dpc2

Got it. Thanks very much for your help. Not the first time and probably not the last!

Syslog attached. Also looks like I have just 1GB of RAM. Could have sworn I had more! syslog_DPC_MAR_16.zip

Will check the memory (I think it's 2GB) and post the syslog. Is safe mode the same as maintenance mode?

I noticed several Reiserfs errors on one disk, so I ran reiserfsck and was told to run reiserfsck --rebuild -tree. That command ran overnight, and in the end the process was killed on Pass #1. Nothing was posted during that pass -- it must have been killed very early. I intend to go home tonight and first run --fix -fixable followed by the --rebuild once again. If that doesn't work, however, any thoughts on why the process would be killed during pass #1? I'm running 5.0.6 and the disk in question is a 3TB Toshiba. Thank you! Dave

You have a bad / loose sata and/or power cables to WD-WMC1T1495377 (disk 5) This is the root cause of the redball This entry is in your syslog prior to the errors: Mar 8 19:45:51 Tower kernel: ata13.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0 Mar 8 19:45:51 Tower kernel: ata13.00: failed command: READ DMA EXT Mar 8 19:45:51 Tower kernel: ata13.00: cmd 25/00:00:58:ff:a1/00:02:41:01:00/e0 tag 0 dma 262144 in Mar 8 19:45:51 Tower kernel: res 51/40:8f:c8:00:a2/00:00:41:01:00/e0 Emask 0x9 (media error) Mar 8 19:45:51 Tower kernel: ata13.00: status: { DRDY ERR } Mar 8 19:45:51 Tower kernel: ata13.00: error: { UNC } While it also does state "media error" which implies a bad sector, I have a feeling that it is merely a by-product of the power / sata connection. Since you just replaced disk10, you probably knocked loose the connections to that drive. You should reseat all of the connections to the drives and try it again. Problems like these are why I'm a huge fan of hot-swap bays for large arrays. Thank you! I ran a smart report and found no issues, so switched cables and rebuilt the disk and all is well.

I've marked this solved, but I'm not entirely sure why it's working. I decided to move all disks into a different server. That didn't work, so I was intending to change motherboards on my existing server and then move all the disks back into that box. Before I swapped MB's I tried booting from the original flash drive and it worked. So at some point I must have created an issue where the BIOS was trying to boot from a disk and wasn't able to, so it froze. I still don't know why Syslinux would freeze when it was trying to boot from the flash, but somehow I guess I got past that. So my advice would be to disconnect all disks and try booting. That seemed to help me pinpoint an eventual BIOS conflict. Good Luck!! Dave

I replaced Disk 10 in my array, which had lots of errors and was going downhill fast. Ran a parity check afterwards and all was well for a while, but finished the check with errors on Disk 5 and a red ball on that drive. Syslog attached. I intended to just switch cables and try again, but though I'd check here first. Thank you! Dave syslog_DPC_MAR_9.zip

Ok, will give those a try. With 5.0rc8 I at least get a new error message: SYSLINUX 3.72 Could not find kernel image: linux boot:

No, I think both MB's are too old to support UEFI. I checked boot order on both and there was no mention of UEFI / EFI / Legacy (except for legacy LAN). Found a dl of 5.0rc8. Trying that now

Thank you! Unfortunately it still didn't boot. I formatted the 8GB stick, made it bootable and was able to bring up 5.0rc11 on another server. Still getting same error message when I try booting on the Gigabyte server, though. Going to go through my files and see if I can find something older that would work.

Would very much appreciate it if you could upload 5rc11 to me. I have back-ups of working sticks, but never kept the downloads. Can you transfer the file to me in this forum? Or e-mail?

Has to be some kind of compatibility issue. Was just able to boot both the old 1GB and new 8GB sticks to UNRAID 5.0.6 on my HTPC. Apparently the Gigabyte MB's in my server and desktop don't like the latest version. I still haven't been able to figure out how to roll back syslinux to an earlier version, though. Any ideas?

What file specifically needs to be an older version? ldlinux.sys? I tried formatting, copying the 5.0.6 files onto the stick, running mkae_bootable and then copying all files from a v4.7 back-up across to the stick. Still getting same error.

