On Fri, Sep 16, 2022 at 5:30 AM Dumitru Ceara wrote:
> I'd like to post an OVS patch to add this along with some more helpers
> (see below) but I have a couple of questions:
>
> 1. What's the correct/desired way of maintaining authorship in this
> case? I can keep all original authors and add my
On Fri, Jul 15, 2022 at 6:10 AM Ilya Maximets wrote:
> Hey! It's been about 9 years since the last major version change.
> And there were lots of large and important changes in OVS in that
> long time frame. Nevertheless, removal of the kernel datapath
> sources from the OVS's git tree seems
On Fri, Jul 8, 2022 at 12:00 PM Ilya Maximets wrote:
> On 7/8/22 17:37, Ben Pfaff wrote:
> > On Fri, Jul 8, 2022 at 1:53 AM David Marchand <mailto:david.march...@redhat.com>> wrote:
> >
> > On Thu, Jul 7, 2022 at 11:21 PM Gregory Rose <ma
On Fri, Jul 8, 2022 at 1:53 AM David Marchand
wrote:
> On Thu, Jul 7, 2022 at 11:21 PM Gregory Rose wrote:
> > > xenserver/openvswitch-xen.spec.in
> >
> > This is a bit of an issue. Does anyone even use xenserver anymore?
> > All of the documentation and build instructions are really old and
>
On Thu, May 12, 2022 at 8:55 AM Aaron Conole wrote:
>
> Ben Pfaff writes:
>
> > Hi Aaron! This will move the following fields from the L4 segment to
> > the L3 segment:
> >
> > struct in6_addr nd_target; /* IPv6 neighbor discovery (ND) target. */
> >
Hi Aaron! This will move the following fields from the L4 segment to
the L3 segment:
struct in6_addr nd_target; /* IPv6 neighbor discovery (ND) target. */
struct eth_addr arp_sha;/* ARP/ND source hardware address. */
struct eth_addr arp_tha;/* ARP/ND target hardware address.
On Tue, May 3, 2022 at 3:15 PM Han Zhou wrote:
> I believe this has been applied already last year.
Oh, I see, we also transitioned to have the main branch be "main" but
I was looking
in the wrong place.
No need to do anything.
Thanks!
___
dev
I am no longer actively working on OVN, so it's appropriate to move myself
to emeritus status.
Signed-off-by: Ben Pfaff
---
MAINTAINERS.rst | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/MAINTAINERS.rst b/MAINTAINERS.rst
index 0d19bd622..fff2cd977 100644
On Thu, Mar 31, 2022 at 02:03:41PM -0400, Mark Michelson wrote:
> @Ben
> Can you please change the channel topic message of #openvswitch so that the
> new meeting time is reflected? Thanks!
I've done this now.
Does anyone know how I would give someone else permission to do this
kind of thing? I
On Wed, Jan 26, 2022 at 06:39:52PM -0800, William Tu wrote:
> On Wed, Jan 26, 2022 at 8:56 AM Ben Pfaff wrote:
> >
> > Until now, the monitor process held its log file open while it waited for
> > the child to exit, and then it reopened it after the child exited. Holding
, then
reopening it afterward.
Signed-off-by: Ben Pfaff
Reported-by: Antonin Bas
VMware-BZ: #2743409
---
v1->v2: Fix memory leak in monitor_daemon() reported by William.
include/openvswitch/vlog.h | 2 +
lib/daemon-unix.c | 5 +-
lib/vlog.c |
, then
reopening it afterward.
Signed-off-by: Ben Pfaff
Reported-by: Antonin Bas
VMware-BZ: #2743409
---
include/openvswitch/vlog.h | 2 +
lib/daemon-unix.c | 4 +-
lib/vlog.c | 93 ++
3 files changed, 70 insertions(+), 29 deletions
On Fri, Dec 24, 2021 at 10:32:23AM -0500, Numan Siddique wrote:
> On Thu, Dec 23, 2021 at 6:51 PM Ben Pfaff wrote:
> >
> > This came to me. I don't know whether this is something we/I should
> > approve.
