In "MAC flooding", an attacker transmits an overwhelming number of frames
with unique Ethernet source address on a switch port. The goal is to
force the switch to evict all useful MAC learning table entries, so that
its behavior degenerates to that of a hub, flooding all traffic. In turn,
that al
Do you see symbol conflicts? If not, then there's no real problem--the
linker will discard the unused object files.
On Thu, Feb 12, 2015 at 04:08:55AM +, Papudippu, Sreedhar Reddy wrote:
> Hi Ben,
>
> I am trying to write few applications which are pure clients to OVSDB. I need
> to link wi
Hi Ben,
I am trying to write few applications which are pure clients to OVSDB. I need
to link with OVSDB-IDL.
Currently the only way I can link with ovsdb-idl is by linking the with
lib/libopenvswitch.la
Currently lib/libopenvswitch.la has sFlow, LACP, netdev-provider etc code
together.
When I
Vagrant sounds like a real winner. Thanks for working on the kernel
testing infrastructure. I hope that we can start to build up a library
of tests.
On Wed, Feb 11, 2015 at 06:04:31PM -0800, Andy Zhou wrote:
> I have experimented with the user-mode linux about a month back. I was
> not able to g
What kind of problems do you have with the rest of the library? Maybe
we should focus on that.
On Thu, Feb 12, 2015 at 02:56:15AM +, Papudippu, Sreedhar Reddy wrote:
> Hi Ben,
>
> I do understand that making it as a Shared library brings in lot of troubles,
> due to the decoupling.
> But I
testsuite uses mkfifo in its job dispatcher that manages
parallel unit tests. MinGW does not have a mkfifo. This
results in unit tests running serially on Windows. Right
now it takes up to approximately 40 minutes to run all the
unit tests on Windows.
This commit provides a job dispatcher for MinG
Hi Ben,
I do understand that making it as a Shared library brings in lot of troubles,
due to the decoupling.
But I am trying to make it as Static library (ovsdb-idl.a), so that if an
external tool/application use the Ovsdb-ild library without carrying the rest
of the baggage from open-switch/li
>
> Thanks for doing this. I really like changes that make it easier or
> faster to run tests, because then developers are more likely to run
> them.
>
> I don't think that "patch" is currently required to build OVS. I think
> that it's reasonable to require it--it is such a common utility--but I
I have experimented with the user-mode linux about a month back. I was
not able to get ovs user space to run reliably with the
(light weight) host-fs file system. It may run better with a
disk-image (I did not try), but then we willl have to deal with
building
and distributing disk-images, which Va
It's difficult to design and maintain a shared library with a stable
public API and ABI. No one has done the work yet.
On Tue, Feb 10, 2015 at 07:25:47PM +, Papudippu, Sreedhar Reddy wrote:
> Hello all,
>
> I have been wondering why doesn open-vswitch code links the entire
> "lib/libopenvs
On Wed, Feb 11, 2015 at 02:16:51PM -0800, Gurucharan Shetty wrote:
> testsuite uses mkfifo in its job dispatcher that manages
> parallel unit tests. MinGW does not have a mkfifo. This
> results in unit tests running serially on Windows. Right
> now it takes up to approximately 40 minutes to run all
On Wed, Feb 11, 2015 at 4:09 PM, Nithin Raju wrote:
>> On Feb 11, 2015, at 2:15 PM, Gurucharan Shetty wrote:
>>
>> OVS daemon service for Windows creates the pidfile and then
>> registers with the Windows services manager that the service
>> is running. There is a small time gap between the two s
On Tue, Feb 10, 2015 at 4:38 PM, Nithin Raju wrote:
> _BitScanForward() and friends are part of the Windows API and
> take DWORD as parameter type. DWORD is defined to be 'unsigned long'
> in Windows' header files.
>
> We call into these functions from within lib/util.h. Currently, we
> pass argum
> On Feb 11, 2015, at 2:15 PM, Gurucharan Shetty wrote:
>
> OVS daemon service for Windows creates the pidfile and then
> registers with the Windows services manager that the service
> is running. There is a small time gap between the two steps.
> So retry a few times in the test.
>
> Also, prov
On Mon, Feb 9, 2015 at 7:54 AM, Thomas Graf wrote:
> vxlan: Only set has-GBP bit in header if any other bits would be set
>
> This allows for a VXLAN-GBP socket to talk to a Linux VXLAN socket by
> not setting any of the bits.
