On Tue, Feb 2, 2021 at 2:05 AM Mark Michelson wrote:
>
> Signed-off-by: Mark Michelson
Acked-by: Numan Siddique
Numan
> ---
> Documentation/internals/release-process.rst | 20 +---
> 1 file changed, 13 insertions(+), 7 deletions(-)
>
> diff --git
On Wed, Feb 10, 2021 at 1:37 AM Dumitru Ceara wrote:
>
> Scale tests have identified the lflow cache to be one of the main memory
> consumers in ovn-controller. This series refactors the lflow cache code
> and adds configuration knobs to limit the size (in lines and/or memory)
> of the cache.
>
This patch adds GSO support for IPv4 TCP, when userspace-tso is enabled.
Tested using veth sending a TSO packet to OVS, segments to smaller TCP
segment, and forward to netdev-afxdp port at another namespace.
Future work includes:
1. GSO for UDP, and IPv6 TCP/UDP GSO.
2. Tunnel GSO: VxLan GSO,
Hi Numan, only one finding down below.
I know that you've added some end-to-end tests for MAC learning in later
patches, but I think adding some basic unit tests for mac-learn.c is
also useful.
On 2/5/21 1:59 AM, num...@ovn.org wrote:
From: Numan Siddique
This patch moves the 'struct
Resending as my previous try didn't make it on the mailing list.
On 2/9/21 1:06 PM, Dumitru Ceara wrote:
On 2/9/21 12:26 PM, gmingchen(陈供明) wrote:
Hi Dumitru,
Thanks for review.
On 2021/2/8, 6:19 PM, "Dumitru Ceara" wrote:
On 2/7/21 1:25 PM, gmingchen(陈供明) wrote:
> From: Gongming
On 2/8/21 11:09 PM, Mark Michelson wrote:
Hi Dumitru,
Thanks for the v2. I had one small nit on patch 6. Aside from that (and
Numan's comments), looks good.
Acked-by: Mark Michelson
Hi Mark, Numan,
Thanks for the reviews!
I sent a v3 with all the suggested changes.
Regards,
Dumitru
Add a new OVS external-id, "ovn-limit-lflow-cache", through which users
can specify the maximum size of the ovn-controller logical flow cache.
To maintain backwards compatibility the default behavior is to not
enforce any limit on the size of the cache.
When the cache becomes full, the rule is
Add a new OVS external-id, "ovn-memlimit-lflow-cache-kb", through which
users can specify the maximum amount of memory (in KB) ovn-controller
can use for caching logical flows.
To maintain backwards compatibility the default behavior is to not
enforce any memory limit on the size of the cache.
If possible, automatically reclaim heap memory when the lflow cache is
flushed. This can be an expensive operation but cache flushing is not
a very common operation.
This change is inspired by Ilya Maximets' OVS commit:
f38f98a2c0dd ("ovsdb-server: Reclaim heap memory after compaction.")
These are available in ovn-controller logs or when explicitly requested
by users with "ovn-appctl -t ovn-controller coverage/show".
Acked-by: Mark Michelson
Acked-by: Numan Siddique
Signed-off-by: Dumitru Ceara
---
controller/lflow-cache.c | 24
1 file changed, 24
Caching the conjunction id offset for flows that refer to address sets
and/or port groups but do not currently generate conjunctive matches,
e.g., because the port group has only one locally bound port, is wrong.
If ports are later added to the port group and/or addresses are added to
the address
Acked-by: Mark Michelson
Acked-by: Numan Siddique
Signed-off-by: Dumitru Ceara
---
v3:
- change calls to lflow-cache API to use new conventions.
- update UT outputs to match new lflow-cache stats outputs.
---
controller/test-lflow-cache.c | 223 +++
Acked-by: Mark Michelson
Signed-off-by: Dumitru Ceara
---
v3:
- change lflow_cache_get_stats() to populate an output string.
- document lflow-cache/show-stats command.
