-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I did not do anything to consciously disable vlan offload, however
there may have been something in the build that would have done that.
I can still capture, however, only for a period of time before it
stops. As I reported before, if I run the userland/tcpdump-4.1.1
version of tcpdump I get traffic for a while and then it just stops.
If I have another window open that is also watching, they stop together.

If I output to a file, then the dump will generally end within
seconds. If I send it to screen it takes longer, but then I believe
that is mostly the backup of printing to the screen since the elapsed
time as seen from the timestamps will be very short (couple seconds at
most).

This was a new build of the system (from bare metal) to see if the
problem I experienced on another machine was replicated. The one thing
I did on this machine that was not different was that I didn't back
rev the kernel to see if the driver was compiling on
2.6.32-504.el6.x86_64 yet.

[root@IDS-05 PF_RING]# ldd userland/tcpdump-4.1.1/tcpdump
        linux-vdso.so.1 =>  (0x00007fff1f5ff000)
        libpthread.so.0 => /lib64/libpthread.so.0 (0x0000003400800000)
        libnuma.so.1 => /usr/lib64/libnuma.so.1 (0x0000003400000000)
        libnl.so.1 => /lib64/libnl.so.1 (0x0000003402c00000)
        librt.so.1 => /lib64/librt.so.1 (0x0000003400c00000)
        libm.so.6 => /lib64/libm.so.6 (0x0000003401000000)
        libc.so.6 => /lib64/libc.so.6 (0x0000003400400000)
        /lib64/ld-linux-x86-64.so.2 (0x00000033ffc00000)


Weeding out comments the driver load script looks like:

FAMILY=ixgbe
rmmod ixgbe
rmmod pf_ring
echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages
mkdir /mnt/huge
mount -t hugetlbfs nodev /mnt/huge
insmod ../../../../../../kernel/pf_ring.ko
insmod ./ixgbe.ko MQ=1,1,1,1 RSS=8,8,8,8
sleep 1
killall irqbalance
INTERFACES=$(cat /proc/net/dev|grep ':'|grep -v 'lo'|grep -v 'sit'|awk
- -F":" '{print $1}'|tr -d ' ')
for IF in $INTERFACES ; do
        TOCONFIG=$(ethtool -i $IF|grep $FAMILY|wc -l)
        if [ "$TOCONFIG" -eq 1 ]; then
                printf "Configuring %s\n" "$IF"
                ifconfig $IF up
                sleep 1
                bash ../scripts/set_irq_affinity $IF
                ethtool -G $IF rx 32768
                ethtool -G $IF tx 32768
                ethtool -K $IF rxvlan off
done

Using the stock load_drivers.sh script had similar results however it
did seem to run longer before stopping but since that time varies from
run to run there is no way of knowing if that was related.