>
> Signed-off-by: Thomas Graf
> Signed-off-by: Da
On Sat, Feb 07, 2015 at 01:04:57PM +0100, Thomas Graf wrote:
> On 02/06/15 at 10:28pm, Ben Pfaff wrote:
> > We already do a ton of kernel module builds in travis, do you mean that
> > we should do one for every released kernel, or do you mean something
> > else?
>
> I'm primarily thinking of Andy'
testsuite uses mkfifo in its job dispatcher that manages
parallel unit tests. MinGW does not have a mkfifo. This
results in unit tests running serially on Windows. Right
now it takes up to approximately 40 minutes to run all the
unit tests on Windows.
This commit provides a job dispatcher for MinG
If IPv6 link-local address is removed from interface, it is unable to
receive any IPv6 packets, including Route Advertisements.
In save_interface only skip IPv4 "scope link" addresses.
Signed-off-by: Alexey I. Froloff
---
utilities/ovs-save | 5 +++--
1 file changed, 3 insertions(+), 2 deletion
Dear Sir/Madam !
Hope you are fine? My name is Alice Joe from London UK, Please kindly reply so
i can fully introduce myself and also share the reasons why i contacted you.
Thanks best regards.
Alice.
___
dev mailing list
dev@openvswitch.org
http://ope
That's great, I'm looking forward to the new version.
Let me know if you can't figure out the test framework. The best
place to start on it is the "Using Autotest" chapter in the Autoconf
manual.
On Tue, Feb 10, 2015 at 02:37:29PM +, Flynn, Dennis R (Dennis) wrote:
> Ben,
>
> My apologies f
OVS daemon service for Windows creates the pidfile and then
registers with the Windows services manager that the service
is running. There is a small time gap between the two steps.
So retry a few times in the test.
Also, provide a keyword for the test.
Reported-by: Nithin Raju
Signed-off-by: Gu
_BitScanForward() and friends are part of the Windows API and
take DWORD as parameter type. DWORD is defined to be 'unsigned long'
in Windows' header files.
We call into these functions from within lib/util.h. Currently, we
pass arguments of type uint32_t which is type defined to
'unsigned int'. T
Unit tests pass except for 3 failures. 2 of them seem expected, and the third
one is with running ovsdb-server.exe as a Windows service. This does not seem
to be related to my change, and the failure seen without my change as well.
I’ll debug that as a separate issue.
thanks,
— Nithin
> On Feb
It's not the ofpbufs that are leaked. Where is 's' freed? Before your
patch, the data in 's' was put into an ofpbuf, and eventually freed.
After your patch, the data in 's' is always copied a second time, by
ofpbuf_clone_data(), and then the copy is put into an ofpbuf (and
eventually freed), but
vxlan: Only set has-GBP bit in header if any other bits would be set
This allows for a VXLAN-GBP socket to talk to a Linux VXLAN socket by
not setting any of the bits.
Signed-off-by: Thomas Graf
Signed-off-by: David S. Miller
Upstream: db79a621835e ("vxlan: Only set has-GBP
Ben,
My apologies for the delay getting back to you on this topic.
I am working on getting the tests you requested in place. As you suggest I will
focus on adding a few tests to validate that we are sending well-formed LLDP
packets. Specifically I will concentrate on validating the TLVs contain
On 02/11/15 at 11:23am, Geert Uytterhoeven wrote:
> net/openvswitch/flow_netlink.c: In function ‘validate_and_copy_set_tun’:
> net/openvswitch/flow_netlink.c:1749: warning: ‘err’ may be used uninitialized
> in this function
>
> If ipv4_tun_from_nlattr() returns a different positive value than
> O
Most of the information that timeval was reporting for long poll intervals
was comparing per-thread with per-process statistics, which yielded
nonsense a lot of the time.
Signed-off-by: Ben Pfaff
---
lib/timeval.c | 38 +-
1 file changed, 21 insertions(+), 1
On 02/04/15 at 04:45pm, Thomas Graf wrote:
> From: Madhu Challa
>
> Introduces two new NXMs to represent VXLAN-GBP [0] fields.
>
> actions=load:0x10->NXM_NX_TUN_GBP_ID[],NORMAL
> tun_gbp_id=0x10,actions=drop
>
> This enables existing VXLAN tunnels to carry security label
> information such
Pravin,
Based on my test and Kevin's confirmation, I think it is safe to apply this
patch for now. What do you think?