---
controller/lflow-cache.c| 27 +++
controller/lflow-cache.h|2 ++
This abstracts the implementation details of the ovn-controller logical
flow cache and also has refactors how the cache is being used allowing
further commits to expand its functionality.
Acked-by: Mark Michelson
Acked-by: Numan Siddique
Signed-off-by: Dumitru Ceara
---
v3:
- pass 'struct uuid
Scale tests have identified the lflow cache to be one of the main memory
consumers in ovn-controller. This series refactors the lflow cache code
and adds configuration knobs to limit the size (in lines and/or memory)
of the cache.
Patches 1 and 6 fix issues with the already existing lflow cache
It was (and still is) the responsibility of the caller of
convert_match_to_expr() to explicitly free any 'prereqs' expression that
was passed as argument if the expression parsing of the 'lflow' match failed.
However, convert_match_to_expr() now updates the value of '*prereqs' setting
it to NULL
If the I-P engine aborts in the middle of a run, the 'flow_output' node
change handlers or run callback might not be called.
As a side effect this may cause that Logical_Flow IDL tracked changes
are not processed during the iteration. As a consequence, if a
Logical_Flow was removed from the
The ovs-kmod-manage.sh script will run weak-updates even on newer
release kernels installing a non-compatible or un-runnable kernel
module.
Update the script to never install weak-updates onto kernels with
newer major release versions.
VMware-BZ: #2717283
Signed-off-by: Greg Rose
---
On 2/4/21 2:01 PM, Dumitru Ceara wrote:
On 1/28/21 9:35 PM, Mark Michelson wrote:
Signed-off-by: Mark Michelson
---
This patch is based on the "Include OVS as a git submodule." patch,
which, at this time has not been merged into OVN master yet.
---
.github/workflows/test.yml | 37
On 1/27/21 1:52 PM, Numan Siddique wrote:
On Tue, Jan 26, 2021 at 2:02 AM Dumitru Ceara wrote:
On 1/25/21 7:39 PM, Mark Michelson wrote:
OVN developers have had isssues with the current method by which OVS
source code is used by OVN.
* There is no way to record the minimum commit/version of
From: Numan Siddique
When a Gateway router is configured with a load balancer
and it is also configured with options:lb_force_snat_ip=,
OVN after load balancing the destination IP to one of the
backend also does a NAT on the source ip with the
lb_force_snat_ip if the packet is destined to a load
On Tue, Feb 9, 2021 at 8:41 PM Eli Britstein wrote:
>
>
> On 2/8/2021 6:21 PM, Sriharsha Basavapatna wrote:
> > On Mon, Feb 8, 2021 at 7:33 PM Eli Britstein wrote:
> >>
> >> On 2/8/2021 3:11 PM, Sriharsha Basavapatna wrote:
> >>> On Sun, Feb 7, 2021 at 4:58 PM Eli Britstein wrote:
> On
On 2/8/2021 6:21 PM, Sriharsha Basavapatna wrote:
On Mon, Feb 8, 2021 at 7:33 PM Eli Britstein wrote:
On 2/8/2021 3:11 PM, Sriharsha Basavapatna wrote:
On Sun, Feb 7, 2021 at 4:58 PM Eli Britstein wrote:
On 2/5/2021 8:26 PM, Sriharsha Basavapatna wrote:
On Fri, Feb 5, 2021 at 4:55 PM
Hi Dumitru,
Thanks for review.
On 2021/2/8, 6:19 PM, "Dumitru Ceara" wrote:
On 2/7/21 1:25 PM, gmingchen(陈供明) wrote:
> From: Gongming Chen
Hi Gongming,
First of all, thanks for the contribution!
This is not a full review, just some comments for now.
It seems that
Add a new OVS external-id, "ovn-limit-lflow-cache", through which users
can specify the maximum size of the ovn-controller logical flow cache.
