On Tue, Aug 13, 2019 at 8:03 PM Darrell Ball wrote:
> Thanks for the patch
>
> few more comments
>
> On Mon, Aug 12, 2019 at 5:57 PM Yi-Hung Wei wrote:
>
>> This patch derives the timeout policy based on ct zone from the
>> internal data structure that we maintain on dpif layer.
>>
>> It also ad
Thanks for the patch
few more comments
On Mon, Aug 12, 2019 at 5:57 PM Yi-Hung Wei wrote:
> This patch derives the timeout policy based on ct zone from the
> internal data structure that we maintain on dpif layer.
>
> It also adds a system traffic test to verify the zone-based conntrack
> timeo
Thanks for the patch
Some high level comments:
1/ The ct_tp_kill_list code is still in common code
I think we discussed moving that to the dpif backer code
ct_timeout_policy_unref() is adding to this deferred kill list which is
not needed for userspace
datapath.
2/ clear_existing_ct_
On Tue, Aug 13, 2019 at 7:07 PM aginwala wrote:
> Thanks for the fixes. Found a bug in current code which breaks both nbctl
> with local socket and daemon mode on follower nodes. Also nbctl daemon mode
> always tries to connect to leader node by default which also failed to
> connect.
> e.g. expo
Bleep bloop. Greetings Ankur Sharma, I am a robot and I have tried out your
patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
build:
gcc -std=gnu99 -DHAVE_CONFIG_H -I. -I ./include -I ./ovs/include -I
./ovs/include -I ./ovs/lib -
Background:
With c0974331b7a19a87ab8f1f2cec8fbe366af92fa2, we have added
support for E-W workflow for vlan backed DVRs.
This series enables N-S workflow for vlan backed DVRs.
Key difference between E-W and N-S traffic flow is that
N-S flow requires a gateway chassis. A gateway chassis
will be res
With 795d7f24ce0e2ed5454e193a059451d237289542 we have added
support for E-W routing on vlan backed networks by replacing
router port macs with chassis macs.
This replacement of router port mac need NOT be done on
gateway chassis for following reasons:
a. For N-S traffic, gateway chassis will resp
Problem Description:
Redirection to chassisredirect ports happens only via tunnel
encapsulation. As a result, same cannot be leveraged upon for vlan backed
environments.
This series addresses the issue by allowing redirection to happen without
encapsulation.
Design:
===
a. High level design
Background:
With c0974331b7a19a87ab8f1f2cec8fbe366af92fa2, we have added
support for E-W workflow for vlan backed DVRs.
This series enables N-S workflow for vlan backed DVRs.
Key difference between E-W and N-S traffic flow is that
N-S flow requires a gateway chassis. A gateway chassis
will be res
Background:
With c0974331b7a19a87ab8f1f2cec8fbe366af92fa2, we have added
support for E-W workflow for vlan backed DVRs.
This series enables N-S workflow for vlan backed DVRs.
Key difference between E-W and N-S traffic flow is that
N-S flow requires a gateway chassis. A gateway chassis
will be res
On Tue, Aug 13, 2019 at 2:33 PM Yi-Hung Wei wrote:
> On Tue, Aug 13, 2019 at 11:43 AM Darrell Ball wrote:
> > Sure, circling back to this part
> >
> > yep, it is the Linux In-tree kernel module rather than OVS tree module
> >
> > dball@ubuntu:~/ovs$ modinfo openvswitch
> > filename:
> /lib/
On Tue, Aug 13, 2019 at 2:33 PM Yi-Hung Wei wrote:
>
> On Tue, Aug 13, 2019 at 11:43 AM Darrell Ball wrote:
> > btw, similarly
> > make 'check-kernel' fails for the same reasons.
> >
> > Ostensibly, I would have expected 5.0 to be ok.
> > I can dig more on this part later if you wish.
>
> The ct
On Tue, Aug 13, 2019 at 11:43 AM Darrell Ball wrote:
> Sure, circling back to this part
>
> yep, it is the Linux In-tree kernel module rather than OVS tree module
>
> dball@ubuntu:~/ovs$ modinfo openvswitch
> filename:
> /lib/modules/5.0.0-23-generic/kernel/net/openvswitch/openvswitch.k
On Tue, Aug 13, 2019 at 4:25 AM wrote:
> > ---
> > Documentation/faq/releases.rst | 3 +-
> > datapath-windows/include/OvsDpInterfaceCtExt.h | 114 +
> > datapath-windows/ovsext/Netlink/NetlinkProto.h | 8 +-
> > include/windows/automake.mk
On Wed, Aug 7, 2019 at 3:25 PM Yifeng Sun wrote:
>
> This patch backports several critical bug fixes related to
> locking and data consistency in nf_conncount code.
>
> This backport is based on the following upstream net-next upstream commits.
