I read the code and find this
datapath/conntrack.c : 982
if (info->nat && !(key->ct_state & OVS_CS_F_NAT_MASK) &&
(nf_ct_is_confirmed(ct) || info->commit) &&
ovs_ct_nat(net, key, info, skb, ct, ctinfo) != NF_ACCEPT) {
return -EINVAL;
}
I think
Hi Ben,
On Monday 12 June 2017 08:30 PM, Stokes, Ian wrote:
>> Some pmd driver(e.g: vNIC thunderx PMD) want mbuf_size to be multiple of
>> cache_line_size. With out this fix, Netdev-dpdk initialization would fail
>> for those PMD.
>>
>> Signed-off-by: Santosh Shukla
>> Acked-by: Mark Kavanagh
>
- Parse netlink ct attr OVS_CT_ATTR_EVENTMASK
- Add a new CT_EVENT_TYPE, OVS_EVENT_CT_UPDATE which is triggered
only when CT_ATTR_EVENTMASK is set for MARK and LABEL updates.
Signed-off-by: Anand Kumar
---
datapath-windows/include/OvsDpInterfaceCtExt.h | 1 +
datapath-windows/ovsext/Con
On Fri, Jun 09, 2017 at 03:30:42PM -0700, Darrell Ball wrote:
> From: Ben Pfaff
>
> These will receive their first user (outside of hash.h) in the following
> commit.
>
> Signed-off-by: Ben Pfaff
> Acked-by: Darrell Ball
Thanks! I applied this series to master.
__
On Mon, Jun 12, 2017 at 08:06:01PM -0400, Lance Richardson wrote:
> Eliminate a number of instances of undefined behavior related to
> passing NULL in parameters having "nonnull" annotations.
>
> Found with gcc's undefined behavior sanitizer.
>
> Signed-off-by: Lance Richardson
> ---
>
> Postin
A left shift that would produce a result that is not representable
by the type of the expression's result has "undefined behavior"
according to the C language standard. Avoid this by casting values
that could set the upper bit to unsigned types.
Also document and convert a macro to a function.
Wh
On Mon, Jun 12, 2017 at 07:36:48PM -0400, Lance Richardson wrote:
>
>
> - Original Message -
> > From: "Greg Rose"
> > To: "Lance Richardson"
> > Cc: d...@openvswitch.org
> > Sent: Monday, 12 June, 2017 6:44:43 PM
> > Subject: Re: [ovs-dev] [PATCH] byte-order: avoid left shifts with
>
On Mon, Jun 12, 2017 at 09:33:50PM -0700, Ben Pfaff wrote:
> On Mon, Jun 12, 2017 at 04:21:04PM -0700, Joe Stringer wrote:
> > On 9 June 2017 at 08:58, Flavio Leitner wrote:
> > > The daemon is killed leaving resources behind when a test fails.
> > > This fixes to first signal the daemon to exit g
On Mon, Jun 12, 2017 at 04:21:04PM -0700, Joe Stringer wrote:
> On 9 June 2017 at 08:58, Flavio Leitner wrote:
> > The daemon is killed leaving resources behind when a test fails.
> > This fixes to first signal the daemon to exit gracefully.
> >
> > Fixes: 0f28164be02ac ("netdev-linux: make tap de
On 06/12/2017 04:36 PM, Lance Richardson wrote:
- Original Message -
> From: "Greg Rose"
> To: "Lance Richardson"
> Cc: d...@openvswitch.org
> Sent: Monday, 12 June, 2017 6:44:43 PM
> Subject: Re: [ovs-dev] [PATCH] byte-order: avoid left shifts with
unrepresentable results
>
> On 06/
Eliminate a number of instances of undefined behavior related to
passing NULL in parameters having "nonnull" annotations.
Found with gcc's undefined behavior sanitizer.
Signed-off-by: Lance Richardson
---
Posting this as RFC because there is no apparent risk of
unwanted compiler optimizations r
Hello, My name is Kevin Lin and I work with Ethan on Quilt (quilt.io). We
just started using load balancing for the project -- Ethan wanted me to write to
you all with feedback on the load balancer, and to get some feedback on our
approach.
