Corrupt Image?

73 views
Skip to first unread message

Toasterlint

unread,
Dec 28, 2011, 12:27:44 PM12/28/11
to Beagle Board
Hello all, just got my BeagleBone for Christmas and have been playing
around with it for the last few days and have been learning a lot! I
sort of messed things up with node and cloud9, so wanted to start over
with an fresh image. I also had the network issue a few others have
reported after doing an opkg update/upgrade so decided I'd try the
2011.12.26 image now out there(as someone else reported it fixed
networking issue). However, I've run into a problem:

I've downloaded Angstrom-Cloud9-IDE-eglibc-ipk-v2011.12-core-
beaglebone-2011.12.26.img.gz multiple times from the Amazon mirror,
and every time the MD5SUM comes out different. It also errors with an
unexpected end of file when I try to either zcat it to the SD Card or
simply try to gunzip it. For the fun of it I tried testing the file
via Windows and WinRAR and it says it is corrupt as well.

I'm in the process of trying to download the file from the main
angstrom site, but it is going slow(about 10KBps instead of 2.5MBps
from Amazon...thus why I've only done the Amazon site). I'll compare
the MD5SUM if/once I've finished the download from the non-Mirror
site.

Can anyone else verify this? Thanks!

Henry

David Lambert

unread,
Dec 28, 2011, 1:01:21 PM12/28/11
to beagl...@googlegroups.com
I had one bad download from the Angstrom site, but the second download
appeared to work. Here is my md5sum:

eabb54cebd18f4f2d1f3925a043ad5b7
Angstrom-Cloud9-IDE-eglibc-ipk-v2011.12-core-beaglebone-2011.12.26.img.gz

Toasterlint

unread,
Dec 28, 2011, 1:08:09 PM12/28/11
to Beagle Board
I've download from the Amazon mirror 6 times now and each time I get
an MD5 sum that starts with 35... can't remember as I've deleted the
file now. It is going to take me about 8 hours to download from
Angstrom's site, which isn't an option for me right now... may had to
kick off that download tonight when I call it a night.

Edouard Lafargue

unread,
Dec 28, 2011, 3:32:24 PM12/28/11
to beagl...@googlegroups.com
You can try with gunzip -t and see what it tells you. On my system, it complains about an unexpected end of file...

gunzip -t Angstrom-Cloud9-IDE-eglibc-ipk-v2011.12-core-beaglebone-2011.12.26.img.gz 
gzip: Angstrom-Cloud9-IDE-eglibc-ipk-v2011.12-core-beaglebone-2011.12.26.img.gz: unexpected end of file


--
You received this message because you are subscribed to the Google Groups "Beagle Board" group.
To post to this group, send email to beagl...@googlegroups.com.
To unsubscribe from this group, send email to beagleboard...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/beagleboard?hl=en.


Pocius

unread,
Dec 28, 2011, 10:14:45 PM12/28/11
to Beagle Board
I'm also seeing this. I'm getting a checksum of
3561f2a0c85c46c48151d6f13f33b392, along with the unexpected end of
file. In fact, most of the checksums of the files in the repository
seem to be wrong ( unless I did "md5sum -c md5sums" wrong...)

Toasterlint

unread,
Dec 28, 2011, 10:38:42 PM12/28/11
to Beagle Board
So I set up another computer to download from the *non-Mirror* site
and it just completed... I ran a MD5SUM against it and got the same
bad sum of: 3561f2a0c85c46c48151d6f13f33b392

Not sure what to do as no matter what I get the same bad MD5SUM for
this image file and it won't decompress(as expected). I downloaded
the 11.16 version and the MD5SUM checked out fine, so I used it for
now, really wish the 12.26 version worked. Appears Edouard may be
getting the same corrupt file, so is there a way to get someone to re-
upload a valid image?

Benny B. Simonsen

unread,
Dec 29, 2011, 3:47:23 AM12/29/11
to Beagle Board
I see the same thing.
Unexpected end of file (downloaded from the angstrom site)
The file is also ~25MB smaller than the previous one (would not expect
that)

Benny

Jason Kridner

unread,
Dec 29, 2011, 9:52:47 AM12/29/11
to beagl...@googlegroups.com
I confirmed that the mirror has a bad copy of the 12/26 image.  I think the 11/16 image is fine.  I'm in the process of updating the mirror of the 12/26 image.  It may take as much as 8 hours to update.

Koen Kooi

unread,
Dec 29, 2011, 10:01:32 AM12/29/11
to beagl...@googlegroups.com

Op 29 dec. 2011, om 15:52 heeft Jason Kridner het volgende geschreven:

> I confirmed that the mirror has a bad copy of the 12/26 image. I think the 11/16 image is fine. I'm in the process of updating the mirror of the 12/26 image. It may take as much as 8 hours to update.

rsync --partial is your friend :)

Jason Kridner

unread,
Dec 29, 2011, 10:36:49 AM12/29/11
to beagl...@googlegroups.com
I'm using rsync --partial, but something else is wrong with my cron job.

I'm currently pushing to the S3 server.  Should be there in just a few minutes.

Jason Kridner

unread,
Dec 29, 2011, 10:57:33 AM12/29/11
to beagl...@googlegroups.com
On Thu, Dec 29, 2011 at 10:36 AM, Jason Kridner <jkri...@beagleboard.org> wrote:
On Thu, Dec 29, 2011 at 10:01 AM, Koen Kooi <ko...@beagleboard.org> wrote:

Op 29 dec. 2011, om 15:52 heeft Jason Kridner het volgende geschreven:

> I confirmed that the mirror has a bad copy of the 12/26 image.  I think the 11/16 image is fine.  I'm in the process of updating the mirror of the 12/26 image.  It may take as much as 8 hours to update.

rsync --partial  is your friend :)

I'm using rsync --partial, but something else is wrong with my cron job.

I'm currently pushing to the S3 server.  Should be there in just a few minutes.

Update is complete.  I downloaded and verified the md5sum.
Reply all
Reply to author
Forward
0 new messages