On 29 Nov 2021, at 8:38, Chris Mi wrote:
> On 11/25/2021 4:52 PM, Eelco Chaudron wrote:
>>
>> On 15 Nov 2021, at 3:53, Chris Mi wrote:
>>
>>> Some offload actions require functionality that is not netdev
>>> based, but dpif. For example, sFlow action requires to create
>>> a psample netlink
Signed-off-by: Nobuhiro MIKI
---
Documentation/intro/install/afxdp.rst | 1 -
1 file changed, 1 deletion(-)
diff --git a/Documentation/intro/install/afxdp.rst
b/Documentation/intro/install/afxdp.rst
index 6ee69a15f..bfef49860 100644
--- a/Documentation/intro/install/afxdp.rst
+++
On 11/17/21 15:08, Aaron Conole wrote:
> Mike Pattrick writes:
>
>> Currently the ovs-tcpdump utility creates a virtual tunnel to send
>> packets to. This method functions perfectly fine, however, it can
>> greatly impact performance of the monitored port.
>>
>> It has been reported to reduce
On 11/5/21 19:24, Terry Wilson wrote:
> ovsdb-tool join-cluster requires a remote addr, so the existing
> code that tried to join a cluster without it when there was an
> existing $DB_FILE would fail.
>
> Instead, if we are trying to specifically join a cluster and there
> is an existing
On 11/5/21 15:13, Terry Wilson wrote:
> Many python implementations pre-allocate space for multiple
> objects in empty dicts and lists. Using a custom dict-like object
> that only generates these objects when they are accessed can save
> memory.
>
> On a fairly pathological case where the DB has
On 11/19/21 14:25, Eelco Chaudron wrote:
>
>
> On 5 Nov 2021, at 6:44, lic121 wrote:
>
>> remove bad quotes
>>
>> Signed-off-by: lic121
>
> Thanks for the patch, looks good to me.
>
> Acked-by: Eelco Chaudron
Applied. Thanks!
Best regards, Ilya Maximets.
On 11/29/21 16:50, Salvatore Daniele wrote:
> On Mon, Nov 29, 2021 at 8:01 AM Ilya Maximets wrote:
>>
>> On 11/4/21 20:54, Salvatore Daniele wrote:
>>> Hi Ilya,
>>>
>>> Per your feedback on v4, I accounted for igmp_type/code, both in the
>>> work-around in 1/2, and removing it from match.c
On 11/22/21 17:06, Gaëtan Rivet wrote:
> On Thu, Nov 18, 2021, at 16:45, Mike Pattrick wrote:
>> Recently there has been a lot of press about the "trojan source" attack,
>> where Unicode characters are used to obfuscate the true functionality of
>> code. This attack didn't effect OVS, but adding
On 11/19/21 15:15, Aaron Conole wrote:
> Mike Pattrick writes:
>
>> As part of some previous checkpatch work, we discovered that checkpatch
>> isn't always reporting correct line numbers. As it turns out, Python's
>> splitlines function considers several characters to be new lines which
>>
On 11/7/21 13:12, Roi Dayan via dev wrote:
> From: Paul Blakey
>
> Fragmented packets with offset=0 are defragmented by tc act_ct, and
> only when assembled pass to next action, in ovs offload case,
> a goto action. Since stats are overwritten on each action dump,
> only the stats for last
On 11/7/21 11:56, Eli Britstein via dev wrote:
> Following dpdk commits [1]-[3], it is now possible to compile with
> --enable-Werror. Change the default to on, with an option to disable
> using --disable-Werror.
>
> Notes:
> 1. To compile against 21.11-rc1, need to apply [4] and [5] patches.
>
On Tue, Nov 23, 2021 at 10:02 AM Dumitru Ceara wrote:
>
> OVS commit 066741d9c5ca ("ovsdb-idl: Add memory report function.")
> enhanced the IDL to track memory usage. Use it in ovn-controller for
> the Southbound DB and local OVS DB IDLs.
>
> Signed-off-by: Dumitru Ceara
> ---
>
On Fri, Nov 26, 2021 at 5:41 PM James Troup wrote:
>
> l3redirect_port was removed in 15348b7b8; remove references to it from
> comments.
>
> There is a comment in build_gateway_redirect_flows_for_lrouter() that
> supersedes the one removed below.
