On 01.06.2018 16:18, Eelco Chaudron wrote:
> When ping-pong'in a live VM migration between two machines running
> OVS-DPDK every now and then the ping misses would increase
> dramatically. For example:
>
> ===Stream Rate: 3Mpps===
> No Stream_Rate Downtime Totaltime Ping_Loss Moong
ll to
> ovsthread_id_self().
>
> Signed-off-by: Eelco Chaudron
Fixes: f3e7ec254738 ("Update relevant artifacts to add support for DPDK
17.05.1.")
Acked-by: Ilya Maximets
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
| 19 +--
> lib/ovs-router.h| 3 +++
> vswitchd/ovs-vswitchd.c | 2 ++
> 3 files changed, 22 insertions(+), 2 deletions(-)
Hello, Ben.
Is it possible to backport these two patches to stable branch-2.9?
I have constant unit test failures on my local setup.
Best reg
On 23.05.2018 19:47, Tiago Lam wrote:
> From: Mark Kavanagh
>
> Currently, packets are only copied to a single segment in
> the function dpdk_do_tx_copy(). This could be an issue in
> the case of jumbo frames, particularly when multi-segment
> mbufs are involved.
>
> This patch calculates the nu
From: Ivan Dyukov
The change fixes random stp test failure. Accuracy is about 20%.
Failed test is following:
2337: STP - flush the fdb and mdb when topology changed
In some cases, a validation is executed after topology change and
it increase time of stp stabilization. To prevent this, delay
whi
On 06.06.2018 02:02, Ben Pfaff wrote:
> On Mon, Jun 04, 2018 at 04:49:28PM +0300, Ilya Maximets wrote:
>> On 01.04.2018 03:12, Ben Pfaff wrote:
>>> The --disable-system option indicates that the user wants to avoid using
>>> the host's datapath. This is also a go
ges since 2.7. It'll be hard to track down the root cause.
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
ke sense in v2.
Other than that:
Acked-by: Ilya Maximets
>
> Suggested-by: Ben Pfaff
> Signed-off-by: Aaron Conole
> ---
> lib/netdev-dpdk.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
> index a76
CC: Shahaji Bhosle
Sorry, missed you in CC list.
Best regards, Ilya Maximets.
On 15.06.2018 10:44, Ilya Maximets wrote:
>> Hi,
>> I just upgraded from OvS 2.7 + DPDK 16.11 to OvS2.9 + DPDK 17.11 and
>> running into performance issue with 64 Byte packet rate. One interesti
le other ports will be underloaded. This will
be a bad case for tunnelling where we have only few heavy flows.
As I understood, this version of bonding doesn't support any load
statistics.
Best regards, Ilya Maximets.
> Problem:
>
> In OVS-DPDK, flows with output over a
because successful "POLLING" cycles are now included into "PROCESSING".
Best regards, Ilya Maximets.
On 18.06.2018 16:31, Shahaji Bhosle wrote:
> Hi Ilya,
> Thanks for the quick reply,
> Please find the numbers for our PHY-PHY test, please note that with OVS 2.
