[Samba] 3.0.25b-1.1.72-1411 - copy from and to the same samba share
Dragan Krnic
with error "The specified network name is no longer available".
The remaining copy is the same size as source but corrupted.
WinXP logs an obscure NetBT Event ID 4322, which says
"NetBT could not process a request, because at least one
OutOfResources-Exception occurred in the last hour".
One can reproduce this error anytime if one chooses a big enough
file and copies it a second time before the first copy is done.
Samba 3.0.24 does not have this bug.
When a double copy fails there is this line in full_audit log
at the moment the errors pop up:
Jul 16 14:52:15 p92 smbd_audit: [2007/07/16 14:52:15, 0]
lib/util_sock.c:read_data(534)
Jul 16 14:52:15 p92 smbd_audit: read_data: read failure for 4 bytes
to client 172.24.204.176. Error = Connection reset by peer
Is this a known problem ?
Can I produce more diagnostics to help fix it ?
--
To unsubscribe from this list go to the following URL and read the
instructions: https://lists.samba.org/mailman/listinfo/samba
Jeremy Allison
> A copy from a samba share to the same share frequently fails
> with error "The specified network name is no longer available".
>
> The remaining copy is the same size as source but corrupted.
>
> WinXP logs an obscure NetBT Event ID 4322, which says
> "NetBT could not process a request, because at least one
> OutOfResources-Exception occurred in the last hour".
>
> One can reproduce this error anytime if one chooses a big enough
> file and copies it a second time before the first copy is done.
> Samba 3.0.24 does not have this bug.
>
> When a double copy fails there is this line in full_audit log
> at the moment the errors pop up:
>
> Jul 16 14:52:15 p92 smbd_audit: [2007/07/16 14:52:15, 0]
> lib/util_sock.c:read_data(534)
> Jul 16 14:52:15 p92 smbd_audit: read_data: read failure for 4 bytes
> to client 172.24.204.176. Error = Connection reset by peer
>
> Is this a known problem ?
> Can I produce more diagnostics to help fix it ?
Open a bug and attach an ethereal/wireshare network trace please.
Jeremy.
Giulio Orsero
> On Mon, Jul 16, 2007 at 03:08:01PM +0200, Dragan Krnic wrote:
> > A copy from a samba share to the same share frequently fails
> > with error "The specified network name is no longer available".
> >
> > The remaining copy is the same size as source but corrupted.
> >
> > WinXP logs an obscure NetBT Event ID 4322, which says
> > "NetBT could not process a request, because at least one
> > OutOfResources-Exception occurred in the last hour".
I'm having the exact same issue with 3.0.25b, didn't happen with
previous versions.
It seems to be happening randomly.
Josh Kelley
> > WinXP logs an obscure NetBT Event ID 4322, which says
> > "NetBT could not process a request, because at least one
> > OutOfResources-Exception occurred in the last hour".
>
I've been tracking this same problem and managed to get a Wireshark
capture, so I posted it to
https://bugzilla.samba.org/show_bug.cgi?id=4796. If there's any other
information I can provide or testing I can perform, please let me
know.
Josh Kelley
Jeremy Allison
> On 7/16/07, Jeremy Allison <j...@samba.org> wrote:
> >On Mon, Jul 16, 2007 at 03:08:01PM +0200, Dragan Krnic wrote:
> >> WinXP logs an obscure NetBT Event ID 4322, which says
> >> "NetBT could not process a request, because at least one
> >> OutOfResources-Exception occurred in the last hour".
> >
> >Open a bug and attach an ethereal/wireshare network trace please.
>
> I've been tracking this same problem and managed to get a Wireshark
> capture, so I posted it to
> https://bugzilla.samba.org/show_bug.cgi?id=4796. If there's any other
> information I can provide or testing I can perform, please let me
> know.
I've just posted a fix for this as an attachment for this bug (#4796).
If someone would like to test it for 3.0.25c I'd be very grateful.
Thanks,
Jeremy.
