So I have a GFH with a 3TB SATA drive that's been working like a champ for years now, running transmission and serving up files via samba. Awesome little device actually. Anyway, I came back one day and it was off the network. So I rebooted it and it came up from in the rescue system, after flashing orange (for led_error) during the uboot process. So I fired up netconsole for the next rebby floodo1 - Debian
wow Bodhi, digging this one up from the past I am, but it turns out that my problem was that my fstab had /deb/sda3 instead of /dev/sda3 in it, and instead of erroring out it was just hanging. Eventually I used the rescue system to figure this out. Fixed fstab and it booted fine :) (using arc 3338)by floodo1 - Debian
PS- you may want to search posts that I've made because I've gone through EXACTLY what you're going through and along the way I posted quite a few times on this forum. Maybe I posted more than I think, but I have the feeling that if you start with my oldest posts and just step your way forward you will see me having your problems, and then people offering solutions. PPA- I searcby floodo1 - Debian
OK so, the bad eraseblock message means your kernel doesnt have the patch that it needs to deal with MTD properly, or that your ArcNumber is set wrong, causing your kernel not to 'use the patch' (as it were). I can't give you the specific steps but I'm pretty sure I can give you a good outline. Davygravy shall be your hero for making his NSA320 archive (https://www.dropboxby floodo1 - Debian
Running off SATA HD. Is there a reason to avoid running syslog? I'm having to troubleshoot a lot of things as I make this system more and more functional, so it would be mighty handy to have. This is the first *nix system I've used that didn't "come with" syslog, so I'm wondering if there is a specific reason to not use. My guess is that it doesn't make sense foby floodo1 - Debian
guess I should have looked around on my own. Wasn't too hard to find something that will do nicely: https://wiki.archlinux.org/index.php/Full_System_Backup_with_rsyncby floodo1 - Debian
I can't use the USB3 base because it does some 4k sector conversion magic so the drive appears differently than when it's used over SATA directly. I figured it would be pretty simple because copying the files should be sufficient. I mean I don't need to copy the partition fully or anything right? I mean I started from a tarball, so can't I just make my own tarball of the fby floodo1 - Debian
I don't have the timer module :(by floodo1 - Debian
After reading around I experimented with things like 1200-1400 and it wasn't beneficial either.by floodo1 - Debian
So I finally have my 3TB sata drive booting and functioning well on my GFH. Thing of it is that I'd like to make some pretty big changes so I need to backup my existing root file system. i've got plenty of USB sticks that will hold the data, but not the partition. That is, I've got a 20gb root partition, that's sitting at 3.8gb used. I have plenty of 4gb and 8gb usb sticks thaby floodo1 - Debian
Yes, you must load the heartbeat module if you would like the double pulse that it provides. Really the whole LED system is VERY easy to use and understand using simple 'ls' command to list the directory /sys/class/leds/ You will see the available LEDs. For GoFlex Home this should be all three LEDs (the upper combo orange/green LED (where orange overrides green) and the lower whitby floodo1 - Debian
So I've umm "proven" it to myself that this thing doesn't support jumbo frames. Earlier I tried jumbo frames, realized it slowed down my file xfer massively so I returned to 1500. Went about making tons of other changes to things, then suddenly Samba performance is horrible, to the point where it won't even list a certain directory. I spend HOURS troubleshotting samba perby floodo1 - Debian
Is anyone else experiencing this? Without jumbo frames (MTU=1500) on the GFH, but with the rest of my network set to 4000/4088 (whatever you want to call it), I get about 28.7 MB/S to the GFH, which I'm fine with. Just surprised that no matter what else I set the MTU to it's significantly slower (as in 1/10th the speed if I set it to 9000) than the usual 1500.by floodo1 - Debian
Pretty sure the problem turned out to be that I had /deb/sda3 in my fstab and it was hanging on trying to mount a partition from /deb :) can't believe that I went for so long without my GFH all because of one bad letter in my fstab :( Good thing this guy had a simiar probelm (similar enough to give me an idea to look at fstab) http://www.plugcomputer.org/plugforum/index.php?topic=5807.0by floodo1 - Debian