3054|rconn|INFO|br-int<->unix#2876: connected
2018-06-18T12:55:10.609Z|03055|rconn|INFO|br-int<->unix#2877: connected
2018-06-18T12:55:12.609Z|03056|rconn|INFO|br-int<->unix#2878: connected
2018-06-18T12:55:12.609Z|03057|rconn|INFO|br-int<->unix#2879: connected
2018-06-1
On 18.06.2018 18:07, Ben Pfaff wrote:
> On Mon, Jun 18, 2018 at 05:18:49PM +0300, Ilya Maximets wrote:
>>> On Wed, May 23, 2018 at 09:28:59PM -0700, Ben Pfaff wrote:
>>>> On Wed, May 23, 2018 at 06:06:44PM -0700, Han Zhou wrote:
>>>>> On Wed, Ma
Hi,
According to your log, your NIC has limited size of tx queues:
2018-06-19T04:34:46.106Z|00089|dpdk|ERR|PMD: Unsupported size of TX queue
(max size: 1024)
This means that you have to configure 'n_txq_desc' <= 1024 in order to
configure your NIC
On 18.06.2018 22:25, Ben Pfaff wrote:
> On Mon, Jun 18, 2018 at 09:31:43AM -0700, Han Zhou wrote:
>> On Mon, Jun 18, 2018 at 8:34 AM, Ilya Maximets
>> wrote:
>>>
>>> On 18.06.2018 18:07, Ben Pfaff wrote:
>>>> On Mon, Jun 18, 2018 at 05:18:49PM +030
On 18.06.2018 22:24, Ben Pfaff wrote:
> On Mon, Jun 18, 2018 at 06:34:13PM +0300, Ilya Maximets wrote:
>> On 18.06.2018 18:07, Ben Pfaff wrote:
>>> On Mon, Jun 18, 2018 at 05:18:49PM +0300, Ilya Maximets wrote:
>>>>> On Wed, May 23, 2018 at 09:28:59PM -0700, Ben P
On 19.06.2018 19:03, Ben Pfaff wrote:
> On Tue, Jun 19, 2018 at 10:58:36AM +0300, Ilya Maximets wrote:
>> On 18.06.2018 22:24, Ben Pfaff wrote:
>>> On Mon, Jun 18, 2018 at 06:34:13PM +0300, Ilya Maximets wrote:
>>>> On 18.06.2018 18:07, Ben Pfaff wrote:
>>&g
-by: Ben Pfaff
Fixes: c9a9b9b00bf5 ("rconn: Introduce new invariant to fix assertion
failure in corner case.")
Signed-off-by: Ilya Maximets
---
lib/rconn.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/rconn.c b/lib/rconn.c
index 7af4e73.
ybe it's the performance issue with bnxt driver that you're using.
There was too many changes in that driver:
30 files changed, 17189 insertions(+), 3358 deletions(-)
Best regards, Ilya Maximets.
On 20.06.2018 01:18, Shahaji Bhosle wrote:
> Hi Ilya,
> This issue is a release b
On 19.06.2018 22:41, Ravi Kerur wrote:
> Hi,
>
>
> On Tue, Jun 19, 2018 at 12:27 AM Ilya Maximets <mailto:i.maxim...@samsung.com>> wrote:
>
> Hi,
> According to your log, your NIC has limited size of tx queues:
>
> 2018-06-19T04:34:46.106Z|000
1 queue 10.4+10.4 14
> 16.11 testpmd 6 queue 21.5 + 21.5 60
> 1 queue 10.4+10.4 14
>
>
> On Wed, Jun 20, 2018 at 4:52 AM, Ilya Maximets <mailto:i.maxim...@samsung.com&g
On 26.06.2018 12:19, Eelco Chaudron wrote:
>
>
> On 22 Jun 2018, at 21:03, Lam, Tiago wrote:
>
>> On 18/06/2018 12:28, Eelco Chaudron wrote:
>>>
>>>
>>> On 11 Jun 2018, at 18:21, Tiago Lam wrote:
>>>
A new mutex, 'nonpmd_mp_mutex', has been introduced to serialise
allocation and free o
On 26.06.2018 15:32, Eelco Chaudron wrote:
>
>
> On 26 Jun 2018, at 12:02, Ilya Maximets wrote:
>
>> On 26.06.2018 12:19, Eelco Chaudron wrote:
>>>
>>>
>>> On 22 Jun 2018, at 21:03, Lam, Tiago wrote:
>>>
>>>> On 18/06/2018 12:28
So, what about this patch?
Best regards, Ilya Maximets.
On 20.06.2018 10:44, Ilya Maximets wrote:
> Recent assertion failure fix changed rconn workflow for unreliable
> connections (such as connections from ovs-ofctl) from
>
> |rconn|DBG|br-int<->unix#151: entering ACTIV
st regards, Ilya Maximets.
On 27.06.2018 04:48, Shahaji Bhosle wrote:
> Hi Ilya,
> Just wanted to check if you found anything interesting. Or anything we can
> try. Thanks, Shahaji
>
> On Wed, Jun 20, 2018 at 9:01 AM, Shahaji Bhosle <mailto:shahaji.bho...@broadcom.com&g
his.
> Thanks, Shahaji
>
>
> On Mon, Jul 2, 2018 at 10:55 AM, Ilya Maximets <mailto:i.maxim...@samsung.com>> wrote:
>
> Hi.