For context, we have a number of containers connected to
- Original Message -
> From: "Greg Rose"
> To: "Lance Richardson"
> Cc: d...@openvswitch.org
> Sent: Monday, 12 June, 2017 6:44:43 PM
> Subject: Re: [ovs-dev] [PATCH] byte-order: avoid left shifts with
> unrepresentable results
>
> On 06/12/2017 01:13 PM, Lance Richardson wrote:
> > A
On 9 June 2017 at 08:58, Flavio Leitner wrote:
> The daemon is killed leaving resources behind when a test fails.
> This fixes to first signal the daemon to exit gracefully.
>
> Fixes: 0f28164be02ac ("netdev-linux: make tap devices persistent")
> Suggested-by: Joe Stringer
> Co-authored-by: Ben P
On Thu, Jun 8, 2017 at 6:20 PM, Ben Pfaff wrote:
>
> On Thu, Jun 08, 2017 at 04:21:17PM -0700, Han Zhou wrote:
> > On Thu, Jun 8, 2017 at 4:14 PM, Ben Pfaff wrote:
> > >
> > > On Tue, Jun 06, 2017 at 05:24:19PM -0700, Han Zhou wrote:
> > > > On Tue, Jun 6, 2017 at 3:56 PM, Ben Pfaff wrote:
> > >
This keeps lines from exceeding 79 columns.
Signed-off-by: Ben Pfaff
---
lib/ofp-print.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/ofp-print.c b/lib/ofp-print.c
index 5b37ed8f46ba..6b994cd17c94 100644
--- a/lib/ofp-print.c
+++ b/lib/ofp-print.c
@@ -218,9 +218,
Signed-off-by: Ben Pfaff
---
ofproto/ofproto.c | 6 --
tests/ofproto.at | 23 ---
2 files changed, 29 deletions(-)
diff --git a/ofproto/ofproto.c b/ofproto/ofproto.c
index 5794dfd54fc5..f5aaf0b1fd44 100644
--- a/ofproto/ofproto.c
+++ b/ofproto/ofproto.c
@@ -3483,12 +348
Signed-off-by: Ben Pfaff
---
lib/ofp-parse.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/ofp-parse.c b/lib/ofp-parse.c
index d7f74c949ed7..528b75b4f4e1 100644
--- a/lib/ofp-parse.c
+++ b/lib/ofp-parse.c
@@ -674,7 +674,7 @@ parse_ofp_packet_out_str__(struct ofputil_pack
Signed-off-by: Ben Pfaff
---
ofproto/ofproto-dpif-xlate.c | 5 -
1 file changed, 5 deletions(-)
diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
index 33af5e0a29e4..c7df2d1bdc38 100644
--- a/ofproto/ofproto-dpif-xlate.c
+++ b/ofproto/ofproto-dpif-xlate.c
@@ -4296,11 +
The code now suppresses the default packet_type.
Signed-off-by: Ben Pfaff
---
tests/ofp-print.at | 12 ++--
tests/ofproto.at | 2 +-
2 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/tests/ofp-print.at b/tests/ofp-print.at
index fc444f62bf4c..a00e3f3fe74b 100644
--- a/tes
Tests that modify the system state or that require special privileges
should only be in the system tests, not the regular unit tests.
Signed-off-by: Ben Pfaff
---
tests/packet-type-aware.at | 462 -
1 file changed, 462 deletions(-)
diff --git a/tests/
From: Zoltán Balogh
Introducing packet_type in OF 1.5 packet-out.
Partly based on Jean Tourrilhes's work.
Add test cases for OF1.5 packet-out
Add negative test case for OF1.5 packet-out
Modify wildcarding and packet-out test printout.
Signed-off-by: Jean Tourrilhes
Signed-off-by: Zoltan Balogh
Signed-off-by: Ben Pfaff
---
tests/ofproto.at| 4 ++--
tests/system-userspace-packet-type-aware.at | 2 +-
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/tests/ofproto.at b/tests/ofproto.at
index dc0a90c873ed..67eeced58ebd 100644
--- a/tests/ofproto.at
On 06/12/2017 01:13 PM, Lance Richardson wrote:
A left shift that would produce a result that is not representable
by the type of the expression's result has "undefined behavior"
according to the C language standard. Avoid this by casting values
that could set the upper bit to unsigned types.