Franz Sirl
> On Tue, Jul 17, 2007 at 05:07:11PM -0400, Josh Kelley wrote:
>> On 7/16/07, Jeremy Allison <j...@samba.org> wrote:
>>> On Mon, Jul 16, 2007 at 03:08:01PM +0200, Dragan Krnic wrote:
>>>> WinXP logs an obscure NetBT Event ID 4322, which says
>>>> "NetBT could not process a request, because at least one
>>>> OutOfResources-Exception occurred in the last hour".
>>> Open a bug and attach an ethereal/wireshare network trace please.
>> I've been tracking this same problem and managed to get a Wireshark
>> capture, so I posted it to
>> https://bugzilla.samba.org/show_bug.cgi?id=4796. If there's any other
>> information I can provide or testing I can perform, please let me
>> know.
>
> I've just posted a fix for this as an attachment for this bug (#4796).
> If someone would like to test it for 3.0.25c I'd be very grateful.
Hi,
I can confirm that a re-build of the current SUSE 3.0.25b RPMs plus
today's diff between SAMBA_3_0_RELEASE and SAMBA_3_0_25 fixes the
strange file corruption problem I was seeing with LIB.EXE from
VisualStudio6 working on a Samba share. Though this was already true
with yesterdays diff, but didn't work with a diff from last week.
Franz.
Jeremy Allison
> Hi,
>
> I can confirm that a re-build of the current SUSE 3.0.25b RPMs plus
> today's diff between SAMBA_3_0_RELEASE and SAMBA_3_0_25 fixes the
> strange file corruption problem I was seeing with LIB.EXE from
> VisualStudio6 working on a Samba share. Though this was already true
> with yesterdays diff, but didn't work with a diff from last week.
Great - thanks for the update. I'm hoping I've now got this
right for 3.0.25c.
Thanks,
Jeremy.
Franz Sirl
> On Wed, Jul 18, 2007 at 04:01:45PM +0200, Franz Sirl wrote:
>> Hi,
>>
>> I can confirm that a re-build of the current SUSE 3.0.25b RPMs plus
>> today's diff between SAMBA_3_0_RELEASE and SAMBA_3_0_25 fixes the
>> strange file corruption problem I was seeing with LIB.EXE from
>> VisualStudio6 working on a Samba share. Though this was already true
>> with yesterdays diff, but didn't work with a diff from last week.
>
> Great - thanks for the update. I'm hoping I've now got this
> right for 3.0.25c.
Ok, since it seems you are heading for a 3.0.25c release soon, I'd
better report this little build failure with the python stuff:
building 'smb' extension
gcc -pthread -fno-strict-aliasing -DNDEBUG -O2 -march=i586 -mtune=i686
-fmessage-length=0 -Wall -D_FORTIFY_SOURCE=2 -g -O2 -g -m32 -march=i586
-mtune=i686 -fmessage-length=0 -D_FORTIFY_SOURCE=2 -D_GNU_SOURCE
-D_LARGEFILE64_SOURCE -DIDMAP_RID_SUPPORT_TRUSTED_DOMAINS
-D_SAMBA_BUILD_=3 -fPIC -I/usr/include/python2.4 -c python/py_smb.c -o
build/temp.linux-i686-2.4/python/py_smb.o -O2 -g -m32 -march=i586
-mtune=i686 -fmessage-length=0 -D_FORTIFY_SOURCE=2 -D_GNU_SOURCE
-D_LARGEFILE64_SOURCE -DIDMAP_RID_SUPPORT_TRUSTED_DOMAINS
-D_SAMBA_BUILD_=3 -DHAVE_CONFIG_H -D_LARGEFILE64_SOURCE
-D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -DLDAP_DEPRECATED -O2 -g -m32
-march=i586 -mtune=i686 -fmessage-length=0 -D_FORTIFY_SOURCE=2
-D_GNU_SOURCE -D_LARGEFILE64_SOURCE -DIDMAP_RID_SUPPORT_TRUSTED_DOMAINS
-D_SAMBA_BUILD_=3 -Iinclude -I./include -I. -I. -I./lib/replace
-I./lib/talloc -I./tdb/include -I./libaddns -I./librpc -DHAVE_CONFIG_H
-D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE
-DLDAP_DEPRECATED -I/home/fsirl/BUILD/samba-3.0.25b/source/lib
-D_SAMBA_BUILD_=3
In file included from /usr/include/python2.4/Python.h:8,
from ./python/py_common.h:33,
from ./python/py_smb.h:24,
from python/py_smb.c:21:
/usr/include/python2.4/pyconfig.h:838:1: warning: "_POSIX_C_SOURCE"
redefined
In file included from /usr/include/stdio.h:28,
from ./lib/replace/replace.h:39,
from include/includes.h:29,
from ./python/py_common.h:24,
from ./python/py_smb.h:24,
from python/py_smb.c:21:
/usr/include/features.h:154:1: warning: this is the location of the
previous definition
python/py_smb.c: In function ‘py_smb_connect’:
python/py_smb.c:51: error: wrong type argument to unary exclamation mark
python/py_smb.c: In function ‘py_smb_session_request’:
python/py_smb.c:71: warning: assignment discards qualifiers from pointer
target type
error: command 'gcc' failed with exit status 1
make: *** [python_ext] Error 1
error: Bad exit status from /var/tmp/rpm-tmp.90950 (%build)
This is on SLES10SP1 with gcc-4.1.2 and python-2.4.2.
Franz
Franz Sirl
> On Wed, Jul 18, 2007 at 04:01:45PM +0200, Franz Sirl wrote:
>> Hi,
>>
>> I can confirm that a re-build of the current SUSE 3.0.25b RPMs plus
>> today's diff between SAMBA_3_0_RELEASE and SAMBA_3_0_25 fixes the
>> strange file corruption problem I was seeing with LIB.EXE from
>> VisualStudio6 working on a Samba share. Though this was already true
>> with yesterdays diff, but didn't work with a diff from last week.
>
> Great - thanks for the update. I'm hoping I've now got this
> right for 3.0.25c.
Sorry, but I have to take back my words, partly at least :-( .
Today the corruption problem was back when I compiled our SW on a samba
share. Now this got me really confused, because yesterday I did the
"gmake clean all" procedure for our SW at least a dozen times, and
everytime it was OK. But today already on the first try linking would
bail out because LIB.EXE generated a corrupted library.
What the heck I thought, why does it fail again. As a first measure I
rebuilt samba with todays additional fix to locking.c, but this didn't
change anything. Then, as a preparation to run a tcpdump, I looked at
the smbd process responsible for my share with lsof. Hmm, I wondered,
why does it use "netbios-ssn" as a local port, my PC is running uptodate
WXPSP2 (german, without IE7), shouldn't it use "microsoft-ds" then?
Ok, disconnected and reconnected the share, and sure enough the
connection to the share would use "microsoft-ds" now. Restarted the
build of our SW, and voila! everything is fine now and the SW builds
until the end.
So this corruption only happens for "netbios-ssn"/port 139 connections
now (? not sure about this, but I can try with original 3.0.25b too if
you want), but with "microsoft-ds"/port 445 connections all works like a
charm.
Is this enough information for you or do you want me to do some smbd
debug and/or tcpdump debugging?
Franz.
Jeremy Allison
> Jeremy Allison wrote:
> >On Wed, Jul 18, 2007 at 04:01:45PM +0200, Franz Sirl wrote:
> >>Hi,
> >>
> >>I can confirm that a re-build of the current SUSE 3.0.25b RPMs plus
> >>today's diff between SAMBA_3_0_RELEASE and SAMBA_3_0_25 fixes the
> >>strange file corruption problem I was seeing with LIB.EXE from
> >>VisualStudio6 working on a Samba share. Though this was already true
> >>with yesterdays diff, but didn't work with a diff from last week.
> >
> >Great - thanks for the update. I'm hoping I've now got this
> >right for 3.0.25c.
>
> Ok, since it seems you are heading for a 3.0.25c release soon, I'd
> better report this little build failure with the python stuff:
Separate bugzilla bug please. Also - who uses the python bindings ?
I know Jerry was wanting to remove them as they're not currently
used/properly maintained.
Jeremy.