> Sorry for late response.
>
> Looking at your perf data, I see functions like "dp_packet_batch_size"
>
h time your test works for one case?
10. What is the NUMA/threads topology of your system? (To be sure that your
cpu cores are not siblings/located on different NUMA)
Best regards, Ilya Maximets.
On 09.07.2018 20:36, Jay Ding wrote:
> Hi Ilya,
>
> Here is the test result for performanc
Not a full review.
One comment inline.
Best regards, Ilya Maximets.
On 10.07.2018 00:13, Vishal Deep Ajmera wrote:
> OVS reads packets in batches from a given port and packets in the
> batch are subjected to potentially 3 levels of lookups to identify
> the datapath megaflow entry
'alg_exp_entry' is allocated on stack memory, but could be used via
'alg_exp' pointer inside 'write_ct_md' function, i.e. outside its scope.
CC: Darrell Ball
Fixes: bd5e81a0e596 ("Userspace Datapath: Add ALG infra and FTP.")
Signed-off-by: Ilya Maxime
ions was intended to hide internals of batch manipulations.
Also, as I understand, you're storing 'map_cnt' for each missed packet.
So, why not just use 'n_missed' for that purpose?
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On 10.07.2018 15:30, Eelco Chaudron wrote:
>
>
> On 10 Jul 2018, at 13:06, Tiago Lam wrote:
>
>> From: Mark Kavanagh
>>
>> Currently, jumbo frame support for OvS-DPDK is implemented by
>> increasing the size of mbufs within a mempool, such that each mbuf
>> within the pool is large enough to co
On 10.07.2018 17:48, Vishal Deep Ajmera wrote:
>>
>> This is potentially dangerous from the future modifications and hard to read
>> for
>> reviewer/person who tries to understand how it works.
>>
>> Current implementation will fail if someone will change the logic of
>> 'DP_PACKET_BATCH_REFILL_FO
Comments inline.
It'll be nice if you will add me in CC for the next versions.
Thanks.
Best regards, Ilya Maximets.
On 11.07.2018 15:22, Vishal Deep Ajmera wrote:
> OVS reads packets in batches from a given port and packets in the
> batch are subjected to potentially 3 levels of
mbuf, but will be much more elegant, and will allow to eliminate
all the OVS_NOT_REACHED cases.
Best regards, Ilya Maximets.
On 24.07.2018 17:25, Tiago Lam wrote:
> Overview
>
> This patchset introduces support for multi-segment mbufs to OvS-DPDK.
> Multi-segment mbufs are ty
'key.len' should be calculated before inserting to EMC, otherwise
resulting entry will match with any packet with the same hash.
CC: Yipeng Wang
Fixes: 60d8ccae135f ("dpif-netdev: Add SMC cache after EMC cache")
Signed-off-by: Ilya Maximets
---
lib/dpif-netdev.c | 2 +
lt timeout applied to control utils in a subshell.
* tests refactored to be more readable.
* testsuite execution speed up.
Ilya Maximets (9):
tests: Simplify the setting of aliases.
tests: Set default timeout for utils in subshell.
tests: Disable syslog by default for control utils.
test
There is no need to create a separate function for each alias.
This will simplify adding new default options and utils.
Signed-off-by: Ilya Maximets
---
tests/ovs-macros.at | 35 +--
1 file changed, 5 insertions(+), 30 deletions(-)
diff --git a/tests/ovs
.
Signed-off-by: Ilya Maximets
---
tests/ofproto-macros.at | 12 +++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/tests/ofproto-macros.at b/tests/ofproto-macros.at
index 2a56ae6..96219cd 100644
--- a/tests/ofproto-macros.at
+++ b/tests/ofproto-macros.at
@@ -107,7 +107,17
syslog messages from unit tests are not useful and only litter the
system logs on build / test machines.
Signed-off-by: Ilya Maximets
---
tests/ofproto-macros.at | 2 +-
tests/ovs-macros.at | 6 --
2 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/tests/ofproto-macros.at b
'-v' option removed. Found no reason to have fully verbose
output from these ovs-ofctl calls.
Signed-off-by: Ilya Maximets
---
tests/ofproto.at | 16
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/tests/ofproto.at b/tests/ofproto.at
index d6c6e9a..f57c
We can not just make an alias for deamons because many of them
has logging options in their command lines.