Thanks,
Alex Wang,
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
From: Jarno Rajahalme
Date: Thu, 5 Feb 2015 13:40:49 -0800
> OVS userspace already probes the openvswitch kernel module for
> OVS_ACTION_ATTR_SET_MASKED support. This patch adds the kernel module
> implementation of masked set actions.
>
> The existing set action sets many fields at once. Whe
Caveat: Unit tests are running and not complete yet. I’ll reply back once they
are complete.
thanks,
-- Nithin
> On Feb 10, 2015, at 4:38 PM, Nithin Raju wrote:
>
> _BitScanForward() and friends are part of the Windows API and
> take DWORD as parameter type. DWORD is defined to be 'unsigned lo
Processing commands for cont...@bugs.debian.org:
> reassign 763428 linux-image-3.16.0-4-amd64
Bug #763428 [openvswitch-switch] openvswitch-switch: openvswitch doesn't work
anymore since kernel 3.16 update
Bug reassigned from package 'openvswitch-switch' to
'linux-image-3.16.0-4-amd64'.
No longer
On Fri, Feb 06, 2015 at 11:14:13PM -0800, Ben Pfaff wrote:
> On Tue, Feb 03, 2015 at 11:53:29PM -0800, Shu Shen wrote:
> > The previous workflow is to convert dot diagrams into .pic format and
> > embed into manpages; double borders and arrows were not used in dot but
> > introduced in .pic convers
Hi Ben,
> > This commit enables the eviction mechanism on the basis of
> > importance as
> > per the openflow specification 1.4.
> >
> > ovs-ofctl -O OpenFlow14 mod-table evict
> > -Enable eviction on of . Eviction adds a mechanism
> > enabling the switch to automatically eliminate entrie
您好||
dev@openvswitch.org
您在做外贸？您的客户资源太少？解决方案如下：
国外客户主动开发方式--双喜软件
是一款外贸客户联系方式搜索软件
功能：关键词搜索网站
设置你们产品的关键词或者你们的目标客户的关键词
进入全球互联网上搜索您的目标客户信息
功能：智能邮件推广
提取客户邮箱信息，软件自动发送开发信联系客户
客户有意向采购您的产品就会回复询盘给您.
（软件搜索-不限制关键词，不限制国家引擎）
（软件群发-多主题 多内容 多帐号 防止屏蔽）
联系我们观看软件操作演示QQ服务号码（137 895 8561）
Hello all,
I have been wondering why doesn open-vswitch code links the entire
"lib/libopenvswitch.la" into every binary in the code.
Isn't that very in-efficient? Shouldn't we create separate libraries based on
some logical divide, and then link the libraries based on the need.
We should have
Thanks! I applied to master and branch-2.[310].
On Wed, Feb 11, 2015 at 02:59:59PM -0800, Alex Wang wrote:
> Thx for the fix!
>
> Acked-by: Alex Wang
>
> On Wed, Feb 11, 2015 at 2:54 PM, Ben Pfaff wrote:
>
> > Most of the information that timeval was reporting for long poll intervals
> > was
Thx for the fix!
Acked-by: Alex Wang
On Wed, Feb 11, 2015 at 2:54 PM, Ben Pfaff wrote:
> Most of the information that timeval was reporting for long poll intervals
> was comparing per-thread with per-process statistics, which yielded
> nonsense a lot of the time.
>
> Signed-off-by: Ben Pfaff
Most of the information that timeval was reporting for long poll intervals
was comparing per-thread with per-process statistics, which yielded
nonsense a lot of the time.
Signed-off-by: Ben Pfaff
---
v1->v2: Really fix the whole problem (thanks Alex!)
v2->v2.1: Repost now that the list is back up
On Wed, Feb 11, 2015 at 2:23 AM, Geert Uytterhoeven
wrote:
> net/openvswitch/flow_netlink.c: In function ‘validate_and_copy_set_tun’:
> net/openvswitch/flow_netlink.c:1749: warning: ‘err’ may be used uninitialized
> in this function
>
> If ipv4_tun_from_nlattr() returns a different positive value
I keep getting a bounce on rory.sex...@intel.com ("550 #5.1.0 Address
rejected."). Is that email address correct?
