Franklin

I recently replaced the motherboard on my unRAID server. I initially was having difficulty with the system booting, which I believe was a result of a couple of PCIe 4 port SATA controllers. After reading through the storage devices and controllers forum, I replaced these with an LSI Logic SAS 9207-8i controller and a 2 port SATA controller with an ASMEDIA chipset. The system booted up promptly after this replacement and appears to be working normally. However I noticed an error message in the log page (see attached screen shot) that references a PCIe Bus Error. I am hoping someone in this forum can tell me the source of this error and, if needed, how to fix. Diagnostics are attached to this message as well. Thank you in advance for any assistance. Franklin tower-diagnostics-20180226-2136.zip

I don't mind buying new RAM, if that will help with the upgrade problem. This is a system that I purchased pre-assembled from Limetech probably close to 10 years ago. Still works great!

I believe 1 GB of RAM. From system log: Jan 26 14:07:15 TowerIDE kernel: Memory: 968160K/1046712K available (6119K kernel code, 614K rwdata, 2004K rodata, 916K init, 592K bss, 78552K reserved, 0K cma-reserved)

Apologize in advance that I have only rudimentary understanding of unRAID, but have been satisfied user for many years. Currently running unRAID 6.1.8. Have no plugins other than those built-in (Dynamix webGUI and unRAID). Everything has been working normally for more than a year (and still is, except for upgrade failure). Attempting to upgrade to 6.4.0. Plugin page shows this update as available, but when I try it fails (see attached screenshot). Tried to run diagnostics, but all I get is "404 File not found." Attempted manual upgrade by copying 6.4.0 bzimage and bzroot to flash drive, but could not get system to come online. Tried installing Fix Common Problems plugin, but get another error (see attached screenshot). System devices screenshot and portion of system log created after reboot and repeat attempt at upgrade attached as well. Any suggestions on how to successfully upgrade would be greatly appreciated. Partial system log.txt

Joe, to "set a new disk configuration" is the following the method for doing that? Telnet into server, login as root, type in: initconfig

How about this strategy to replace parity disk without losing data on disk 10? Stop current data rebuild (moving at snail's pace due to errors on parity disk, and probably inaccurate?). Unassign disk 10 and parity disk. Power down. Reinstall original disk 10 (which has the original data and hasn't been used for anything else) and new parity replacement. Power on. Reassign disk 10, Reassign parity disk, and start parity rebuild? Any better way of doing this?

Joe L. Thank you once again for your input (you have saved me many times before)! That drive is brand new, so aggravating to RMA it! Since the data rebuild on disk 10 did not complete (87.3% and stalled), how should I proceed? If I just cancel data rebuild, stop array, and replace parity drive won't data on disk 10 be lost? I do have the original disk 10 and could use it, but not sure how to proceed. Warm regards, Frank

I have an MD1200/IDE/Sata unRAID server built for me by Lime Technology several years ago. There is one parity disk, and 11 data disks. I am using unRAID version 4.7. I am attempting to upgrade to larger, Sata disks. I first replaced parity drive and everything went smoothly. I then replaced disk 10 with a larger Sata disk and started the data rebuild process. It seemed like everything was going well, but upon returning home tonight I found the process stalled. I followed instructions in this forum to get a syslog, which I believe suggests a problem with the new parity drive? I then obtained a smartctl report for the parity drive. Unfortunately, I do not understand these logs, nor do I know how to proceed from here. Any help is greatly appreciated! Addendum: My syslog is too large to attach, even zipped (it is 16MB)! It appears that this is secondary to numerous entries about read errors for disk 0 (the parity disk). Any thoughts? smart1.txt

I have an MD1200/IDE unRAID server built for me by Lime Technology several years ago. There is one parity drive and 11 data drives. After a recent power outage, a disk (#7) went bad and I replaced it with a slightly larger disk. The data rebuild began and appeared to finish. However upon reviewing the main web interface, I noticed that drives 10 and 11 had no temperature readings and several write errors. I rebooted, and the same drives were not found/missing. These 2 drives are both connected to the same port on a PCI IDE expansion card using a master/slave IDE connector cable. I replaced the cable thinking it might be the problem, and after a restart both drives reappeared. However, disk 11 has a red ball next to it. There is no longer any option to perform a parity check on the web interface. Movies stored on these 2 drives no longer can be played across my network. Any help in figuring out what might be the problem would be most appreciated. I will attach a syslog (which I have little hope of deciphering). syslog-2012-07-24.txt

Kevin, thank you for such an outstanding and detailed response - exactly what I was looking for! I ordered the sata Icy Dock racks and new hard drives today. The current system has only 3 TB, so the capacity will be quickly replaced. I may well end up upgrading the motherboard in the future, just as you did. Frank1940, I think your points are well founded. I actually have another, relatively new server that was built by Limetech as well, so the old server isn't something I really "need". I do, however like to tinker, and the case is nice looking [/img]. I hope to keep it around for a couple more years. The parity checks do take 3 times longer than my new server, but it hasn't really bothered me because I run them at night.

I have an MD1200/IDE unRAID server built for me by Lime Technology several years ago. Works great, but the 5 - 10 year old IDE drives are starting to fail, and finding replacements has become quite challenging (and expensive). My question is whether this server could be converted to use sata drives? The motherboard is an ASUS P5LD2-VM, (which does have 4 sata ports). The Coolermaster case has 12 drives with Icy Dock IDE Removable trays. Four drives connect to the motherboard IDE connectors, and the remaining 8 via two pci IDE expansion cards. If this transition is possible, what would be the best (and most cost effective) way of proceeding? Could it be done gradually, using both IDE and sata drives initially?

Joe L., Thanks so much for the help! I replaced disk11 and data rebuild is in progress at the expected speed. Looks like it was a failing drive. Is there an easier way to detect this kind of problem? By the way, I post infrequently but read the forums here regularly. You are an incredible resource for all of us using unRaid systems!

I apologize in advance for my limited knowledge base with Linux, but I was hoping for some help with what I fear is a problem with one of my unRaid systems. The unit in question was built for me by Tom, and has worked flawlessly for a couple of years. It uses only IDE drives. Today I noticed that it had spontaneously started a parity check, which was progressing very slowly. I stopped the array, powered down, and restarted. I then re initiated a parity check, which started out running usual speed, but quickly slowed to about 1500 KB/sec, much slower than usual. So far no parity errors reported. I viewed the syslog using instructions and the wiki, which I will attach below. I was hoping one of the experts here could review this to see if there might be an explanation for my trouble? Thank you in advance for any assistance. syslog3-11-11.txt