Fo
Enable OVN_Southbound RBAC by default in the sandbox environment,
provide a new option "--no-ovn-rbac" to disable it.
Signed-off-by: Lance Richardson
---
tutorial/ovs-sandbox | 17 +
1 file changed, 13 insertions(+), 4 deletions(-)
diff --git a/tutorial/ovs-sandbox b/tutorial/ov
From: Jan Scheurich
ptap - create packet-type-aware bridge
ptap - legal flow entries in ptap bridge
ptap - triangle bridge setup with L2 and L3 GRE tunnels
First and second unit tests perform basic verification.
The third one is a triangular bridge setup test case. It tests dataplane
in non-
From: Jan Scheurich
Send packet_in for non-Ethernet packets.
Include packet_type in Packet In for ptap bridges.
Signed-off-by: Jan Scheurich
Signed-off-by: Ben Pfaff
---
lib/flow.c | 4
lib/ofp-print.c | 3 +--
ofproto/ofproto-dpif-xlate.c | 10 +-
Since packet_type is now accepted on any bridge, there's no need to test
for it being rejected.
Signed-off-by: Ben Pfaff
---
tests/packet-type-aware.at | 59 +-
1 file changed, 6 insertions(+), 53 deletions(-)
diff --git a/tests/packet-type-aware.at b
Signed-off-by: Ben Pfaff
---
ofproto/ofproto-dpif-xlate.c | 2 +-
ofproto/tunnel.c | 17
ofproto/tunnel.h | 2 +-
vswitchd/vswitch.xml | 62
4 files changed, 58 insertions(+), 25 deletions(-)
diff --git a
Signed-off-by: Ben Pfaff
---
lib/netdev-native-tnl.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/netdev-native-tnl.c b/lib/netdev-native-tnl.c
index 515b9948cabb..7f3cf984e887 100644
--- a/lib/netdev-native-tnl.c
+++ b/lib/netdev-native-tnl.c
@@ -580,7 +580,7 @@ netdev
Locks weren't always released on error paths.
Signed-off-by: Ben Pfaff
---
lib/netdev-native-tnl.c | 11 +++
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/lib/netdev-native-tnl.c b/lib/netdev-native-tnl.c
index 658c871015f9..515b9948cabb 100644
--- a/lib/netdev-native-tnl
From: Jan Scheurich
In netdev_gre_build_header(), GRE protocol and VXLAN next_potocol is set based
on packet_type of flow. If it's about an Ethernet packet, it is set to
ETP_TYPE_TEB. Otherwise, if the name space is OFPHTN_ETHERNET, it is set
according to the name space type.
Signed-off-by: Jan
The policy originally proposed by this patch series was to make support for
a packet type aware pipeline a property of each bridge, via
other-config:packet-type-aware. Global modes like this have implications
for testing, that mean that many features need to be tested in both modes.
This commit c
Signed-off-by: Ben Pfaff
---
include/openvswitch/ofp-util.h | 2 --
1 file changed, 2 deletions(-)
diff --git a/include/openvswitch/ofp-util.h b/include/openvswitch/ofp-util.h
index bbf6ffec5dd3..07723b427ce8 100644
--- a/include/openvswitch/ofp-util.h
+++ b/include/openvswitch/ofp-util.h
@@ -24
An upcoming commit will need to pass an extra piece of data from
nx_put_raw() into all of its direct and indirect calls to nxm_put__().
This commit prepares for that by switching from a "struct ofpbuf *"
parameter to a context structure that, currently, contains just a
struct ofpbuf *. The upcomin
The packet type is even more basic to a packet than L2 fields, which OVS
has always matched on for a long time. It simplifies reasoning about
flow translation to simply always include the match in ODP translation.
At the same time, when we're displaying an OpenFlow flow, as opposed to
a datapath
This will receive its first users in an upcoming commit.