On Mon, Feb 9, 2015 at 8:16 AM, Ben Pfaff wrote:
> It's not the ofpbufs that are leaked. Where is 's' freed? Before your
> patch, the data in 's' was put into an ofpbuf, and eventua
Hi Pravin,
The tests fail because the amount of data that an 'mbuf-type' ofpbuf can store
is now restricted to UINT16_MAX.
In the failed tests, the amount of data added to the ofpbuf far exceeds this
limit; resolving this shortcoming will require modifications to additional lib
functions (in
net/openvswitch/flow_netlink.c: In function ‘validate_and_copy_set_tun’:
net/openvswitch/flow_netlink.c:1749: warning: ‘err’ may be used uninitialized
in this function
If ipv4_tun_from_nlattr() returns a different positive value than
OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS, err will be uninitialized, and
On Tue, Feb 10, 2015 at 2:27 AM, Kavanagh, Mark B
wrote:
> Hi Pravin,
>
> The tests fail because the amount of data that an 'mbuf-type' ofpbuf can
> store is now restricted to UINT16_MAX.
>
> In the failed tests, the amount of data added to the ofpbuf far exceeds this
> limit; resolving this sho
From: Saloni Jain
This commit enables the eviction mechanism on the basis of importance as
per the openflow specification 1.4.
ovs-ofctl -O OpenFlow14 mod-table evict
-Enable eviction on of . Eviction adds a mechanism
enabling the switch to automatically eliminate entries of lower
i
Thx everyone for the help and review,
I apply the patch with following comment:
diff --git a/lib/netdev-dpdk.h b/lib/netdev-dpdk.h
index 9a47165..694899c 100644
--- a/lib/netdev-dpdk.h
+++ b/lib/netdev-dpdk.h
@@ -5,6 +5,9 @@
struct dpif_packet;
+/* Reserves cpu core 0 for all non-pmd threads.
On Mon, Feb 9, 2015 at 2:24 PM, Alex Wang wrote:
> Pravin,
>
> Based on my test and Kevin's confirmation, I think it is safe to apply this
> patch for now. What do you think?
>
If you already have done test with CONFIG_RTE_LIBRTE_MBUF_DEBUG then I
am ok with the patch.
Acked-by: Pravin B Shelar
Hi Ben,
It was deactivated this month, but the former owner made some contributions
towards this patch - I'll remove it from the thread going forward but will
leave it in the commit.
Thanks,
Mark
> -Original Message-
> From: Ben Pfaff [mailto:b...@nicira.com]
> Sent: Monday, February 9
Patches that modify existing code can break expected behaviour.
Flag this by testing the patch with 'make check' prior to submission.
Furthermore, it is not sufficient to only test patches that add files
using 'make distcheck';the compile flags for this target could change
the definition of some f
On Tue, Feb 10, 2015 at 03:19:18PM -0800, Shu Shen wrote:
> On Fri, Feb 06, 2015 at 11:14:13PM -0800, Ben Pfaff wrote:
> > On Tue, Feb 03, 2015 at 11:53:29PM -0800, Shu Shen wrote:
> > > The previous workflow is to convert dot diagrams into .pic format and
> > > embed into manpages; double borders
Hi Ben,
I don't think so - jsonrpc_run() empties the rpc output queue, freeing the
ofpbufs, after their contents are sent to the stream.
Thanks,
Mark
> -Original Message-
> From: Ben Pfaff [mailto:b...@nicira.com]
> Sent: Friday, February 6, 2015 8:41 PM
> To: Kavanagh, Mark B
> Cc: dev
Multiple weak edges between nodes at the same rank overlaps with each other in
a dot/graphviz diagram. The vswitchd.pic used in ovs-vswitchd.conf.db.5 suffers
this problem.
Removing "constraint=false" allows graphviz to rank the nodes using the weak
edages as well so that the nodes at the ends of
Hello all,
I was wondering why "ovsdb-idl" client API code is not made as a separate
independent library.
Assume if someone wants to write a new client for the ovsdb-idl, how will
he/she link the code to the source library?
How will he/she get the get the header file definitions?
Regards
Sreed
From: Geert Uytterhoeven
Date: Wed, 11 Feb 2015 11:23:38 +0100
> net/openvswitch/flow_netlink.c: In function ‘validate_and_copy_set_tun’:
> net/openvswitch/flow_netlink.c:1749: warning: ‘err’ may be used uninitialized
> in this function
>
> If ipv4_tun_from_nlattr() returns a different positive
55 matches
Mail list logo