>
> This request came from Andrew (CCed) and I approved
This came to me. I don't know whether this is something we/I should
approve.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
will setup some time next week to discuss alignment and share our
>proposed MSVC plan.
>
>
>
>
>
> *From:* Ben Pfaff
> *Sent:* Thursday, November 4, 2021 11:15 AM
> *To:* Omar Cardona
> *Cc:* William Tu ; thomas ; <
> d...@openvswitch.org> ; ovs-dev <
On Thu, Nov 4, 2021 at 8:37 AM Omar Cardona wrote:
> > > f. With clang support on Windows, can we remove MSVC?
>
> Please no. We are actively working with the MSVC team to extend support
> for DPDKs dependencies
>
Could you get them to implement C11 atomics? The MSVC support for atomic
On Thu, Sep 30, 2021 at 12:23:57AM -0700, Han Zhou wrote:
> Hi Ben,
>
> I understand that we have difficulties for northd-ddlog progress, but I
> still want to try it before it goes away. I tested with the latest version,
> and it is super fast for most of the operations. With a large NB & SB
>
On Wed, Oct 6, 2021 at 10:01 AM William Tu wrote:
> The patch adds OVS-DPDK supports on Windows.
Fantastic! I'm happy to see this work in progress. OVS(-DPDK) everywhere!
___
dev mailing list
d...@openvswitch.org
On Tue, Oct 05, 2021 at 08:06:45PM -0700, William Tu wrote:
> On Tue, Oct 5, 2021 at 8:05 PM William Tu wrote:
> >
> > On Tue, Oct 5, 2021 at 11:33 AM Numan Siddique wrote:
> > >
> > > On Tue, Oct 5, 2021 at 1:46 PM Han Zhou wrote:
> > > >
> >
On Wed, Oct 06, 2021 at 11:20:48AM +0200, Simon Horman wrote:
> On Tue, Oct 05, 2021 at 09:32:41AM -0700, Ben Pfaff wrote:
> > It has been a long time since I've been able to devote significant time
> > to OVS work. I don't expect that to change. I think that the responsible
I have not been actively maintaining OVN for some time now. I don't expect
that to change. I think that the responsible thing to do is to
acknowledge this properly by transitioning to emeritus status.
Signed-off-by: Ben Pfaff
---
MAINTAINERS.rst | 4 ++--
1 file changed, 2 insertions(+), 2
It has been a long time since I've been able to devote significant time
to OVS work. I don't expect that to change. I think that the responsible
thing to do is to transition myself to emeritus status, with this patch.
Signed-off-by: Ben Pfaff
---
MAINTAINERS.rst | 4 ++--
1 file changed, 2
tch, this might be the right solution. All IGMP packets
> > independent of their type need to be processed by a SLOW_ACTION upcall.
> >
> >
> > I've copied in Flavio, who worked on the IGMP snooping part before, as he
> > might still remember what I argue above is true :
Thanks. I rebased and applied this series to master.
On Mon, Sep 13, 2021 at 02:13:28PM -0400, Mark Michelson wrote:
> Hi Ben, for the series:
>
> Acked-by: Mark Michelson
>
> I had a question on patch 1, but it's minor and otherwise shouldn't block an
> ACK.
>
> On 9
SwitchPortARPForwards already has all the necessary data in it and joining
with SwitchPort just wastes time.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 12
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/northd/ovn_northd.dl b/northd/ovn_northd.dl
index
By interning earlier, we do less copying across the joins with
Router and Switch, because DDlog always copies by value and
copying an interned string is cheap.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 57 +++-
1 file changed, 30 insertions
It's expensive for a long list. All it buys us is sorting the list in
alphabetical order (rather than in order of Intern hash value), which isn't
that valuable.