>
> Signed-off-by: James Troup
Thanks for the
On Fri, Nov 19, 2021 at 10:56 AM Dumitru Ceara wrote:
>
> Patches 1/3 and 2/3 fix some issues with the way the incremental
> processing engine is currently used for northd.
>
> Patch 3/3 enables the "recompute" command for northd too.
>
> Dumitru Ceara (3):
> inc-proc-northd: Don't reset
On 11/26/21 13:13, Kevin Traynor wrote:
> On 03/11/2021 16:57, Suneetha Kalahasthi wrote:
>> FAQ is updated to reflect the latest DPDK for OVS branch 2.13 and 2.14
>>
>> Signed-off-by: Suneetha Kalahasthi
>> ---
>> Documentation/faq/releases.rst | 4 ++--
>> 1 file changed, 2 insertions(+), 2
On Sat, Nov 27, 2021 at 12:12:51AM +0100, Paolo Valerio wrote:
> with the command is now possible to change the aging time of the
> cache entries.
>
> For the existing entries the aging time is updated only if the
> current expiration is greater than the new one. In any case, the next
> refresh
Looks good to me!
Acked-by: Mike Pattrick
On Sat, 2021-11-27 at 03:39 +, Peng He wrote:
> From: Peng He
>
> ipf_postprocess will emit packets into the datapath pipeline ignoring
> the conntrack context, this might casuse weird issues when a packet
> batch has less space to hold all the
This reverts commit c645550bb249 ("odp-util: Always report
ODP_FIT_TOO_LITTLE for IGMP.")
Always forcing a slow path action can result in some over-broad
flows which swallow all traffic and force them to userspace, as reported
in the thread at
On Thu, Nov 18, 2021 at 5:55 AM Mohammad Heib wrote:
>
> Improve packet-in debuggability within pinctrl module
> by printing basic details about each received packet-in
> message, those messages will be printed to the logs only
> when DBG log level is enabled.
>
> Also, add two coverage counters
Greetings,
Registration is now open for the Open vSwitch and OVN annual conference,
taking place this year as a virtual event on December 7 and 8, hosted on
the bluejeans platform. To register, please visit:
https://www.eventbrite.com/e/ovsovn-21-conference-registration-217640256867
There are
On 11/1/21 05:49, Yunjian Wang wrote:
> It is useful to also log when tnl_port_build_header() failed.
>
> Signed-off-by: Yunjian Wang
> ---
> ofproto/ofproto-dpif-xlate.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
>
On Tue, Oct 12, 2021 at 8:18 AM wrote:
>
> From: Mohammad Heib
>
> When the ovn controller receives an ip packet that targets a lport that has
> ACL
> rule to reject ip packets, the controller will reply with TCP_RST or icmp4/6
> unreachable packet
> to notify the sender that the destination
On Thu, Nov 18, 2021 at 9:17 AM Dumitru Ceara wrote:
>
> On 11/9/21 10:08 PM, Numan Siddique wrote:
> +Numan
>
> In general, I'm not completely sure about why sodepends.py needs to be
> part of OVN and why we can't use the OVS version?
> >>> OVS does not have this problem as
On Fri, Nov 26, 2021 at 7:52 AM Dumitru Ceara wrote:
>
> On 11/8/21 10:38, Frode Nordahl wrote:
> > The first test case of adding localport without a
> > requested-chassis option is invalid as this would make the
> > chassis do claim/release thrashing leading to a unpredictable
> > test results.
Thanks!
regards,
Vladislav Odintsov
> On 29 Nov 2021, at 21:18, Numan Siddique wrote:
>
> On Tue, Nov 23, 2021 at 12:08 PM Vladislav Odintsov
> wrote:
>>
>> In ovn-ic enabled setups, which also use VXLAN encapsulation
>> (for instance, for VTEP-switches) interconnected transit switches
>>
On Tue, Nov 23, 2021 at 12:08 PM Vladislav Odintsov wrote:
>
> In ovn-ic enabled setups, which also use VXLAN encapsulation
> (for instance, for VTEP-switches) interconnected transit switches
> (datapaths) were not configured after the patch from "Fixes" tag.
> This was due to the fact that
Without this patch "ovs-appctl dpctl/ct-bkts" produces the following
output:
Total Buckets: 1
Current Connections: 10246
+---+-+
| Buckets | Connections per Buckets |
+---+-+
0..
The purpose of this commit is to split the current way of storing the
conn nodes. Before this patch the nodes were stored into a single cmap
using ct->lock to avoid concurrent write access.