Let's handle them one by one. Additionally, it's a good chance
to wrap all the very long lines for better readability.
Signed-off-by: Ilya Maximets
---
tests/ofproto-macros.at
Logs enabled by 'appctl vlog/set' are commonly only used for
'check_logs' at the end of the test. No need to enable any
other logs except for file.
Patch made automatically by sed replace.
Signed-off-by: Ilya Maximets
---
tests/dpif-netdev.at | 10
This disables syslog logging for:
* ovs-testcontroller
* test-netflow
* test-ovsdb
* test-sflow
* test-unixctl (for cases where it's not needed)
Signed-off-by: Ilya Maximets
---
tests/bridge.at | 3 ++-
tests/ofproto-dpif.at
'--log-file' should go after '-v' arguments to avoid unwanted
'opened log file' messages.
Execution is in a subshell and not covered by aliases.
Signed-off-by: Ilya Maximets
---
tests/ovsdb-cluster.at | 6 +-
1 file changed, 5 insertions(+), 1 deletio
This is the only place where ovsdb-tool produces some logs.
Also, it does not support '--timeout' option thus we can't
just add it to the utils list. Let's add syslog option inplace.
Signed-off-by: Ilya Maximets
---
tests/ovsdb-cluster.at | 2 +-
1 file changed, 1 inse
On 01.08.2018 16:34, Aaron Conole wrote:
> Hi Ilya,
>
> Ilya Maximets writes:
>
>> There is no need to create a separate function for each alias.
>> This will simplify adding new default options and utils.
>>
>> Signed-off-by: Ilya Maximets
lt timeout applied to control utils in a subshell.
* tests refactored to be more readable.
* testsuite execution speed up.
Version 2:
* Fixed accidentially missed '--timeout' in patches 1 and 2. [Aaron]
Ilya Maximets (9):
tests: Simplify the setting of aliases.
tests: Set default
There is no need to create a separate function for each alias.
This will simplify adding new default options and utils.
Signed-off-by: Ilya Maximets
---
tests/ovs-macros.at | 35 +--
1 file changed, 5 insertions(+), 30 deletions(-)
diff --git a/tests/ovs
.
Signed-off-by: Ilya Maximets
---
tests/ofproto-macros.at | 12 +++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/tests/ofproto-macros.at b/tests/ofproto-macros.at
index 2a56ae6..96219cd 100644
--- a/tests/ofproto-macros.at
+++ b/tests/ofproto-macros.at
@@ -107,7 +107,17
syslog messages from unit tests are not useful and only litter the
system logs on build / test machines.
Signed-off-by: Ilya Maximets
---
tests/ofproto-macros.at | 2 +-
tests/ovs-macros.at | 6 --
2 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/tests/ofproto-macros.at b
We can not just make an alias for deamons because many of them
has logging options in their command lines.
Let's handle them one by one. Additionally, it's a good chance
to wrap all the very long lines for better readability.
Signed-off-by: Ilya Maximets
---
tests/ofproto-macros.at
Logs enabled by 'appctl vlog/set' are commonly only used for
'check_logs' at the end of the test. No need to enable any
other logs except for file.
Patch made automatically by sed replace.
Signed-off-by: Ilya Maximets
---
tests/dpif-netdev.at | 10
'-v' option removed. Found no reason to have fully verbose
output from these ovs-ofctl calls.
Signed-off-by: Ilya Maximets
---
tests/ofproto.at | 16
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/tests/ofproto.at b/tests/ofproto.at
index d6c6e9a..f57c
This disables syslog logging for:
* ovs-testcontroller
* test-netflow
* test-ovsdb
* test-sflow
* test-unixctl (for cases where it's not needed)
Signed-off-by: Ilya Maximets
---
tests/bridge.at | 3 ++-
tests/ofproto-dpif.at
'--log-file' should go after '-v' arguments to avoid unwanted
'opened log file' messages.
Execution is in a subshell and not covered by aliases.
Signed-off-by: Ilya Maximets
---
tests/ovsdb-cluster.at | 6 +-
1 file changed, 5 insertions(+), 1 deletio
This is the only place where ovsdb-tool produces some logs.