> a007232 ("netfilter: nf_conncount: fix argument ord
--
Hallo,
Ich bin Azim Hashim Premji, ein indischer Wirtschaftsmagnat, Investor
und Philanthrop. Ich bin der Vorsitzende von Wipro Limited. Ich gab 25
Prozent meines persönlichen Vermögens für wohltätige Zwecke ab. Und ich
habe auch versprochen, den Rest von 25% dieses Jahr 2018 an
Einzelp
On Tue, Aug 13, 2019 at 9:23 AM Han Zhou wrote:
>
> From: Han Zhou
>
> As mentioned in RAFT paper, section 6.2:
>
> Leaders: A server might be in the leader state, but if it isn’t the
current
> leader, it could be needlessly delaying client requests. For example,
suppose a
> leader is partitioned
On Tue, Aug 13, 2019 at 11:01 AM Yi-Hung Wei wrote:
> On Mon, Aug 12, 2019 at 7:35 PM Darrell Ball wrote:
> >
> > Thanks for the patch
> >
> > Not a full review; I just did a quick run of the test using a more
> recent kernel version
> >
> > dball@ubuntu:~/ovs$ uname -r
> > 5.0.0-23-generic
> >
On Mon, Aug 12, 2019 at 7:35 PM Darrell Ball wrote:
>
> Thanks for the patch
>
> Not a full review; I just did a quick run of the test using a more recent
> kernel version
>
> dball@ubuntu:~/ovs$ uname -r
> 5.0.0-23-generic
> dball@ubuntu:~/ovs$ lsb_release -a
> No LSB modules are available.
> Di
Bleep bloop. Greetings Numan Siddique, I am a robot and I have tried out your
patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
checkpatch:
WARNING: Line is 108 characters long (recommended limit is 79)
#270 FILE: rhel/usr_lib_system
Bleep bloop. Greetings Numan Siddique, I am a robot and I have tried out your
patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
checkpatch:
WARNING: Line is 81 characters long (recommended limit is 79)
#141 FILE: .travis/osx-build.sh
Bleep bloop. Greetings Numan Siddique, I am a robot and I have tried out your
patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
checkpatch:
WARNING: Line is 99 characters long (recommended limit is 79)
#650 FILE: utilities/ovn-ctl:34
On 13 Aug 2019, at 18:37, Ilya Maximets wrote:
This is highly useful to see on which core PMD is running by
only looking at the thread name. Thread Id still allows to
distinguish different threads running on the same core over the time:
|dpif_netdev(pmd-c10/id:53)|DBG|Creating 2. subtable
On Tue, Aug 13, 2019 at 01:34:04PM -0300, Flavio Leitner via dev wrote:
> When a packet needs to be encapsulated in userspace, the endpoint
> address needs to be resolved to fill in the headers. If it is not,
> then currently OvS sends either a Neighbor Solicitation (IPv6)
> or an ARP Query (IPv4)
This is highly useful to see on which core PMD is running by
only looking at the thread name. Thread Id still allows to
distinguish different threads running on the same core over the time:
|dpif_netdev(pmd-c10/id:53)|DBG|Creating 2. subtable <...>
|dpif_netdev(pmd-c10/id:53)|DBG|flow_add: <
When a packet needs to be encapsulated in userspace, the endpoint
address needs to be resolved to fill in the headers. If it is not,
then currently OvS sends either a Neighbor Solicitation (IPv6)
or an ARP Query (IPv4) to resolve it.
The problem is that the NS/ARP packet will go through the flow
r
From: Numan Siddique
This patch could have created a new user 'ovn' for ovn services instead
of using 'openvswitch' user. But this would require some amount of work and
proper testing since the new user 'ovn' should be part of 'openvswitch'
group (to access /var/run/openvswitch/db.sock.). If ovs
From: Numan Siddique
"make rpm-fedora" is broken and this patch fixes it. Previous patch
in this series supported building OVN from external OVS sources.
Before running "make rpm-fedora", it is expected that the developer has run
"make dist" in the OVS source folder to generate the
openvswitch-
From: Numan Siddique
With this patch we have to configure OVN to refer to external OVS source/build
directory instead of the ovs subtree.
The new configuration options added are:
* --with-ovs-source=/path/to/ovs/source/dir
* --with-ovs-build=/path/to/ovs/build/dir
Before configuring OVN, user
From: Numan Siddique
Until now, OVN uses the openvswitch rundirs (rundir, logdir, etcdir).
The commit [1] changed the package name from openvswitch to ovn, but
it didn't take into the account the effects of it. When "make install"
is run ovn-ctl utility is copied to /usr/local/share/ovn/scripts f
From: Numan Siddique
This patch series adds support for building OVN from external OVS
sources.