This also updates a test that depended on the sort order.
Suggested-by: Leonid Ryzhyk
Signed-off-by: Ben Pfaff
---
northd
as well just reference it.
Only slight improvement to performance.
Signed-off-by: Ben Pfaff
---
northd/lswitch.dl| 32 ++--
northd/ovn_northd.dl | 29 +
2 files changed, 23 insertions(+), 38 deletions(-)
diff --git a/northd/lswitch.dl b
Signed-off-by: Ben Pfaff
---
northd/lrouter.dl | 9 -
1 file changed, 9 deletions(-)
diff --git a/northd/lrouter.dl b/northd/lrouter.dl
index c0ec6be47..582e07654 100644
--- a/northd/lrouter.dl
+++ b/northd/lrouter.dl
@@ -488,7 +488,6 @@ typedef Router = Router {
is_gateway
syntax to traditional Datalog-style syntax. This is not
completely needed, but the Datalog-style syntax is more versatile
(it supports FlatMap and aggregation), so I tend to make that sort
of change as I refactor things.
Signed-off-by: Ben Pfaff
---
northd/lrouter.dl| 11 --
northd
The LBVIPs already contain parsed versions of the load balancer
keys, but the code was parsing it redundantly.
Slight performance improvement but much cleaner.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 37 -
1 file changed, 12 insertions(+), 25
This makes the get_router_load_balancer_ips() function much faster.
This function is a hot path for the use of load balancers, so it
improves performance overall when they are in use.
Signed-off-by: Ben Pfaff
---
northd/lrouter.dl| 68 +--
northd
DDlog can index equality joins within a expression like this, but not
conditional expression that follow such an expression. This doesn't
actually matter much in this case because ordinarily the expression
will be true (most router ports are enabled).
Signed-off-by: Ben Pfaff
---
northd
continuing to work on performance.
These patches are also available in:
https://github.com/blp/ovs-reviews/tree/ovn-memory-8
v1->v2:
- Add two new patches.
- Rebase and rerun benchmarks (revisions reported above).
Ben Pfaff (9):
ovn-northd-ddlog: Make joins for ARP/ND fl
On Tue, Aug 24, 2021 at 11:07:22PM +0200, Ilya Maximets wrote:
> +count = 0;
> +start = string;
> while ((c = *string++) != '\0') {
> -escape = chars_escaping[c];
> -while ((c2 = *escape++) != '\0') {
> -ds_put_char(ds, c2);
> +if (c >= ' ' && c !=
By interning earlier, we do less copying across the joins with
Router and Switch, because DDlog always copies by value and
copying an interned string is cheap.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 57 +++-
1 file changed, 30 insertions
as well just reference it.
Only slight improvement to performance.
Signed-off-by: Ben Pfaff
---
northd/lswitch.dl| 32 ++--
northd/ovn_northd.dl | 29 +
2 files changed, 23 insertions(+), 38 deletions(-)
diff --git a/northd/lswitch.dl b
Signed-off-by: Ben Pfaff
---
northd/lrouter.dl | 9 -
1 file changed, 9 deletions(-)
diff --git a/northd/lrouter.dl b/northd/lrouter.dl
index c0ec6be47..582e07654 100644
--- a/northd/lrouter.dl
+++ b/northd/lrouter.dl
@@ -488,7 +488,6 @@ typedef Router = Router {
is_gateway
The LBVIPs already contain parsed versions of the load balancer
keys, but the code was parsing it redundantly.
Slight performance improvement but much cleaner.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 37 -
1 file changed, 12 insertions(+), 25
syntax to traditional Datalog-style syntax. This is not
completely needed, but the Datalog-style syntax is more versatile
(it supports FlatMap and aggregation), so I tend to make that sort
of change as I refactor things.
Signed-off-by: Ben Pfaff
---
northd/lrouter.dl| 11 --
northd
DDlog can index equality joins within a expression like this, but not
conditional expression that follow such an expression. This doesn't
actually matter much in this case because ordinarily the expression
will be true (most router ports are enabled).