Also, it does not support '--timeout' option thus we can't
just add it to the utils list. Let's add syslog option inplace.
Signed-off-by: Ilya Maximets
---
tests/ovsdb-cluster.at | 2 +-
1 file changed, 1 inse
to have console output.
Signed-off-by: Ilya Maximets
---
This could be a part of syslog clean up patch-set [1], but it looks
more like a separate independent feature.
[1] https://mail.openvswitch.org/pipermail/ovs-dev/2018-August/350425.html
vtep/ovs-vtep.in | 7 ++-
1 file changed, 6 i
On 01.08.2018 18:54, Ilya Maximets wrote:
> Control utils should be called with the same verbose level
> at least to manage output to system logs. For example, to
> disable unwanted syslog messages in unit tests or to enable
> higher debug levels if needed.
> New argumen
Hello. Thanks for the patch.
I'm not much familiar with that functionality thus someone
else should review the sanity of this patch, but I have few
comments about testing itself. See inline.
Best regard, Ilya Maximets.
> Added test for snoop command to check for the initial handshake
On 03.08.2018 17:44, Aaron Conole wrote:
> Ilya Maximets writes:
>
>> Each run of the testsuite produces millions lines in a system
>> log. This is completely unnecessary and makes it difficult to
>> use system logs on test / build servers.
>>
>> This se
On 06.08.2018 16:47, Timothy Redaelli wrote:
> On Wed, 01 Aug 2018 17:00:11 +0300
> Ilya Maximets wrote:
>
>> Aliases are not inheritable. To add a default options for utils
>> executed in subshell we may try to catch them here and append
>> options explicitly.
>
On 04.08.2018 03:17, Ben Pfaff wrote:
> On Wed, Aug 01, 2018 at 05:00:09PM +0300, Ilya Maximets wrote:
>> Each run of the testsuite produces millions lines in a system
>> log. This is completely unnecessary and makes it difficult to
>> use system logs on test / build servers.
On 07.08.2018 00:12, Ben Pfaff wrote:
> On Mon, Aug 06, 2018 at 06:18:19PM +0300, Ilya Maximets wrote:
>> On 04.08.2018 03:17, Ben Pfaff wrote:
>>> On Wed, Aug 01, 2018 at 05:00:09PM +0300, Ilya Maximets wrote:
>>>> Each run of the testsuite produces millions lines
, which was
a bit modified.
Version 2:
* Fixed accidentially missed '--timeout' in patches 1 and 2. [Aaron]
Ilya Maximets (9):
tests: Simplify the setting of aliases.
tests: Set default timeout for utils in subshell.
tests: Disable syslog by default for control utils.
te
There is no need to create a separate function for each alias.
This will simplify adding new default options and utils.
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
---
tests/ovs-macros.at | 35 +--
1 file changed, 5 insertions(+), 30 deletions(-)
diff
.
Signed-off-by: Ilya Maximets
---
tests/ofproto-macros.at | 12 +++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/tests/ofproto-macros.at b/tests/ofproto-macros.at
index 2a56ae6..df8e7c4 100644
--- a/tests/ofproto-macros.at
+++ b/tests/ofproto-macros.at
@@ -107,7 +107,17
syslog messages from unit tests are not useful and only litter the
system logs on build / test machines.
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
---
tests/ofproto-macros.at | 2 +-
tests/ovs-macros.at | 6 --
2 files changed, 5 insertions(+), 3 deletions(-)
diff --git a
We can not just make an alias for daemons because many of them
has logging options in their command lines.
Let's handle them one by one. Additionally, it's a good chance
to wrap all the very long lines for better readability.
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
'-v' option removed. Found no reason to have fully verbose
output from these ovs-ofctl calls.
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
---
tests/ofproto.at | 16
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/tests/ofproto.at b/tests/ofprot
Logs enabled by 'appctl vlog/set' are commonly only used for
'check_logs' at the end of the test. No need to enable any
other logs except for file.
Patch made automatically by sed replace.
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
---
tests/dpif-netdev.at
'--log-file' should go after '-v' arguments to avoid unwanted
'opened log file' messages.
Execution is in a subshell and not covered by aliases.