Yes, I've primarily been using FAT32. I haven't found a way to format FAT, possibly because the new stick is 8GB and FAT doesn't like that? And yes, have run the .bat file and all files, including the hidden ones, seem to be present and accounted for. Strange then. Although it does not say so explicitly, the error message seems to imply that the syslinux.cfg file (in the syslinux folder) cannot be found. It obviously got far enough to invoke the syslinux loader. Having said that, I seem to remember something about some motherboards having problems with the latest syslinux versions. In these case an older version of syslinux worked for them. That would make sense. I've upgraded UNRAID versions several times, but always by overwriting bzimage and bzroot files. I may have quite happily been using an old version os syslinux which I've now wiped out.

Yes, I've primarily been using FAT32. I haven't found a way to format FAT, possibly because the new stick is 8GB and FAT doesn't like that? And yes, have run the .bat file and all files, including the hidden ones, seem to be present and accounted for.

Here's the full error: SYSLINUX 4.03 2010-10-22 ERROR: No configuration file found No DEFAULT or UI configuration directive found boot:

This is starting to drive me nuts, so time to ask for help. I've had an UNRAID server running for years, most recently on 5.0.6. A recent syslog review indicated that the USB stick might be corrupt, so I decided to pull it, reformat, and start fresh. Since that time I have not been able to boot. I have formatted using Windows XP and Windows 7, Mac OS X and Ubuntu. I've tried FAT 32, exFAT and EXT4. Have also tried versions 5.x and 6.x of UNRAID. Have also changed boot sequence in the BIOS from USB-HDD to USB-FDD. When I boot I get one of two error messages, either Invalid UI or DEVICE configuration or boot disk not recognized, replace and try again. The original stick was a Sandisk Cruzer 1GB. I bought a Sandisk Cruzer 8GB and am having exactly the same issues. Have also tried booting my regular desktop PC with both sticks and get the same errors. Both my desktop and the UNRAID server will boot to Ubuntu from a USB stick, so I can't figure out why they won't boot to UNRAID. UNRAID motherboard is a Gigabyte GA-G33M-DSR2. Thank you!

Now I'm afraid the flash drive is toast. I stopped and restarted the array. No red balls. Parity sync started and no errors showing, but showing 4 billion reads and 17 billion writes on the flash.

Did you remove the disks and pack them separately? If you shipped them in hot-swap cages, the weight of the disks and any vibrations during the move may very well have caused microscopic fractures on the PCB's and/or the plug-in connectors. There have been several issues caused by doing this on other threads in this forum ... the most recent on a system that was shipped overseas to Germany where the user ended up replacing all 3 of his hot-swap cages because of these issues. Nope! Just left everything mounted as it was, packed it well and made the move. I'd moved the same box from Singapore to Texas a few years ago without incident, but maybe I wasn't so lucky this time.

Testdisk shows a bad sector count on the flash drive. Should I repair that with testdisk? Sorry for being so helpless, but I'd rather go slowly and ask along the way than mess something up. Yes to repair flash. Disk 10 smart report is the one that looks worrisome. But you can't do anything about that while disk 13 is red. What did you find out about your cables and connections when you checked them? Do you have locking SATA cables? Any reason something might have come loose? I've been through the server and checked every cable connection -- no issues that I could find. I did move across the country about six months ago, but I'd have thought that anything that was going to come loose would have shown up months ago. I'll repair the flash and reboot. I'll repair the flash drive

And yes, you're exactly right that webGui-latest.plg is zero byte. Would the Upgrade webGui utility fix that?

I've attached four smart reports. Disk 13 is the one which currently has a red ball, but previously Disks 3 and 8 both went red, and Disk 10 just before that. I ran a short smart test on all four and then ran the reports. I'll check the flash and see what I can find. Thank you! smart13.txt smart03.txt smart08.txt smart10.txt

It does - which suggest the file that was attached is not the correct syslog. Turns out that was the right file! My clock and calendar must need to be updated. I run the machine headless and had never noticed that it was way off. So the syslog that I posted does reflect the current situation. Thank you!

I realized that was probably what happened as soon as I hit send. My bad! Will post the right file tonight.