Signed-off-by: Ben Pfaff
---
northd
This makes the get_router_load_balancer_ips() function much faster.
This function is a hot path for the use of load balancers, so it
improves performance overall when they are in use.
Signed-off-by: Ben Pfaff
---
northd/lrouter.dl| 68 +--
northd
continuing to work on performance.
Ben Pfaff (7):
ovn-northd-ddlog: Make joins for ARP/ND flows slightly more efficient.
ovn-northd-ddlog: Derive load balancer IP addresses in new
LoadBalancer.
ovn-northd-ddlog: Reverse order of joins for connection tracking
flows.
ovn-northd-ddlog
On Fri, Aug 20, 2021 at 03:43:15PM +0200, Frode Nordahl wrote:
> Data retrieved from netlink and friends may include link layer
> address. Add type to nl_attr_type and min/max functions to allow
> use of nl_policy_parse with this type of data.
>
> While this will not be used by Open vSwitch
On Wed, Aug 18, 2021 at 08:42:47PM -0400, num...@ovn.org wrote:
> From: Numan Siddique
>
> This patch adds 2 new APIs in the ovsdb-idl client library
> - ovsdb_idl_has_table() and ovsdb_idl_has_column_in_table() to
> query if a table and a column is present in the IDL or not. This
> is
On Mon, Aug 16, 2021 at 03:52:49PM -0300, Flavio Leitner wrote:
> On Tue, Jul 20, 2021 at 11:41:37AM -0700, Ben Pfaff wrote:
> > The OVS review process has greatly slowed over the last few years. This
> > is partly because I haven't been able to spend as much time on review,
>
On Tue, Aug 17, 2021 at 11:47:50AM -0400, Numan Siddique wrote:
> On Thu, Aug 12, 2021 at 1:45 PM Ben Pfaff wrote:
> >
> > Apologies for repost; I forgot to tag this as "ovn" in the first
> > posting.
> >
> > For the benchmark at
> > https://mail.o
There's another perspective from the PostgreSQL community detailed at
LWN. Here's a "guest link" to the article, titled "PostgreSQL's
commitfest clog":
https://lwn.net/SubscriberLink/865754/6c0eb771f8436789/
The paragraph that starts out "There is a longstanding expectation
within the PostgreSQL
This has little effect on performance in the current benchmark, but it
seems like the cleanest approach to me.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 38 ++
1 file changed, 22 insertions(+), 16 deletions(-)
diff --git a/northd/ovn_northd.dl b
With the benchmark at
https://mail.openvswitch.org/pipermail/ovs-dev/2021-July/385333.html,
this reduces memory consumption by 300 MB and elapsed time by a few
seconds.
Signed-off-by: Ben Pfaff
---
northd/ipam.dl | 2 +-
northd/lswitch.dl| 36
-July/385333.html,
this reduces memory consumption by 1.3 GB.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 579 ++-
1 file changed, 292 insertions(+), 287 deletions(-)
diff --git a/northd/ovn_northd.dl b/northd/ovn_northd.dl
index 844add024
from 53 GB to 17 GB and elapsed time
from 14 minutes to 5 minutes.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/northd/ovn_northd.dl b/northd/ovn_northd.dl
index dd69126ff..38b97aa6b 100644
--- a/northd
With the benchmark at
https://mail.openvswitch.org/pipermail/ovs-dev/2021-July/385333.html,
this reduces memory consumption from 72 GB to 66 GB and elapsed time by
a few seconds.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 140 +--
1 file changed
s headroom for more improvement, because Leonid has some
cleverness that I haven't been able to successfully work in yet.
