unRAID server, only 1 parity, 1 cache, and 2 data discs. All 4 discs were scanned thoroughly with HD tune before putting into unRAID . I transfered about 1TB of data to the server, with 500GB left the parity showed 1119 errors. The file transfer failed (was going 594k/s, but it wasn't updating). I rebooted unraid and now it says the parity disk is uninstalled.
January 22, 201016 yr Author I just took the drive out and put it into Windows, no SMART errors, windows detects and formats the drive correctly. What gives? Bad SATA Cable?
January 22, 201016 yr We'll never know unless you saved the syslog from your first attempt at installing a parity drive. In any case, it sounds as if you had either a bad SATA cable, or a loose SATA cable, or a loose/bad power cable. Please provide a SMART report on the drive too once you put it back into the unRAID server. Basically, to be able to provide any help, you need to supply a syslog. Instructions are in the wiki.
January 22, 201016 yr Author Nope, I didn't grab the syslog before rebooting, i'm brand new to unRAID and didn't know it cleared it after reboot. I do have more info though. I replaced the SATA cable, and it now detects my drive. However, I switched back to the old SATA cable and it (still) detects my drive. I verified that the cables were in all the way, so it couldn't of been a loose cable.
January 22, 201016 yr Nope, I didn't grab the syslog before rebooting, i'm brand new to unRAID and didn't know it cleared it after reboot. I do have more info though. I replaced the SATA cable, and it now detects my drive. However, I switched back to the old SATA cable and it (still) detects my drive. I verified that the cables were in all the way, so it couldn't of been a loose cable. Did you verify that even though it was originally inserted all the way it was making good electrical contact? It is as likely it was the power cable... regardless, it is good the server sees the drive now. Are you looking for guidance now? or are you ok and proceeding with a new calculation of parity? The syslog is written to memory... The whole of unRAID runs in a in-memory file-system. The syslog is lost every time you reboot unless you save it before you reboot. Instructions to grab a copy of the syslog are in the wiki under troubleshooting if you need help. Joe L.
January 22, 201016 yr Author I'm going to remove the parity drive until I get all my data transfered, then add it and parity sync. I'll see how it goes from there.. Thanks again Joe.
January 22, 201016 yr I'm going to remove the parity drive until I get all my data transferred, then add it and parity sync. I'll see how it goes from there.. Thanks again Joe. just be aware that your data will not be "protected" from a disk failure until the parity drive is installed and parity copied. If you intend to reuse any of the existing drives you'll want to put the parity drive in place before destroying the original copy on the old drive. Joe L.
January 22, 201016 yr Author I'm going to remove the parity drive until I get all my data transferred, then add it and parity sync. I'll see how it goes from there.. Thanks again Joe. just be aware that your data will not be "protected" from a disk failure until the parity drive is installed and parity copied. If you intend to reuse any of the existing drives you'll want to put the parity drive in place before destroying the original copy on the old drive. Joe L. Will do. I will post a syslog after the parity check completes. I have another 4ish hours before everything is transfered over, then the parity sync.. which could take awhile. So.. some time tomorrow I'll post my log.
January 23, 201016 yr Tyrindor, a suggestion if I may... It is usually a good idea to burn-in a new disk before trusting that it is good to assign it in your array. And that goes for both data and parity disks. Have a look at this thread: http://lime-technology.com/forum/index.php?topic=2817.0
January 23, 201016 yr Author Tyrindor, a suggestion if I may... It is usually a good idea to burn-in a new disk before trusting that it is good to assign it in your array. And that goes for both data and parity disks. Have a look at this thread: http://lime-technology.com/forum/index.php?topic=2817.0 I use HDtune in windows to write zeros to the drive, and then scan for errors. It takes about 16 hours. I don't get what that would do differently? All these drives had zeros wrote to them.
January 23, 201016 yr Author Heres my syslog. I just filled up the only 2 drives in my system, then added the parity drive and cache drive and did a parity sync. 0 errors found. I'm using the same cables as before, so I have no idea what caused those errors. See any errors? I have no ideas what any of this means.. but heres some stuff I found that might not be normal. Jan 22 16:47:43 Server kernel: Phoenix BIOS detected: BIOS may corrupt low RAM, working around it. Jan 22 16:47:43 Server kernel: ACPI Warning: \_SB_.PCI0._OSC: Return type mismatch - found Integer, expected Buffer 20090521 nspredef-940 Jan 22 16:47:43 Server kernel: _OSC evaluation returned wrong type Jan 22 16:47:43 Server kernel: _OSC evaluation returned wrong type syslog-2010-01-23.txt
January 23, 201016 yr Likely crappy ACPI in BIOS. May need to boot with acpi=off See this: http://lime-technology.com/wiki/index.php?title=Boot_Codes
January 23, 201016 yr Author Likely crappy ACPI in BIOS. May need to boot with acpi=off See this: http://lime-technology.com/wiki/index.php?title=Boot_Codes Can I just turn ACPI off and not worry about it affecting my current data? Is it possible that this has corrupted some of my current data? This mobo has been used quite a bit for unRAID, haven't seen anyone else having to do that. My mobo: http://www.newegg.com/Product/Product.aspx?Item=N82E16813182142
January 23, 201016 yr Can I just turn ACPI off and not worry about it affecting my current data? acpi=off will not affect your current data. One minor inconvenoence will be that you won't be able to completely power down your server through software. You'll have to go there and press the power button after the system is halted.
January 23, 201016 yr You can also look to see if a BIOS update is available... or... you can ignore the errors about the buggy BIOS in the syslog. Linux is detecting them, and dealing with them.
January 23, 201016 yr Author You can also look to see if a BIOS update is available... or... you can ignore the errors about the buggy BIOS in the syslog. Linux is detecting them, and dealing with them. Latest BIOs, first thing I did. Everything else in my log is fine though, correct? I just took a look at the level 1 testing on this motherboard. That guy's log has the same ACPI errors, so it makes me feel better.
Archived
This topic is now archived and is closed to further replies.