With this commit a single connection can be stored into one or two (at
most) CONNTRACK_BUCKETS available
From: Peng He
Currently, when doing NAT, the userspace conntrack will use an extra
conn for the two directions in a flow. However, each conn has actually
the two keys for both orig and rev directions. This patch introduces a
key_node[CT_DIR_MAX] member in the conn which consists of key and a
From: Gaetan Rivet
Multiple lookups are done to stored timeout policies, each time blocking
the global 'ct_lock'. This is usually not necessary and it should be
acceptable to get policy updates slightly delayed (by one RCU sync
at most). Using a CMAP reduces multiple lock taking and releasing in
From: Gaetan Rivet
Change the data structure from hmap to cmap for zone limits.
As they are shared amongst multiple conntrack users, multiple
readers want to check the current zone limit state before progressing in
their processing. Using a CMAP allows doing lookups without taking the
global
This series aims to share the work done so far, and to start a
discussion of a slightly different approach in terms of the way we
handle the connections.
It's not considered ready as further work and extensive tests are
needed.
There are two main logical changes that the series tries to
On 10/21/21 11:18, Nir Anteby via dev wrote:
> V1:
> Add hardware offload support for GRE flows (experimental).
> Available only if DPDK experimantal APIs enabled during the build.
>
> V2:
> Rebase version
>
> v3:
> Fixed rebase compilation
>
> v4:
> *Rebase version
> *Fixed bug found by
On 10/15/21 17:04, David Marchand wrote:
> When troubleshooting multiqueue setups, having per queue statistics helps
> checking packets repartition in rx and tx queues.
>
> Per queue statistics are exported by most DPDK drivers (with capability
> RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS). But since OVS
On Mon, Nov 29, 2021 at 8:01 AM Ilya Maximets wrote:
>
> On 11/4/21 20:54, Salvatore Daniele wrote:
> > Hi Ilya,
> >
> > Per your feedback on v4, I accounted for igmp_type/code, both in the
> > work-around in 1/2, and removing it from match.c entirely in 2/2.
> >
> > I was not sure the best way
On 11/22/21 01:09, Ilya Maximets wrote:
> With the next commit reference counting of json objects will take
> significant part of the CPU time for ovsdb-server. Inlining them
> to reduce the cost of a function call.
>
> Signed-off-by: Ilya Maximets
> ---
Looks good to me, thanks!
Acked-by:
On 11/22/21 01:09, Ilya Maximets wrote:
> ovsdb_atom_string and json_string are basically the same data structure
> and ovsdb-server frequently needs to convert one to another. We can
> avoid that by using json_string from the beginning for all ovsdb
> strings. So, the conversion turns into
Maxime Coquelin writes:
> Instead of waiting 10 seconds for testpmd to start, this
> patch makes use of OVS_WAIT_UNTIL() macro to wait for
> the Virtio device readiness notification in ovs-vswitchd
> logs.
>
> Signed-off-by: Maxime Coquelin
> ---
Acked-by: Aaron Conole
On 11/4/21 20:54, Salvatore Daniele wrote:
> Hi Ilya,
>
> Per your feedback on v4, I accounted for igmp_type/code, both in the
> work-around in 1/2, and removing it from match.c entirely in 2/2.
>
> I was not sure the best way to proceed regarding the issue I had
> mentioned on v4 where
On 11/22/21 00:46, Ilya Maximets wrote:
> ssl_send() clones the data before sending, but if SSL_write() succeeds
> at the first attempt, this is only a waste of CPU cycles.
>
> Trying to send the original buffer instead and only copying remaining
> data if it's not possible to send it all right
On Thu, Nov 25, 2021 at 6:00 PM Flavio Leitner wrote:
>
> On Thu, Nov 25, 2021 at 09:22:20AM +0100, Frode Nordahl wrote:
> > On Wed, Nov 24, 2021 at 9:31 PM Flavio Leitner wrote:
> > >
> > > On Mon, Nov 15, 2021 at 10:40:47AM +0100, Frode Nordahl wrote:
> > > > On Mon, Sep 13, 2021 at 4:23 AM
On 11/26/21 23:18, Numan Siddique wrote:
> On Fri, Nov 26, 2021 at 11:16 AM Dumitru Ceara wrote:
>>
>> There are various costs (e.g., not being able to perform hardware
>> offload in some cases) when using check_pkt_larger() so the CMS
>> can now limit the impact by bypassing the packet length
42 matches
Mail list logo