This series is also available here:
https://github.com/blp/ovs-reviews/tree/ovn-memory-4
Ben Pfaff (7):
ovn-northd-ddlog: Merge TaggedFlow and MeteredFlow into Flow.
ovn-northd-
On Tue, Aug 10, 2021 at 05:17:56PM -0700, Sergey Madaminov wrote:
> Checking if '__has_builtin' is defined and then defining OVS_PREFETCH to
> be '__builtin_prefetch' if it is available. To preserve backwards
> compatibility, the previous way to define OVS_PREFETCH macro is also
> there.
Thanks
On Thu, Aug 12, 2021 at 08:53:51AM -0700, Ben Pfaff wrote:
> For the benchmark at
> https://mail.openvswitch.org/pipermail/ovs-dev/2021-July/385333.html,
> this reduces memory consumption from 115 GB to 17 GB and elapsed time
> from 19 minutes to 5 minutes.
>
> I think there's
From: Ben Pfaff
This has little effect on performance in the current benchmark, but it
seems like the cleanest approach to me.
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 38 ++
1 file changed, 22 insertions(+), 16 deletions(-)
diff --git a/northd
From: Ben Pfaff
With the benchmark at
https://mail.openvswitch.org/pipermail/ovs-dev/2021-July/385333.html,
this reduces memory consumption by 300 MB and elapsed time by a few
seconds.
Signed-off-by: Ben Pfaff
---
northd/ipam.dl | 2 +-
northd/lswitch.dl| 36
From: Ben Pfaff
With the benchmark at
https://mail.openvswitch.org/pipermail/ovs-dev/2021-July/385333.html,
this reduces memory consumption from 72 GB to 66 GB and elapsed time by
a few seconds.
Signed-off-by: Ben Pfaff
Signed-off-by: Ben Pfaff
---
northd/ovn_northd.dl | 140
From: Ben Pfaff
It was possible for these rules to generate multiple Flow records that
differed only in their stage_hint. This caused a lot of duplication for
the load balancer benchmark.
With the benchmark at
https://mail.openvswitch.org/pipermail/ovs-dev/2021-July/385333.html,
this reduces
From: Ben Pfaff
The stage_hint only shows 32 bits of the uuid, so it's cheaper to omit
the rest for the internal representation. Also, this is just a hint, so
we might as well use zero to mean None and save the cost of the Option
wrapper.
With the benchmark at
https://mail.openvswitch.org
to successfully work in yet.
Ben Pfaff (7):
ovn-northd-ddlog: Merge TaggedFLow and MeteredFlow into Flow.
ovn-northd-ddlog: Use cheaper representation for stage_hint.
ovn-northd-ddlog: Intern Stage.
Intern all the matches and actions.
ovn-northd-ddlog: Get rid of duplicate flows caused
On Fri, Jul 23, 2021 at 08:26:51PM +0800, Tonghao Zhang wrote:
> On Wed, Jul 21, 2021 at 2:41 AM Ben Pfaff wrote:
> >
> > The OVS review process has greatly slowed over the last few years. This
> > is partly because I haven't been able to spend as much time on review,
On Thu, Jul 22, 2021 at 12:14:23PM -0700, Ben Pfaff wrote:
> On Wed, Jul 21, 2021 at 01:22:19PM +0200, Ilya Maximets wrote:
> > On 7/21/21 12:52 AM, Ben Pfaff wrote:
> > > On Tue, Jul 20, 2021 at 08:29:14PM +0200, Ilya Maximets wrote:
> > >> On 7
On Wed, Jul 21, 2021 at 01:22:19PM +0200, Ilya Maximets wrote:
> On 7/21/21 12:52 AM, Ben Pfaff wrote:
> > On Tue, Jul 20, 2021 at 08:29:14PM +0200, Ilya Maximets wrote:
> >> On 7/7/21 8:51 PM, Ben Pfaff wrote:
> >>> +/* From an OpenFlow point of view,
ninitialized in this
>function [-Werror=maybe-uninitialized]
> 284 | if (retval == n && dp_packet_size(>rxbuf) > 2) {
> |^
>
> This breaks the build with --enable-Werror. Initializing 'n' to
> avoid the warning.