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
---
tests/ovsdb-cluster.at | 6 +-
1 file changed, 5 insertions
This disables syslog logging for:
* ovs-testcontroller
* test-netflow
* test-ovsdb
* test-sflow
* test-unixctl (for cases where it's not needed)
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
---
tests/bridge.at | 3 ++-
tests/ofproto-dp
This is the only place where ovsdb-tool produces some logs.
Also, it does not support '--timeout' option thus we can't
just add it to the utils list. Let's add syslog option inplace.
Signed-off-by: Ilya Maximets
Acked-by: Aaron Conole
---
tests/ovsdb-cluster.at | 2 +
ocessing, packets from
>> flow map are *appended* to per-flow buffer.
>>
>
> Thanks for the V6 Vishal, looking at this today myself.
>
> Ilya, has the v6 addressed your concerns from the v5?
Hi.
Ian, I had no much time. So, I'm still looking at the patch.
Vishal, It'
On 25.09.2017 18:07, Bodireddy, Bhanuprakash wrote:
> Hi Ilya,
>
> This series needs to be rebased. Few comments below.
Hi. Thanks for review.
I just returned from vacation and starting working on this.
Comments inline.
Best regards, Ilya Maximets.
>
>> While processing
h Policer functionality but
> Just wondering may_steal isn't needed with this change.
This was added by a recent commit
"netdev-dpdk: Execute QoS Checking before copying to mbuf."
And yes, above commit will be mostly reverted, because 'may_steal' is always
true.
>
ter.
* Time based RFC: fixed assert on n_output_batches <= 0.
Version 2:
* Rebased on current master.
* Added time based batching RFC patch.
* Fixed mixing packets with different sources in same batch.
Ilya Maximets (7):
dpif-netdev: Keep latest measured time for PMD thread.
ry and simplify possible modifications
in current logic.
This patch introduces new structure 'dp_netdev_pmd_thread_ctx'
contained by 'struct dp_netdev_pmd_thread' to store any processing
context of this PMD thread. For now, only time and cycles moved to
that structure. Can be extended
is thread local and located in send port cache.
Signed-off-by: Ilya Maximets
---
lib/dpif-netdev.c | 75 +++
1 file changed, 64 insertions(+), 11 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index 1a3d8b4..c6f5c23 100644
---
Not needed anymore because 'may_steal' already handled on
dpif-netdev layer and always true;
Signed-off-by: Ilya Maximets
---
lib/dpif-netdev.c | 2 +-
lib/netdev-bsd.c | 4 ++--
lib/netdev-dpdk.c | 65 +--
lib/netd
Cutlen already applied while processing OVS_ACTION_ATTR_OUTPUT.
Signed-off-by: Ilya Maximets
---
lib/netdev-bsd.c | 2 +-
lib/netdev-dpdk.c | 5 -
lib/netdev-dummy.c | 2 +-
lib/netdev-linux.c | 4 ++--
4 files changed, 4 insertions(+), 9 deletions(-)
diff --git a/lib/netdev-bsd.c b/lib
This fanction will provide monotonic time in microseconds.
Signed-off-by: Ilya Maximets
---
lib/timeval.c | 22 ++
lib/timeval.h | 2 ++
2 files changed, 24 insertions(+)
diff --git a/lib/timeval.c b/lib/timeval.c
index dd63f03..be2eddc 100644
--- a/lib/timeval.c
+++ b/lib
asuring to ensure desired resolution of 'tx-flush-interval'.
Signed-off-by: Ilya Maximets
---
lib/dpif-netdev.c| 141 ---
vswitchd/vswitch.xml | 16 ++
2 files changed, 127 insertions(+), 30 deletions(-)
diff --git a/lib/dpif
Signed-off-by: Ilya Maximets
---
lib/dpif-netdev.c | 27 ++-
1 file changed, 22 insertions(+), 5 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index 3ddb711..1d5321a 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif-netdev.c
@@ -344,6 +344,8 @@ enum
As soon as stats printed for all the files, lines should be
counted for each line separately.
Signed-off-by: Ilya Maximets
---
utilities/checkpatch.py | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/utilities/checkpatch.py b/utilities/checkpatch.py
index 185ddaf..33feb6b
Lines should be counted for each file separately.