Signed-off-by: Ben Pfaff
---
include/openvswitch/ofpbuf.h | 1 +
lib/ofpbuf.c | 18 ++
2 files changed, 19 insertions(+)
diff --git a/include/openvswitch/ofpbuf.h b/include/openvswitch/ofpbuf.h
index bc25b
Also, OVS really supports the packet-type field in any version, since it
accepts OXM even in OpenFlow 1.0.
Signed-off-by: Ben Pfaff
---
vswitchd/vswitch.xml | 9 -
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/vswitchd/vswitch.xml b/vswitchd/vswitch.xml
index 829c13688f31
A previous patch moved this but it's not clear why, so move it back.
Signed-off-by: Ben Pfaff
---
ofproto/ofproto-dpif-xlate.c | 32
1 file changed, 16 insertions(+), 16 deletions(-)
diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
index
Signed-off-by: Ben Pfaff
---
include/openvswitch/ofp-errors.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/include/openvswitch/ofp-errors.h b/include/openvswitch/ofp-errors.h
index aeb58e012fd3..7d75386352d8 100644
--- a/include/openvswitch/ofp-errors.h
+++ b/include/open
Signed-off-by: Ben Pfaff
---
ofproto/ofproto-dpif.c | 9 +
ofproto/ofproto-provider.h | 3 +++
ofproto/ofproto.c | 7 ++-
3 files changed, 18 insertions(+), 1 deletion(-)
diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
index 752d2304a163..098db18c7ffa 100644
From: Jan Scheurich
New boolean parameter "other-config:packet-type-aware' in bridge table.
Pass non-Ethernet packets unchanged into packet-type-aware bridges.
Do not convert packet type when sending packet from packet-type-aware
bridge to a port.
Only include field MFF_PACKET_TYPE in matchabale
Signed-off-by: Ben Pfaff
---
lib/meta-flow.xml | 2 ++
lib/ofp-parse.c | 2 ++
2 files changed, 4 insertions(+)
diff --git a/lib/meta-flow.xml b/lib/meta-flow.xml
index 82daaaffcb44..283a1cb0bcb8 100644
--- a/lib/meta-flow.xml
+++ b/lib/meta-flow.xml
@@ -314,6 +314,8 @@ tcp,tp_src=0x07c0/0xfff
Signed-off-by: Ben Pfaff
---
lib/flow.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/lib/flow.h b/lib/flow.h
index e76adac7e0d4..f61d6c3aa106 100644
--- a/lib/flow.h
+++ b/lib/flow.h
@@ -83,8 +83,7 @@ void format_flags(struct ds *ds, const char
*(*bit_to_string)(uint32_t
Signed-off-by: Ben Pfaff
---
lib/meta-flow.xml | 131 +-
1 file changed, 90 insertions(+), 41 deletions(-)
diff --git a/lib/meta-flow.xml b/lib/meta-flow.xml
index c94288e4c294..82daaaffcb44 100644
--- a/lib/meta-flow.xml
+++ b/lib/meta-flow.xm
From: Jan Scheurich
Allow packet type namespace OFPHTN_ETHERTYPE as alternative pre-requisite
for matching L3 protocols (MPLS, IP, IPv6, ARP etc).
Change the meta-flow definition of packet_type field to use the new
custom format MFS_PACKET_TYPE representing "(NS,NS_TYPE)".
Parsing routine for M
I don't see value in making packet_type formatting different from other
fields.
Also, format_packet_type_masked() added a trailing comma to its output,
which is surprising, so this patch removes it.
Signed-off-by: Ben Pfaff
---
lib/flow.c | 42 ---
lib/match.c
Signed-off-by: Ben Pfaff
---
include/openvswitch/meta-flow.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/include/openvswitch/meta-flow.h b/include/openvswitch/meta-flow.h
index 3a17d7b0618e..fc109501d869 100644
--- a/include/openvswitch/meta-flow.h
+++ b/include/openvs
This series is based on Zoltan Balogh's series here:
https://patchwork.ozlabs.org/patch/770490/
https://patchwork.ozlabs.org/patch/770487/
https://patchwork.ozlabs.org/patch/770495/
https://patchwork.ozlabs.org/patch/770498/
https://patchwork.ozlabs.org/patch/770488/
https://patchwork.ozlabs.org/pa
Greetings!
Would you be interested in reviewing our *Boat Owners List* within the USA ?