>
>
On Wed, Jul 21, 2021 at 10:37:18AM +0200, Eelco Chaudron wrote:
>
>
> On 20 Jul 2021, at 20:41, Ben Pfaff wrote:
>
> > The OVS review process has greatly slowed over the last few years. This
> > is partly because I haven't been able to spend as much time on revi
On Tue, Jul 20, 2021 at 08:29:14PM +0200, Ilya Maximets wrote:
> On 7/7/21 8:51 PM, Ben Pfaff wrote:
> > +/* From an OpenFlow point of view, goto_table and write_metadata
> > are
> > + * instructions, not actions. This means that to use
The OVS review process has greatly slowed over the last few years. This
is partly because I haven't been able to spend as much time on review,
since I was once the most productive reviewer. Ilya has been able to
step up the amount of review he does, but that still isn't enough to
keep up with
On Tue, Jul 20, 2021 at 06:26:48PM +0100, Anton Ivanov wrote:
>
> On 19/07/2021 21:52, Ben Pfaff wrote:
> > On Fri, Jul 16, 2021 at 02:42:32PM +0100, anton.iva...@cambridgegreys.com
> > wrote:
> > > From: Anton Ivanov
> > >
> > > If we are not obt
On Tue, Jul 20, 2021 at 08:19:35AM +0100, Anton Ivanov wrote:
>
> On 19/07/2021 21:52, Ben Pfaff wrote:
> > On Fri, Jul 16, 2021 at 02:42:32PM +0100, anton.iva...@cambridgegreys.com
> > wrote:
> > > From: Anton Ivanov
> > >
> > > If we are not obt
On Wed, Jul 07, 2021 at 11:51:50AM -0700, Ben Pfaff wrote:
> Open vSwitch supports OpenFlow "instructions", which were introduced in
> OpenFlow 1.1 and act like restricted kinds of actions that can only
> appear in a particular order and particular circumstances. OVS di
On Fri, Jul 16, 2021 at 02:42:33PM +0100, anton.iva...@cambridgegreys.com wrote:
> From: Anton Ivanov
>
> time_poll() makes an excessive number of time_msec() calls
> which incur a performance penalty.
>
> 1. Avoid time_msec() call for timeout calculation when time_poll()
> is asked to skip
On Fri, Jul 16, 2021 at 02:42:32PM +0100, anton.iva...@cambridgegreys.com wrote:
> From: Anton Ivanov
>
> If we are not obtaining any useful information out of the poll(),
> such as is a fd busy or not, we do not need to do a poll() if
> an immediate_wake() has been requested.
>
> This cuts out
On Thu, Jul 15, 2021 at 06:10:40PM -0700, Han Zhou wrote:
> On Thu, Jul 15, 2021 at 4:57 PM Ben Pfaff wrote:
>
> > On Thu, Jul 15, 2021 at 04:34:51PM -0700, Han Zhou wrote:
> > > Fixes: 895e02ec0be6("ovn-sbctl.c Add logical flows count numbers"
On Sun, Jul 18, 2021 at 04:22:02AM +, Wilson Peng wrote:
> Hi, d...@openvswitch.org ,
>
> I would like to send out one patch for some issue found on datapah-windows in
> OVS repo,
>
> I have sent the patch via the email attached in the mail and the guided doc to
> Submit the patch is got
On Mon, Jul 19, 2021 at 01:27:16PM +0200, David Marchand wrote:
> This is a simple cleanup with a script of mine.
>
> Signed-off-by: David Marchand
> ---
> This script is a bit too silly/simplistic and requires manual review
> (too many false positives), but I don't think it is worth investing
On Fri, Jul 16, 2021 at 10:42:01PM +0200, Ilya Maximets wrote:
> Hi, everyone.
>
> branch-2.16 was just created. Please, test it and report issues!