To maintain backwards compatibility the default behavior is to not
enforce any limit on the size of the cache.
When the cache becomes full, the rule is
These are available in ovn-controller logs or when explicitly requested
by users with "ovn-appctl -t ovn-controller coverage/show".
Acked-by: Mark Michelson
Acked-by: Numan Siddique
Signed-off-by: Dumitru Ceara
---
controller/lflow-cache.c | 24
1 file changed, 24
Acked-by: Mark Michelson
Acked-by: Numan Siddique
Signed-off-by: Dumitru Ceara
---
v3:
- change calls to lflow-cache API to use new conventions.
- update UT outputs to match new lflow-cache stats outputs.
---
controller/test-lflow-cache.c | 223 +++
Acked-by: Mark Michelson
Signed-off-by: Dumitru Ceara
---
v3:
- change lflow_cache_get_stats() to populate an output string.
- document lflow-cache/show-stats command.
---
controller/lflow-cache.c| 27 +++
controller/lflow-cache.h|2 ++
This abstracts the implementation details of the ovn-controller logical
flow cache and also has refactors how the cache is being used allowing
further commits to expand its functionality.
Acked-by: Mark Michelson
Acked-by: Numan Siddique
Signed-off-by: Dumitru Ceara
---
v3:
- pass 'struct uuid
It was (and still is) the responsibility of the caller of
convert_match_to_expr() to explicitly free any 'prereqs' expression that
was passed as argument if the expression parsing of the 'lflow' match failed.
However, convert_match_to_expr() now updates the value of '*prereqs' setting
it to NULL
On 2/8/21 8:54 PM, William Tu wrote:
> On Mon, Feb 8, 2021 at 6:36 AM Ilya Maximets wrote:
>>
>> On 2/8/21 1:56 AM, Yi Yang (杨燚)-云服务集团 wrote:
>>> Yes, GSO is ok, but GRO may be have that issue, I didn't see that issue in
>>> my openstack environment, so maybe it will be great if we can have a
On 2021/02/09 4:13, William Tu wrote:
On Mon, Feb 8, 2021 at 8:57 AM Ilya Maximets wrote:
On 2/6/21 5:15 PM, William Tu wrote:
On Mon, Feb 1, 2021 at 5:48 PM Yi Yang (杨燚)-云服务集团 wrote:
Thanks Ilya, net_tap PMD is handling tap device on host side, so it can
leverage vnet header to do
On 2/8/21 11:35 PM, William Tu wrote:
> On Mon, Feb 8, 2021 at 4:58 AM Ilya Maximets wrote:
>>
>> On 2/7/21 5:05 PM, Toshiaki Makita wrote:
>>> On 2021/02/07 2:00, William Tu wrote:
On Fri, Feb 5, 2021 at 1:08 PM Gregory Rose wrote:
> On 2/4/2021 7:08 PM, William Tu wrote:
>> On
On 2021/02/05 12:03, William Tu wrote:
On Thu, Feb 4, 2021 at 5:15 PM William Tu wrote:
Hi Toshiaki,
Thanks for the patch. I have some questions inline.
On Thu, Jul 30, 2020 at 7:55 PM Toshiaki Makita
wrote:
This provider offloads classifier to software XDP.
It works only when a custom
On 2021/02/05 2:36, William Tu wrote:
Hi Toshiaki,
Thanks for the patch. I've been testing it for a couple days.
I liked it a lot! The compile and build process all work without any issues.
Hi, thank you for reviewing!
Sorry for taking time to reply. It took time to remember every detail of
On Tue, 9 Feb 2021 at 00:58, Gregory Rose wrote:
> On 1/14/2021 1:09 AM, Michal Kazior wrote:
> > From: Michal Kazior
[...]
>
> Hi Michal,
>
> The patch looks fine to me. Applies cleanly to master
> and did not cause any regressions in 'make check'.
>
> I'm curious since I don't ever use OVS
36 matches
Mail list logo