*Our Databases Include:-* *1.* Sailboat Owners List *2.* Powerboat
Owners List *3.* Harley Davidson Owners List *4.* Travelers List *5.*
Cruise Travelers List *6.* Fishing Enthusiasts List *7.* Outdoor
En
It has been useful to attach gdb to the running process. With this commit
we make it a little easier, as the daemon will have a gdbserver process
attached and listening on a specific port.
Signed-off-by: Aaron Conole
---
utilities/ovs-ctl.in | 2 +-
utilities/ovs-lib.in | 14 ++
2
A left shift that would produce a result that is not representable
by the type of the expression's result has "undefined behavior"
according to the C language standard. Avoid this by casting values
that could set the upper bit to unsigned types.
Found via gcc's undefined behavior sanitizer.
Signe
On 9 June 2017 at 15:16, Yin Lin wrote:
> From: Anand Kumar
>
> Add support for parsing netlink attributes related to NAT
> in conntrack.
>
> Co-Authored-by: Yin Lin
> Co-Authored-by: Darrell Ball
> Signed-off-by: Anand Kumar
> Signed-off-by: Yin Lin
> Signed-off-by: Darrell Ball
>
Thanks.
Hi Bhanu,
Bhanuprakash Bodireddy writes:
> This commit introduces the initial keepalive support by adding
> 'keepalive' module and also helper and initialization functions
> that will be invoked by later commits.
>
> This commit adds new ovsdb column "keepalive". It shows the overall
> datapath
On Mon, Jun 12, 2017 at 08:18:22AM -0700, Greg Rose wrote:
> On 06/09/2017 07:47 PM, Ben Pfaff wrote:
> >Signed-off-by: Ben Pfaff
> >
> >
> Acked-by: Greg Rose
Thanks! I applied this to master.
___
dev mailing list
d...@openvswitch.org
https://mail.op
Thanks for the ack. I applied this to master and branch-2.7.
On Mon, Jun 12, 2017 at 10:24:06AM -0700, Jarno Rajahalme wrote:
> Acked-by: Jarno Rajahalme mailto:ja...@ovn.org>>
>
> Maybe it would be worthwhile elaborating on the subtle difference in the
> commit message? I.e., “in_port” should
On Mon, Jun 12, 2017 at 01:24:37PM -0400, Lance Richardson wrote:
> > From: "Ben Pfaff"
> > To: d...@openvswitch.org
> > Cc: "Ben Pfaff"
> > Sent: Monday, 12 June, 2017 12:54:42 PM
> > Subject: [ovs-dev] [PATCH] ofp-util: Initialize tunnel metadata for
> > OpenFlow1.0 matches.
> >
> > F
Hi Bhanu,
Bhanuprakash Bodireddy writes:
> With this commit the keepalive framework is invoked when enable-keepalive
> is set to 'true' in ovsdb. As part of initialization the shared memory block
> is created and initialized. Also the keepalive data structure is initialized
> and the callback fu
> From: "Ben Pfaff"
> To: d...@openvswitch.org
> Cc: "Ben Pfaff"
> Sent: Monday, 12 June, 2017 12:54:42 PM
> Subject: [ovs-dev] [PATCH] ofp-util: Initialize tunnel metadata for OpenFlow
> 1.0 matches.
>
> Found by valgrind.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/ofp-util.c | 1 +
> 1 file
Acked-by: Jarno Rajahalme mailto:ja...@ovn.org>>
Maybe it would be worthwhile elaborating on the subtle difference in the commit
message? I.e., “in_port” should match to special OpenFlow port number that
outputs to the incoming port, while normally output to the incoming port would
be suppresse
Keepalive feature is aimed at achieving Fastpath Service Assurance
in OVS-DPDK deployments. It adds support for monitoring the packet
processing cores(PMD thread cores) by dispatching heartbeats at regular
intervals. Incase of heartbeat misses additional health checks are
enabled on the PMD thread
This commit adds support to display the extended keepalive statistics
stored in SHM block. The status can be displayed as follows.