>
> Official release date according to our release process should be on
> Monday, August 16th (1 month from now).
>
> Best regards, Ilya Maximets.
On Fri, Jul 16, 2021 at 09:43:17AM +0200, Dumitru Ceara wrote:
> On 7/16/21 3:05 AM, Han Zhou wrote:
> > On Thu, Jul 15, 2021 at 5:03 PM Ben Pfaff wrote:
> >
> >> On Mon, Jul 12, 2021 at 10:08:10AM +0200, Dumitru Ceara wrote:
> >>> Whenever a Load_Balan
On Mon, Jul 12, 2021 at 10:08:10AM +0200, Dumitru Ceara wrote:
> Whenever a Load_Balancer is updated, e.g., a VIP is added, the following
> sequence of events happens:
>
> 1. The Southbound Load_Balancer record is updated.
> 2. The Southbound Datapath_Binding records on which the Load_Balancer is
reproducible improvement.
> IMHO it should be even more noticeable at high scale.
>
> Best regards,
>
> A
>
> On 09/07/2021 19:45, Ben Pfaff wrote:
> > On Fri, Jul 09, 2021 at 06:19:06PM +0100, anton.iva...@cambridgegreys.com
> > wrote:
> > > From: Anton I
On Thu, Jul 15, 2021 at 04:34:51PM -0700, Han Zhou wrote:
> Fixes: 895e02ec0be6("ovn-sbctl.c Add logical flows count numbers")
> Signed-off-by: Han Zhou
> ---
> tests/ovn-sbctl.at | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/tests/ovn-sbctl.at
On Mon, Jul 12, 2021 at 04:42:27PM -0700, Ben Pfaff wrote:
> On Thu, Jul 08, 2021 at 08:59:24PM +0200, Dumitru Ceara wrote:
> > Hi Ben,
> >
> > As discussed earlier, during the OVN meeting, I've noticed a new
> > performance issue with ovn-northd-ddlog when running it
ry which load less than or equal to 'ideal_delta'.
+/* There is no entry with load less than or equal to 'ideal_delta'.
* Lets try closest one. The closest is the last in sorted list. */
struct bond_entry *closest;
Acked-by: Ben Pfaff
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On Tue, Jul 13, 2021 at 12:17:48PM -0400, Numan Siddique wrote:
> On Fri, Jul 2, 2021 at 5:36 PM Ben Pfaff wrote:
> >
> > These were part of the OVS tutorial, which isn't relevant for OVN.
> >
> > Signed-off-by: Ben Pfaff
>
> Acked-by: Numan Siddique
On Mon, Jul 12, 2021 at 06:15:29PM +0100, anton.iva...@cambridgegreys.com wrote:
> From: Anton Ivanov
>
> time_poll() makes an excessive number of time_msec() calls
> which incur a performance penalty.
>
> 1. Avoid time_msec() call for timeout calculation when time_poll()
> is asked to skip
On Mon, Jul 12, 2021 at 06:15:28PM +0100, anton.iva...@cambridgegreys.com wrote:
> From: Anton Ivanov
>
> If we are not obtaining any useful information out of the poll(),
> such as is a fd busy or not, we do not need to do a poll() if
> an immediate_wake() has been requested.
>
> This cuts out
On Thu, Jul 15, 2021 at 08:45:00PM +0200, Lorenzo Bianconi wrote:
> > On Thu, Jul 15, 2021 at 11:18:12AM -0700, Han Zhou wrote:
> > > On Wed, Jul 14, 2021 at 1:34 AM Lorenzo Bianconi <
> > > lorenzo.bianc...@redhat.com> wrote:
> > > > This should avoid some work by doing the cheapest check (the
xes: 1b1d2e6daa56 ("ovsdb: Introduce experimental support for clustered
databases.")