Signed-off-by: Ilya Maximets
---
Version 2:
* Fixed commit message.
utilities/checkpatch.py | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/utilities/checkpatch.py b/utilities/checkpatch.py
index 185ddaf..33feb6b
ovs-appctl can wait indefinitely while executing an exit for a
dead service. Let's add a timeout (10 seconds should be
reasonable) to exit calls to avoid hanging up of the testsuite
in such cases.
Signed-off-by: Ilya Maximets
---
tests/ovs-macros.at | 6 +++---
1 file changed, 3 inser
DPDK application (testpmd) inside
guest while heavy traffic flows to this VM.
Fix that by obtaining and checking the 'vid' only once.
CC: Ciara Loftus
Fixes: 0a0f39df1d5a ("netdev-dpdk: Add support for DPDK 16.07")
Signed-off-by: Ilya Maximets
---
lib/netdev-dpdk.c | 14
------
Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On 09.10.2017 09:16, Ilya Maximets wrote:
> On 08.10.2017 12:32, Gao Zhenyu wrote:
>> Hi llya,
>>
>> Thanks for working it. Your patch tried to eliminate the may_steal in dpdk
>> qos, because may_steal handled on dpif-netdev layer and always true.
>> But in
v, qid, batch); <---your patch releases
> some packets in a batch
> dp_packet_delete_batch(batch, may_steal); <---it releases
> all packets in this batch, may hit issue I think
> }
>
> 2017-10-09 14:16 GMT+08:00 Ilya Maximets <mailto:i.max
, I guess, is that I'm using ARMv8 machine for that.
It could be not so easy to reproduce on x86 system (I didn't try).
Best regards, Ilya Maximets.
>
> Thanks,
> Billy.
>
>> -Original Message-
>> From: ovs-dev-boun...@openvswitch.org [mailto:ovs-dev
cenario
on my ARMv8 system:
1.33 with 50ms vs. 1.42 Gbps with 0ms and 1.53 Gbps with 500ms.
I'll share more detailed test results, but it's clear that best time
interval is highly system dependent. This means that we should not make
assumptions about it.
Best regards, Ilya Maximets.
>
month for that:
* https://mail.openvswitch.org/pipermail/ovs-dev/2017-September/338708.html
Maybe it'll help or my case will give you some hints about the root cause.
Unfortunately I had no much time to investigate the issue deeply.
Best regards, Ilya Maximets.
___
Any other thoughts about this patch-set?
It's in kind of 'undecided' state for a long time.
Best regards, Ilya Maximets.
On 04.08.2017 17:17, Ilya Maximets wrote:
> Version 3:
> * Added comment to EM_FLOW_INSERT_INV_PROB_SHIFT.
>
> Ilya Maximets (2):
>
On 25.10.2017 20:28, Ben Pfaff wrote:
> On Fri, Oct 13, 2017 at 01:03:18PM +, Bodireddy, Bhanuprakash wrote:
>>> This fanction will provide monotonic time in microseconds.
>>
>> [BHANU] Typo here with function.
>>
>>>
>>> Signed-off-b
On 26.10.2017 10:12, Ilya Maximets wrote:
> On 25.10.2017 20:28, Ben Pfaff wrote:
>> On Fri, Oct 13, 2017 at 01:03:18PM +, Bodireddy, Bhanuprakash wrote:
>>>> This fanction will provide monotonic time in microseconds.
>>>
>>> [BHANU] Typo here with func
On 11.10.2017 11:51, Eelco Chaudron wrote:
> On 05/10/17 17:05, Ilya Maximets wrote:
>> In current implementation 'now' variable updated once on each
>> receive cycle and passed through the whole datapath via function
>> arguments. It'll be better to kee
gt; can wait in output batch for sending.
>>
>> dpif-netdev turned to microsecond resolution for time measuring to ensure
>> desired resolution of 'tx-flush-interval'.
>>
>> Signed-off-by: Ilya Maximets
>> ---
>> lib/dpif-netdev.c| 141
>> +
assert on n_output_batches <= 0.
Version 2:
* Rebased on current master.
* Added time based batching RFC patch.
* Fixed mixing packets with different sources in same batch.
Ilya Maximets (7):
dpif-netdev: Keep latest measured time for PMD thread.
dpif-netdev: Output packet
1 - 100 of 7269 matches
Mail list logo