User avatar
pnelsonsr
Posts: 20
Joined: Sat Mar 30, 2013 6:40 pm

Can not get a good SD

Sat Mar 30, 2013 7:44 pm

I got my raspberry pi with a debian 7 4GB drive and this works fine. But I'm a fedora dude so I went and got rpfr-f18-final.zip and unziped it to rpfr-f18-final.img and rpfr-f18-final.img.md5sum. I have a 16GB class 6 SD card. I have tried and tried to create an SD that will boot but I can not get one that will. I have created the SD both with the installer and manually with dd. The creation process works without any errors. When I re-insert the SD card I can read and explore /boot on /dev/sdg1 and /root on /dev/sdg2 without any problems. I thought that maybe the card was bad so to test this I copied my debian 7 SD card with dd if=/dev/sdg of=my-d7-sd.img bs=2M and then put that on the 16GB SD card with dd if=my-d7-sd.img of=/dev/sdg bs=2M which seemed to work fine. I tried to boot the 16GB SD card and it booted fine into debian 7. I even expanded the root image using raspi-config and rebooted. All went great.

I'm really frustrated because I want fedora 18 (it's my distro and I want it). I'm not sure what else to try. When I try and boot the Fedora 18 SD I get the Raspberry Pi Icon in upper left corner! And then a few errors. Like:

Code: Select all

[ 2.284163] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #133687: comm systemd: bad extra_isize (34182 != 256)
[ 2.602970] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #133687: comm systemd: bad extra_isize (34182 != 256)
[ 2.617594] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #133687: comm systemd: bad extra_isize (34182 != 256)
[ 2.723060] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #133699: comm systemd: bad extra_isize (11050 != 256)
[ 2.890963] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #133718: comm systemd: bad extra_isize (10347 != 256)
[ 2.957078] EXT4-fs error (device mmcblk0p2): ext4_iget:3811: inode #133695: comm systemd: bad extra_isize (47258 != 256)
[ 2.971167] systemd[1]: Failed to start default target: Unit sysinit.target failed to load: Input/output error. See system logs and 'systemctl status sysinti.target' for details
[ 2.600781] systemd-cgroups-agent[50]: Failed to get D-Bus connection: Failed to connect to socket /org/fredesktop/systemd1/private: Connection refused
If I leave it here (because the ACT light is blinking randomly). After about 10 minutes the screen goes blank and that seems to be it. Anyone know anything else I can try to get a fedora raspberry pi SD to work?

herbthom
Posts: 19
Joined: Sun Mar 10, 2013 6:10 pm

Re: Can not get a good SD

Sun Mar 31, 2013 9:54 pm

Try a different / smaller card?

I've successfully installed and run rpfr-f18-final.img on several brands (Sandisk, Sony, Kingston) , sizes (4GB and 8GB), and classes (4 and 10) of SD cards on two Rev 1 Model B boards.

Don't have a 16GB card here to try.

Raspoootin
Posts: 12
Joined: Mon Apr 01, 2013 3:10 am

Re: Can not get a good SD

Mon Apr 01, 2013 7:42 pm

The problem is owned by 1 of 2 possible entities:

The distro, or the SD card.
Reduce the scenario to hardware and software minimums to prove who owns the problem.

Take the SD card that is known to fail the fedora installation.
I would do 2 different complete installs on the same SD card just to be sure.
Now, use that exact same SD card and do the latest raspbian wheezy install on it.
Try to boot the wheezy distro and expand the root partition to fill the whole card.
If successful, then that is a pointer at something not quite right about the fedora install.

After proving who owns the problem I would fresh format the same SD card.
Then do the fedora install again on the exact same card.
If the problem comes back then you have a choice to make.

User avatar
pnelsonsr
Posts: 20
Joined: Sat Mar 30, 2013 6:40 pm

Re: Can not get a good SD

Mon Apr 01, 2013 9:44 pm

I have tried it on a 2GB card and a 4GB card. Same same. In my long (sorry...) OP I stated I had loaded raspbian wheezy up on the card and it worked fine. In fact I'm running off it now on that card. I have tried it on another 16GB card and that had the same result. I do not think it is the card (completely...) but something is going wrong. Some additional info: Yesterday I got a logwatch entry on the Fedora 18 system that I was using to create the Fedora SD. I'm not sure if this means anything but here is the error:

Code: Select all

 --------------------- Kernel Begin ------------------------ 
 WARNING:  Kernel Errors Present
    EXT4-fs (sdg2): error count: 15 ...:  1 Time(s)
    EXT4-fs (sdg2): initial error at 2: ext4_iget: ...:  1 Time(s)
    EXT4-fs (sdg2): last error at 1364661492: e ...:  1 Time(s)
    EXT4-fs (sdg2): warning: mounting fs with errors, running e2fsck ...:  2 Time(s)
    EXT4-fs error (device sdg2): e ...:  18 Time(s)
 ---------------------- Kernel End ------------------------- 
So I search the log file to find these:

Code: Select all

Mar 29 19:28:34 nh kernel: [378910.872492] sd 8:0:0:0: [sdg] 31711232 512-byte logical blocks: (16.2 GB/15.1 GiB)
Mar 29 19:28:34 nh kernel: [378910.873493] sd 8:0:0:0: [sdg] No Caching mode page present
Mar 29 19:28:34 nh kernel: [378910.873498] sd 8:0:0:0: [sdg] Assuming drive cache: write through
Mar 29 19:28:34 nh kernel: [378910.875736] sd 8:0:0:0: [sdg] No Caching mode page present
Mar 29 19:28:34 nh kernel: [378910.875740] sd 8:0:0:0: [sdg] Assuming drive cache: write through
Mar 29 19:28:34 nh kernel: [378910.877161]  sdg: sdg1 sdg2
Mar 29 19:28:35 nh udisksd[1862]: Mounted /dev/sdg1 at /run/media/root/boot on behalf of uid 0
Mar 29 19:28:35 nh kernel: [378911.469991] EXT4-fs (sdg2): warning: mounting fs with errors, running e2fsck is recommended
Mar 29 19:28:35 nh kernel: [378911.471135] EXT4-fs (sdg2): mounted filesystem with ordered data mode. Opts: (null)
Mar 29 19:28:35 nh udisksd[1862]: Mounted /dev/sdg2 at /run/media/root/rootfs on behalf of uid 0
Mar 29 19:28:35 nh kernel: [378911.670118] EXT4-fs error (device sdg2): ext4_iget:3888: inode #134208: comm pool: bad extra_isize (31936 != 256)
Mar 29 19:28:36 nh kernel: [378912.189736] EXT4-fs error (device sdg2): ext4_iget:3888: inode #133644: comm pool: bad extra_isize (54273 != 256)
Mar 29 19:28:36 nh kernel: [378912.192490] EXT4-fs error (device sdg2): ext4_iget:3888: inode #133672: comm pool: bad extra_isize (37500 != 256)
Mar 29 19:28:36 nh kernel: [378912.468738] EXT4-fs error (device sdg2): ext4_iget:3888: inode #133639: comm pool: bad extra_isize (49255 != 256)
Mar 29 19:28:36 nh kernel: [378912.478363] EXT4-fs error (device sdg2): ext4_iget:3888: inode #133939: comm pool: bad extra_isize (47056 != 256)
Mar 29 19:28:36 nh kernel: [378912.527737] EXT4-fs error (device sdg2): ext4_iget:3888: inode #126630: comm pool: bad extra_isize (52178 != 256)
Mar 29 19:28:36 nh kernel: [378912.530494] EXT4-fs error (device sdg2): ext4_iget:3888: inode #126637: comm pool: bad extra_isize (11556 != 256)
Mar 29 19:28:36 nh kernel: [378912.538362] EXT4-fs error (device sdg2): ext4_iget:3888: inode #126671: comm pool: bad extra_isize (59692 != 256)
Mar 29 19:28:36 nh kernel: [378912.541122] EXT4-fs error (device sdg2): ext4_iget:3888: inode #126945: comm pool: bad extra_isize (55763 != 256)
So I tried inserting the SD and un-mounting both /boot and /root. I then ran fsck -y /dev/sdg2. It took a few minutes with lots of stuff being fixed but it did complete. I also did /dev/sdg1 and then did both again to make sure. Then I ejected the SD and plugged it back in. Here is what the logs showed now:

Code: Select all

Apr  1 14:20:11 nh kernel: [619608.038062] sd 8:0:0:0: [sdg] 31711232 512-byte logical blocks: (16.2 GB/15.1 GiB)
Apr  1 14:20:11 nh kernel: [619608.039679] sd 8:0:0:0: [sdg] No Caching mode page present
Apr  1 14:20:11 nh kernel: [619608.039685] sd 8:0:0:0: [sdg] Assuming drive cache: write through
Apr  1 14:20:11 nh kernel: [619608.041808] sd 8:0:0:0: [sdg] No Caching mode page present
Apr  1 14:20:11 nh kernel: [619608.041813] sd 8:0:0:0: [sdg] Assuming drive cache: write through
Apr  1 14:20:11 nh kernel: [619608.043230]  sdg: sdg1 sdg2
Apr  1 14:20:12 nh udisksd[1862]: Mounted /dev/sdg1 at /run/media/root/boot on behalf of uid 0
Apr  1 14:20:12 nh kernel: [619608.649335] EXT4-fs (sdg2): mounted filesystem with ordered data mode. Opts: (null)
Apr  1 14:20:12 nh udisksd[1862]: Mounted /dev/sdg2 at /run/media/root/rootfs on behalf of uid 0
So feeling like just maybe I had it... I tried it in the Pi. It gave me some new errors:

Code: Select all

systemd-fsck[53]: rootfs: clean, .....
systemd-logind[101]: Failed to fully start up daemon: Connection refused
systemd-logind[156]: Failed to get system D-Bus connection: Did not receive a reply.  Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken
Is there a way to manually create the rootfs file system?

User avatar
pnelsonsr
Posts: 20
Joined: Sat Mar 30, 2013 6:40 pm

Re: Can not get a good SD

Mon Apr 01, 2013 11:28 pm

I was trying to mount the rootfs partition straight from the img file and I noticed something. parted rpfr-f18-final.img didn't show partitions correctly. So I md5sum'ed it and it didn't match rpfr-f18-final.img.md5sum... I did do a check when I downloaded the img file but had problems with the SD reader on that system so I transfered it to another and did not check the md5sum on the new location. So I re-downloaded it from http://scotland.proximity.on.ca/raspber ... 18/images/, unzipped it and checked the md5sum. With that being good I installed it on the SD and tried in on the Raspberry Pi. Success! Man I wasted a ton of time on this! Arg... But I learned a bunch about images! Just trying to keep that in mind!

ChampS
Posts: 7
Joined: Fri Feb 01, 2013 5:12 pm

Re: Can not get a good SD

Tue Apr 02, 2013 12:41 pm

I'm using the Samsung MB-SPAGA/EU Class 10 Plus Serie SDHC 16GB

http://www.amazon.de/gp/product/B00569J ... UTF8&psc=1

It works great and I don't have Problems with Fedora Remix

Return to “Pidora / Fedora”