Signed-off-by: Dumitru Ceara
Signed-off-by: Ben Pfaff
---
ovsdb/ovsdb-server.c | 8 +++-
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/ovsdb/ovsdb-server.c b/ovsdb/ovsdb-server.c
inde
On Thu, Jul 15, 2021 at 04:28:12PM +0100, anton.iva...@cambridgegreys.com wrote:
> From: Anton Ivanov
>
> 1. Make latch behave as described and documented - clear all
> outstanding latch writes when invoking latch_poll().
> 2. Decrease the size of the latch buffer. Less stack usage,
> less cache
On Thu, Jul 15, 2021 at 11:18:12AM -0700, Han Zhou wrote:
> On Wed, Jul 14, 2021 at 1:34 AM Lorenzo Bianconi <
> lorenzo.bianc...@redhat.com> wrote:
> > This should avoid some work by doing the cheapest check (the one on
> > UseLogicalDatapathGroups) before any joins. DDlog is probably
> >
On Wed, Jul 14, 2021 at 06:23:15PM -0400, num...@ovn.org wrote:
> From: Numan Siddique
>
> Fixes: 895e02ec0be6("ovn-sbctl.c Add logical flows count numbers")
> Signed-off-by: Numan Siddique
Acked-by: Ben Pfaff
On Wed, Jul 14, 2021 at 09:12:10PM +0100, Anton Ivanov wrote:
> On 14/07/2021 19:33, Ben Pfaff wrote:
> > On Wed, Jul 14, 2021 at 05:36:36PM +0100, anton.iva...@cambridgegreys.com
> > wrote:
> > > From: Anton Ivanov
> > >
> > > There is no point to a
On Wed, Jul 14, 2021 at 05:36:36PM +0100, anton.iva...@cambridgegreys.com wrote:
> From: Anton Ivanov
>
> There is no point to add 512 bytes on the stack
> every time latch is polled. Alignment, cache line thrashing,
> etc - you name it.
Do you have evidence this is a real problem?
> The
On Tue, Jul 13, 2021 at 12:00:13PM -0700, Ben Pfaff wrote:
> On Sat, Jul 10, 2021 at 02:56:53PM +0200, Lorenzo Bianconi wrote:
> > ack, no worries. I folded the patch here:
> > https://github.com/LorenzoBianconi/ovn/tree/CoPP-v7
> >
> > however CoPP tests
On Sat, Jul 10, 2021 at 02:56:53PM +0200, Lorenzo Bianconi wrote:
> ack, no worries. I folded the patch here:
> https://github.com/LorenzoBianconi/ovn/tree/CoPP-v7
>
> however CoPP tests available in system-ovn.at and for
> ovn-northd/ovn-controller
> are failing with DDlog. Can you please
On Thu, Jul 08, 2021 at 08:59:24PM +0200, Dumitru Ceara wrote:
> Hi Ben,
>
> As discussed earlier, during the OVN meeting, I've noticed a new
> performance issue with ovn-northd-ddlog when running it against a
> database from one of our more recent scale tests:
>
>
like:
>
> Time spent on processing nb_cfg 39:
> ovn-northd delay before processing: -1623916000627ms
> ovn-northd completion: 14408ms
> ovn-controller(s) completion: 36204ms
>
> Signed-of
On Fri, Jul 09, 2021 at 06:19:06PM +0100, anton.iva...@cambridgegreys.com wrote:
> From: Anton Ivanov
>
> If we are not obtaining any useful information out of the poll(),
> such as is a fd busy or not, we do not need to do a poll() if
> an immediate_wake() has been requested.
>
> This cuts out
On Wed, Jul 07, 2021 at 05:17:17PM -0700, Han Zhou wrote:
> On Wed, Jul 7, 2021 at 3:39 PM Ben Pfaff wrote:
> >
> > Hi, I've been talking to Shay Vargaftik (CC'd), also a researcher at
> > VMware, about some work he's done on optimizing load balancers. What
> > he's
1 - 100 of 24009 matches
Mail list logo