So I just realized that i can use chroot to analyze boot problems without getting the kernel to use netconsole. So I fired up the NAND based rescue, mounted my non-booting SATA partition, chrooted over there, and here is my dmesg output. Can anyone tell me why my GFH never shows up on the network? To me this dmesg output looks fine, but I cant ping it or ssh or anything (I have it set for staticby floodo1 - Debian
CONFIG_NETCONSOLE=m CONFIG_NETCONSOLE_DYNAMIC=y what does 'm' mean?by floodo1 - Debian
interesting point. If I use Jeff's scripts will it work?by floodo1 - Debian
fyi PlugPlayer also works with foobar2000 uPnP component, so you can pull double duty with it ;)by floodo1 - Debian
I followed the instructions but when I reboot I get nothing and the machine hangs at some point. If I take my usb stick and revert the uInitrd and uImage to the backups then everything boots again. Therefore I suspect there is a problem with one of these two commands: # update-initramfs -u # mkimage -A arm -O linux -T ramdisk -C gzip -a 0x00000000 -e 0x00000000 -n initramfs-3.2.28 -d /bootby floodo1 - Debian
So, thanks to the help from this forum I was able to get my GFH setup with the rescue system, and a uboot that functions flawlessly with nc. So I am able to fully control/monitor the boot process and boot the rescue system from NAND, usb devices that have appropriately labeled partitions, as well as boot from the SATA 3tb drive. I have the 3tb drive setup with the first partition at like 20gb extby floodo1 - Debian
changing arcNumber made no difference. Pretty sure I'm going to start fresh with the OS partition on my sata drive. Since rescue boots consistently and so do all my usb sticks I think I'm going to just salvage my transmission and samba config files and wipe the sata drive and put a new filesystem on there. Hopefully that will boot correctly, then I'll make my first priority gettby floodo1 - Debian
???????? That is the nuttiest uboot environment that I have ever seen. Pretty sure your uboot should be simpler, like Optim's setup here: arcNumber=3338 baudrate=115200 bootcmd_usb=run usb_init; run usb_load_uimage; run set_bootargs_usb; run usb_boot; bootdelay=3 console=ttyS0,115200 ethact=egiga0 ethaddr=XXXXXXXXXX led_error=orange blinking led_exit=green off led_init=greenby floodo1 - uBoot
3089, pretty sure it's davy's uboot, uname -a (can't run it since I can't boot SATA) should be 3.3.2-kirkwood-wide or something.........basically the SATA rootfs was created by using davygravy's nsa320 filesystem ( http://forum.doozan.com/read.php?2,7806 ), then installing the 3.3.2 kernel .deb from davygravy (http://forum.doozan.com/read.php?2,7676).by floodo1 - Debian
if you have the wrong arc_number and you are experiencing the Too Few Good Blocks error, then you need a special kernel roll that davygravy whipped up. You see the patch for Too Few Good Blocks error is rolled into many of davygravy's kernels, specifically his 3.3.2 that supports most devices, and rescue v2.8, which is great! .... ONLY IF your arc_number is set properly. But when your arc_nuby floodo1 - uBoot
sorry, didn't mention that fsck was the first thing I did. I've done it multiple times at this point. Because I can boot into rescue and from usb I have access to basically anything that a fully functioning goflex can do. Already checked the filesystem itself, and tried to revert some of my changes from yesterday. Also tried to enable kernel netconsole, by following more than one guideby floodo1 - Debian
actually I tried heartbeat yesterday and it was working at least for the green ledby floodo1 - Debian
dollars to donuts your uboot environment doesn't have netconsole configured correctly. My suggestion is to unplug your 3tb drive, boot off your sata or another usb, then post the output of 'fw_printenv' so that we can take a look at how your uboo is configured. I too had a problem with netconsole not working and I ended up simply needing to change a few uboot environment variabby floodo1 - uBoot
I have the exact same problem. Had everything working perfectly off SATA, then a 'halt' later and everything appears fine from the uboot output (finds kernel image, and initramfs and says 'starting kernel') but then nothing! ~~~~~~~~~ Actually I don't have the same problem. Seems that your uInitrd file wasn't found during boot. My system finds that as well asby floodo1 - uBoot
also any tips on how to troublesheet the sata filesystem would be greatly appreciated. and yes, I know the best solution is to install a physical serial connection, then nc doesn't matter.....almost there but not yet...probably 2 more weeks ;) hahaby floodo1 - Debian