Yours,

John

On 11/8/2014 1:26 AM, Alfredo Cardigliano wrote:
> Hi John that command disables vlan offload (tag stripping), this is
> strange (already disabled?) but you should still be able to capture
> from the interface, right?
> 
> Alfredo
> 
>> On 07 Nov 2014, at 23:03, John Ives <[email protected]>
>> wrote:
>> 
> Since I have enough boxes dedicated to this project I could run
> one with the 2.6.32-504.el6.x86_64 kernel while still trying to get
> to the bottom of the ZC issues on another.  I can report that the 
> ixgbe-3.22.3-zc driver does now compile on this kernel. However
> when trying to run the load_drivers script it gets an error at
> "ethtool -K $IF rxvlan off". When this runs (and after the set ring
> commands e.g. ethtool -G ran without error), I get "Could not
> change any device features." I found a similar conversation that
> happened in Mar 2013, however that was for DNA and the discussion
> revolved around snort. Whel I am not even to the snort part yet.
> 
> Any suggestions?
> 
> John
> 
> 
> On 10/21/2014 01:29 PM, John Ives wrote:
>>>> Has there been any progress on this as I am experiencing the
>>>> same issue building 6.0.3 (trying to get to the ZC DAQ) fresh
>>>> from svn on a RHEL 6.6 box (kernel 2.6.32-504.el6.x86_64 as
>>>> well).
>>>> 
>>>> Yours,
>>>> 
>>>> John
>>>> 
>>>> On 10/16/14 9:08 AM, Scott Finlon wrote:
>>>>> make -C /lib/modules/2.6.32-504.el6.x86_64/build 
>>>>> SUBDIRS=/usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src
>>>>>
>>>>>
>
>>>>> 
modules make[1]: Entering directory
>>>>> `/usr/src/kernels/2.6.32-504.el6.x86_64' CC [M] 
>>>>> /usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe_main.o
>>>>
>>>>>
>>>>
>>>>
>>>>> 
In file included from
>>>> /usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe_main.c:51:
>>>>>
>>>>
>
>>>> 
/usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe.h:
>>>>> In function ‘ixgbe_qv_unlock_napi’: 
>>>>> /usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe.h:602:
>>>>>
>>>>>
>
>>>>> 
error: too many arguments to function ‘napi_gro_flush’
>>>>> /usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe_main.c:
>>>>>
>>>>>
>
>>>>> 
At top level:
>>>>> /usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe_main.c:9778:
>>>>>
>>>>>
>
>>>>> 
error: unknown field ‘ndo_busy_poll’ specified in initializer
>>>>> /usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe_main.c:9778:
>>>>>
>>>>>
>
>>>>> 
warning: initialization from incompatible pointer type make[2]: ***
>>>>> [/usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src/ixgbe_main.o]
>>>>>
>>>>>
>
>>>>> 
Error 1 make[1]: ***
>>>>> [_module_/usr/local/src/PF_RING/drivers/PF_RING_aware/intel/ixgbe/ixgbe-3.22.3-zc/src]
>>>>>
>>>>>
>
>>>>> 
Error 2 make[1]: Leaving directory
>>>>> `/usr/src/kernels/2.6.32-504.el6.x86_64' make: ***
>>>>> [default] Error 2
>>>> 
>>>> 
>>>>> I am running: Linux 2.6.32-504.el6.x86_64 #1 SMP Tue Sep 16
>>>>>  01:56:35 EDT 2014 x86_64 x86_64 x86_64 GNU/Linux
>>>> 
>>>>> Scott Finlon, CISSP GCIA GCIH
>>>>> ----------------------------------- Information Security
>>>>> Engineer The University of Scranton email :
>>>>> [email protected] phone : 570-941-6168 
>>>>> -----------------------------------
>>>> 
>>>> 
>>>>> From: "[email protected]<mailto:[email protected]>" 
>>>>> <[email protected]<mailto:[email protected]>>
>>>>> Reply-To: 
>>>>> "[email protected]<mailto:[email protected]>"
>>>>>
>>>>>
>
>>>>> 
<[email protected]<mailto:[email protected]>>
>>>> 
>>>> 
>>>> Date: Thursday, October 16, 2014 at 12:06 PM
>>>>> To: 
>>>>> "[email protected]<mailto:[email protected]>"
>>>>>
>>>>>
>
>>>>> 
<[email protected]<mailto:[email protected]>>
>>>> 
>>>> 
>>>> Subject: Re: [Ntop-misc] Error compiling ixgbe drivers in
>>>> RHEL
>>>> 
>>>>> Hi Scott probably something has changed in your kernel
>>>>> version, please try also with ixgbe-3.22.3-zc to make sure
>>>>> it has not been already fixed.
>>>> 
>>>>> Best Regards Alfredo
>>>> 
>>>>> On 16 Oct 2014, at 18:00, Scott Finlon 
>>>>> <[email protected]<mailto:[email protected]>>
>>>>>  wrote:
>>>> 
>>>>> I upgrade the kernel on my RHEL boxes today, and when I 
>>>>> attempted to recompile the ixgbe drivers I was met with
>>>>> the following error:
>>>> 
>>>>> make -C /lib/modules/2.6.32-504.el6.x86_64/build 
>>>>> SUBDIRS=/usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src
>>>>>  modules make[1]: Entering directory 
>>>>> `/usr/src/kernels/2.6.32-504.el6.x86_64' CC [M] 
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_main.o
>>>>
>>>>>
>>>>
>>>>
>>>>> 
In file included from