The first patch adds support to run OVN services using separate
directores
- Default run time dir - /usr/local/var/run/ovm
- Default log dir - /usr/loca/var/log/ovn
- Default db dir - /usr/loca
From: Han Zhou
As mentioned in RAFT paper, section 6.2:
Leaders: A server might be in the leader state, but if it isn’t the current
leader, it could be needlessly delaying client requests. For example, suppose a
leader is partitioned from the rest of the cluster, but it can still
communicate wit
From: Han Zhou
candiate_retrying is used to determine if the current node is disconnected
from the cluster when the node is in candiate role. However, a node
can flap between candidate and follower role before a leader is elected
when majority of the cluster is down, so is_connected() will flap,
From: Han Zhou
When clustered mode is used, the client needs to retry connecting
to new servers when certain failures happen. Today it is allowed to
retry new connection only if multiple remotes are used, which prevents
using LB VIP with clustered nodes. This patch makes sure the retry
logic work
From: Han Zhou
Fixes: commit 5a9b53a5 ("ovsdb raft: Fix duplicated transaction execution when
leader failover.")
Signed-off-by: Han Zhou
---
ovsdb/raft.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/ovsdb/raft.c b/ovsdb/raft.c
index c60ef41..1c38b3b 100644
--- a/ovsdb/ra
This allows to decrease code duplication and avoid using Linux-specific
functions (this might be useful in the future if we'll try to allow
running OvS+DPDK on FreeBSD).
Signed-off-by: Ilya Maximets
---
lib/dpdk.c | 27 ---
1 file changed, 12 insertions(+), 15 deletions(-
Unlike 'rte_get_tsc_cycles()' which doesn't need any specific
initialization, 'rte_get_tsc_hz()' could be used only after successfull
call to 'rte_eal_init()'. 'rte_eal_init()' estimates the TSC frequency
for later use by 'rte_get_tsc_hz()'. Fairly said, we're not allowed
to use 'rte_get_tsc_cycle
New functions to get and set CPU affinity using CPU dumps.
This will abstract OS specific implementation details from the
cross-platform code.
Signed-off-by: Ilya Maximets
---
lib/ovs-numa.c | 74 +-
lib/ovs-numa.h | 2 ++
2 files changed, 75 inse
Ilya Maximets (3):
ovs-numa: Add dump based thread affinity functions.
dpif-netdev-perf: Fix TSC frequency for non-DPDK case.
dpdk: Use ovs-numa provided functions to manage thread affinity.
lib/dpdk-stub.c| 6
lib/dpdk.c | 33 ++-
lib/dpdk.h
On 07.08.2019 17:21, Nitin Katiyar wrote:
> Each PMD updates the global sequence number for RCU synchronization
> purpose with other OVS threads. This is done at every 1025th iteration
> in PMD main loop.
>
> If the PMD thread is responsible for polling large number of queues
> that are carrying t
Передайте это сообщение начальнику по продажам
---
Если вашей фирме нужна свежая база данных потенциальных клиентов для обзвона
или рассылки, закажите базу данных по этому адресу http://bzonl.info/db/
--
А также приглашаем воспользоваться предложением возврат
> -Original Message-
> From: ovs-dev-boun...@openvswitch.org boun...@openvswitch.org> On Behalf Of Justin Pettit
> Sent: Tuesday, August 6, 2019 8:28 AM
> To: Yi-Hung Wei
> Cc: ovs dev ; Alin Gabriel Serdean
>
> Subject: Re: [ovs-dev] [PATCH v2 4/9] ct-dpif, dpif-netlink: Add conntrac
> -Original Message-
> From: ovs-dev-boun...@openvswitch.org boun...@openvswitch.org> On Behalf Of Yi-Hung Wei
> Sent: Tuesday, August 13, 2019 3:52 AM
> To: d...@openvswitch.org
> Subject: [ovs-dev] [PATCH v3 4/9] ct-dpif, dpif-netlink: Add conntrack
> timeout policy support
>
> This pat
Bleep bloop. Greetings Numan Siddique, I am a robot and I have tried out your
patch.
Thanks for your contribution.
I encountered some error that I wasn't expecting. See the details below.
checkpatch:
WARNING: Line is 99 characters long (recommended limit is 79)
#635 FILE: utilities/ovn-ctl:34
When a packet is received, the RSS hash is calculated if it is not
already available. The Exact Match Cache (EMC) entry is then looked up
using this RSS hash.
When a MPLSoGRE encapsulated packet is received, the GRE header is
popped, the RSS hash is invalidated and the packet is recirculated for
f
From: Numan Siddique
Until now, OVN uses the openvswitch rundirs (rundir, logdir, etcdir).
The commit [1] changed the package name from openvswitch to ovn, but
it didn't take into the account the effects of it. When "make install"
is run ovn-ctl utility is copied to /usr/local/share/ovn/scripts f
Thanks! Applied on master and backported back until branch-2.10.
Alin.
From: Anand Kumar
Sent: Thursday, August 1, 2019 5:41 AM
To: Alin Serdean ; d...@openvswitch.org
Cc: Danting Liu
Subject: Re: [ovs-dev] [PATCH] [windows][wmi] Switch from internal port to all
ports defined
Acked-
>
> Hi,
>
> why not a static function in the header file? So it gets inlined.
>
> Regards,
> --
> Matteo Croce
> per aspera ad upstream
Thanks Matteo for looking into this patch-set. Yes I agree. I will address your
suggestion in the next revision.
Warm Regards,
Vishal
_
48 matches
Mail list logo