Pyrit fails to export: file write error...

[GoogleCodeExporter]

What steps will reproduce the problem?
1. After batching, I issued:
# pyrit -e cubo30 -o cowpt.out export_cowpatty

What is the expected output? What do you see instead?

Pyrit 0.3.1-dev (svn r228) (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Connecting to storage at 'file://'...  connected.
Exporting to 'cowpt.out'...
19459597 entries written (315165.7/s)...Error while loading results
6e812d31e5b83131f924380df2b1d9a2 for ESSID 'cubo30'
Traceback (most recent call last):
  File "/usr/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/lib/python2.5/site-packages/pyrit_cli.py", line 106, in
initFromArgv
    func(self, **options)
  File "/usr/lib/python2.5/site-packages/pyrit_cli.py", line 313, in
export_cowpatty
    for results in storage.iterresults(essid):
  File "/usr/lib/python2.5/site-packages/cpyrit/storage.py", line 228, in
iterresults
    yield self[essid, key]
  File "/usr/lib/python2.5/site-packages/cpyrit/storage.py", line 388, in
__getitem__
    raise IOError("File-format for '%s' unknown." % filename)
NameError: global name 'filename' is not defined

I reissued command many times, getting the same error at exactly the same line.

What version of the product are you using? On what operating system?
Pyrit 0.3.1-dev on Backtrack 4.

Additional informations:
Many segmentation faults during batchprocess.

Original issue reported on code.google.com by segac...@gmail.com on 10 Apr 2010 at 3:41

[GoogleCodeExporter]

segaccia, upgrede to python2.6 and recompile.

Original comment by pyrit.lo...@gmail.com on 10 Apr 2010 at 6:06

[GoogleCodeExporter]

Many thanks for quick reply.

But the pyrit version I used is pre-packaged with bt4, so (I think), it must 
work
out-of-the-box...

Bt4 uses python for his own internal tasks, so upgrading manually to 2.6 (that 
is not
packaged) can hang my system horribly...

Original comment by segac...@gmail.com on 10 Apr 2010 at 1:43

[GoogleCodeExporter]

Ok, installed python 2.6 leaving 2.5 installed.

Uninstalled pyrit 0.3.1-dev.

Manually builded and installed pyrit 0.3.0 via "python2.6 setup.py..etc..". Done
successful installation.

Got the same error at the same point :(

Pyrit 0.3.0 (C) 2008-2010 Lukas Lueg http://pyrit.googlecode.com
This code is distributed under the GNU General Public License v3+

Connecting to storage at 'file://'...  connected.
Exporting to 'cow.out'...
19459597 entries written (187103.6/s)...Error while loading results
6e812d31e5b83131f924380df2b1d9a2 for ESSID 'cubo30'
Traceback (most recent call last):
  File "/usr/bin/pyrit", line 6, in <module>
    pyrit_cli.Pyrit_CLI().initFromArgv()
  File "/usr/local/lib/python2.6/site-packages/pyrit_cli.py", line 106, in initFromArgv
    func(self, **options)
  File "/usr/local/lib/python2.6/site-packages/pyrit_cli.py", line 313, in
export_cowpatty
    for results in storage.iterresults(essid):
  File "/usr/local/lib/python2.6/site-packages/cpyrit/storage.py", line 228, in
iterresults
    yield self[essid, key]
  File "/usr/local/lib/python2.6/site-packages/cpyrit/storage.py", line 388, in
__getitem__
    raise IOError("File-format for '%s' unknown." % filename)
NameError: global name 'filename' is not defined

Original comment by segac...@gmail.com on 10 Apr 2010 at 2:05

[GoogleCodeExporter]

segaccia,
to me pyrit in BT4 gave only segfault and headhache.
I wiped out BT4 and definitively move to debian stable (5.0.4): I user only .deb
included into debian, with exception, of course, of ATI driver 10.2 I got from 
ATI
site and python source .tgz file of 2.6.4 version.
Ablout Pithon2.6 I downloaded the package of source, then compile and 
installed. then
I used it to compile last svn pyrit.
Works fine, segfault disappeared 100%.

Original comment by pyrit.lo...@gmail.com on 10 Apr 2010 at 11:29

[GoogleCodeExporter]

Thanks again for tips...

Have you had other (general) troubles with bt4?

Furthermore, have you tried pyrit opencl addon with your ati card? (It does not 
work
with my nvidia card, despite of full opencl support by latest nvidia 
driver...). 

Original comment by segac...@gmail.com on 10 Apr 2010 at 11:58

[GoogleCodeExporter]

this error was fixed in 0.3.1-dev r231. It went unnoticed because its the error
message that Pyrit shows when encountering a broken/unknown file in the
filesystem-storage.

You've probably killed Pyrit or the whole box while Pyrit was writing to a 
file. The
filesystem truncates the file to zero bytes in that case. Either try to find 
that
file by hand (it should be somewhere in ~/.pyrit/blobspace/essids/*/) or update 
to r231

Original comment by lukas.l...@gmail.com on 11 Apr 2010 at 2:43

[GoogleCodeExporter]

Thanks a lot for reply, Lukas.

I absolutely haven't killed nor pyrit nor the linux box during the process. 
Note, 
furthermore, that pyrit stops writing to file at the same point in any attempt.

Surely I'll upgrade to r231, however I would be pleased to solve the problem 
with 
current release, before.

Another interesting issue are the many segfault with pre-packaged version of 
pyrit in 
bt4, as noted already by pyrit.lover. Any about?

Many thanks again.

Original comment by segac...@gmail.com on 13 Apr 2010 at 11:13

[GoogleCodeExporter]

Maybe some other filesystem-related problem. Look out for zero-length files in
~/.pyrit/blobspace/essid/*/ and delete those. I'll close this issue for now...

I'm perfectly willing to blame the python-build in bt4 for the segfaults you are
probably talking about. The segfaults occur within the interpreter in
interpreted-only code. I can't reproduce the problem outside of bt4 and people 
who
updated their python installation reported the problem to go away.

Original comment by lukas.l...@gmail.com on 14 Apr 2010 at 5:34

• Changed state: Fixed

[GoogleCodeExporter]

By the way, write error recovered following your advice. Thanks.

Original comment by segac...@gmail.com on 14 Apr 2010 at 8:47

[GoogleCodeExporter]

Had the same problem after a power outage - after deleting all zero sized 
files, pyrit worked as expected (Ubuntu 10.04, pyrit 0.3.0, python 2.6.5). I 
can also confirm a lot of segmentation faults in BackTrack, so I no longer use 
it.

Original comment by dkaloya...@gmail.com on 29 Jul 2010 at 7:20