>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_osdep.h:37,
>>>>>
>>>>
>
>>>> 
from
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_type.h:65,
>>>>
>>>>>
>>>>
>>>>
>>>>> 
from
>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_dcb.h:28,
>>>>>
>>>>
>
>>>> 
from
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe.h:47,
>>>>
>>>>>
>>>>
>>>>
>>>>> 
from
>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_main.c:50:
>>>>>
>>>>
>
>>>> 
/usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/kcompat.h:4063:
>>>>> error: nested redefinition of ‘enum pkt_hash_types’ 
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/kcompat.h:4063:
>>>>>
>>>>>
>
>>>>> 
error: redeclaration of ‘enum pkt_hash_types’
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/kcompat.h:4064:
>>>>>
>>>>>
>
>>>>> 
error: redeclaration of enumerator ‘PKT_HASH_TYPE_NONE’
>>>>> include/linux/skbuff.h:613: note: previous definition of 
>>>>> ‘PKT_HASH_TYPE_NONE’ was here 
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/kcompat.h:4065:
>>>>>
>>>>>
>
>>>>> 
error: redeclaration of enumerator ‘PKT_HASH_TYPE_L2’
>>>>> include/linux/skbuff.h:614: note: previous definition of 
>>>>> ‘PKT_HASH_TYPE_L2’ was here 
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/kcompat.h:4066:
>>>>>
>>>>>
>
>>>>> 
error: redeclaration of enumerator ‘PKT_HASH_TYPE_L3’
>>>>> include/linux/skbuff.h:615: note: previous definition of 
>>>>> ‘PKT_HASH_TYPE_L3’ was here 
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/kcompat.h:4067:
>>>>>
>>>>>
>
>>>>> 
error: redeclaration of enumerator ‘PKT_HASH_TYPE_L4’
>>>>> include/linux/skbuff.h:616: note: previous definition of 
>>>>> ‘PKT_HASH_TYPE_L4’ was here 
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_main.c:9371:
>>>>>
>>>>>
>
>>>>> 
error: unknown field ‘ndo_busy_poll’ specified in initializer
>>>>> /usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_main.c:9371:
>>>>>
>>>>>
>
>>>>> 
warning: initialization from incompatible pointer type make[2]: ***
>>>>> [/usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src/ixgbe_main.o]
>>>>>
>>>>>
>
>>>>> 
Error 1 make[1]: ***
>>>>> [_module_/usr/local/src/PF_RING/drivers/DNA/ixgbe-3.21.2-DNA/src]
>>>>>
>>>>>
>
>>>>> 
Error 2 make[1]: Leaving directory
>>>>> `/usr/src/kernels/2.6.32-504.el6.x86_64' make: ***
>>>>> [default] Error 2
>>>> 
>>>>> I see Ubuntu users had a similar problem back in August
>>>>> with the OS back porting stuff into the kernel that wasn’t
>>>>> officially there. Is there a similar patch for RHEL yet?
>>>> 
>>>>> Thanks! -Scott
>>>> 
>>>> 
>>>>> Scott Finlon, CISSP GCIA GCIH
>>>>> ----------------------------------- Information Security
>>>>> Engineer The University of Scranton email :
>>>>> [email protected]<mailto:[email protected]>
>>>>>
>>>>> 
phone : 570-941-6168 -----------------------------------
>>>> 
>>>>> _______________________________________________ Ntop-misc 
>>>>> mailing list 
>>>>> [email protected]<mailto:[email protected]>
>>>>
>>>>>
>>>>
>>>>
>>>>> 
http://listgateway.unipi.it/mailman/listinfo/ntop-misc
>>>> 
>>>> 
>>>> 
>>>> 
>>>>> _______________________________________________ Ntop-misc 
>>>>> mailing list [email protected] 
>>>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc
>>>> 
>>>> 
>>>> 
>>>> _______________________________________________ Ntop-misc
>>>> mailing list [email protected] 
>>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc
>>>> 
> 
>> _______________________________________________ Ntop-misc mailing
>> list [email protected] 
>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc
> 
> _______________________________________________ Ntop-misc mailing
> list [email protected] 
> http://listgateway.unipi.it/mailman/listinfo/ntop-misc
> 

- -- 
- -------------------------------------------------------------------------
John Ives
Information Security & Policy                       Phone (510) 229-8676
University of California, Berkeley
- -------------------------------------------------------------------------
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iQEcBAEBAgAGBQJUYTtzAAoJEJkidK6qbyws6+kIAIsvg8TD3xyPJh28n61fakpV
tJ+YWIiOE/3/YPnb5jHIKtzr7bEZfqg+aIduQcAIxxquDlVC0SRRzfNikSntgK7D
ZyBF5xvwRlspYbhh8yBibh0qwhU++OesGzeRqTOL3mcjzAeEuLdWpTJ3HnGOm4h3
En8tYDiEWExyzDRijI9GJZnJfKv3aRv/p2UGE0gBsBxNxswag2vkNDlDPaA+KSzx
owA4R/SEDBsLVD17fb13Hl+NeF0tLoOxax2W06mhsvqIVgzRQuD1gcPwGSmDqi1N
Rj/uFsfctETRyXBRUBVsPpi2989/ttCt2JGwHYK+FhgSWDzF6AzsT/PewGfmSBw=
=sgpw
-----END PGP SIGNATURE-----
_______________________________________________
Ntop-misc mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-misc

Reply via email to