$ ovs-appctl keepalive/pmd-xstats-show
keepAlive Status : Enabled
keepAlive Interval: 1000 ms
CORE: 0
PMD thread id : 1269 [ACTIVE]
PMD he
This commit adds the support to check the PMD cycle stats. If the cycles
aren't changing for a duration of time this can be flagged as possible
PMD stall.
Signed-off-by: Bhanuprakash Bodireddy
---
lib/dpif-netdev.c | 18 +++---
lib/dpif-netdev.h | 6 ++
lib/keepalive.c | 52 ++
The keepalive thread sends heartbeats to PMD thread and when PMD fails to
respond to successive heartbeats the PMD is potentially stalled. The PMD
state transition is as below:
ALIVE -> MISSING -> DEAD -> GONE
This commit enables PMD healthchecks when PMD doesn't respond to
heartbeats. This is ne
This commit adds the support to check the packet statistics on the port
polled by PMD thread. If the packets aren't processed due to PMD thread
stall/deadlock the statistics wont update and this can be used by
monitoring framework to confirm PMD failure.
This mechanism has limitation with MQ enabl
This commit adds the initial support in to performing PMD health checks.
The ports handled by the PMD threads are checked for the link status and
the same is updated in to SHM block.
Signed-off-by: Bhanuprakash Bodireddy
---
lib/dpif-netdev.c | 42 --
lib/keepalive.c
This commit enables additional datapath health checks. The checks
are enabled only on a PMD heartbeat failure. On missing three successive
heartbeats additional health checks needs to be performed on respective
PMD thread to confirm the failure.
The datapath health is monitored periodically from k
This commit adds an API to store the PMD thread id in SHM block.
The tid later shall be retrieved by callback function that gets invoked to
check for false positives.
Signed-off-by: Bhanuprakash Bodireddy
---
lib/dpif-netdev.c | 1 +
lib/keepalive.c | 23 +++
lib/keepalive
This commit adds support to query if keepalive status is
enabled/disabled.
$ ovs-appctl keepalive/status
keepAlive Status: Enabled
Signed-off-by: Bhanuprakash Bodireddy
---
lib/keepalive.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/lib/keepalive.c b/lib/keepalive
This commit allows vswitchd thread to update the OVSDB with the
datapath status along with the status of all registered PMD threads.
The status can be monitored using ovsdb-client.
$ ovsdb-client monitor Open_vSwitch Open_vSwitch keepalive
rowaction keepalive
7b7
This commit adds support to query keepalive statistics stored in
posix shared memory block. Datapath health status can be retrieved as
follows:
$ ovs-appctl keepalive/pmd-health-show
Keepalive status
keepalive status : Enabled
keepalive interval: 1000 ms
CORESTATE
This commit implements APIs to retrieve the PMD thread status and return
the status in the below format for each PMD thread.
Format: CORE_="STATUS, last_seen_timestamp"
eg: CORE_1="ALIVE,9220698256784207"
CORE_2="GONE,9220698256786231"
The status is periodically retrieved by kee
This commit adds heartbeat mechanism support for DPDK datapath. Heartbeats
are sent to registered PMD threads at predefined intervals (as set in ovsdb
with 'keepalive-interval').
The heartbeats are only enabled when there is atleast one port added to
the bridge and with active PMD thread polling t
This commit registers the packet processing PMD cores to keepalive
framework. Only PMDs that have rxqs mapped will be registered and
actively monitored by KA framework.
This commit spawns a keepalive thread that will dispatch heartbeats to
PMD cores. The pmd threads respond to heartbeats by markin
This commit introduces helper functions in 'keepalive' module that are
needed to register/unregister PMD threads to KA framework. Also
introduce APIs to mark the PMD core states.
Signed-off-by: Bhanuprakash Bodireddy
---
lib/keepalive.c | 56 ++
With this commit the keepalive framework is invoked when enable-keepalive
is set to 'true' in ovsdb. As part of initialization the shared memory block
is created and initialized. Also the keepalive data structure is initialized
and the callback functions are registered if DPDK datapath is enabled.
This commit introduces the initial keepalive support by adding
'keepalive' module and also helper and initialization functions
that will be invoked by later commits.
This commit adds new ovsdb column "keepalive". It shows the overall
datapath status and the health of the cores running datapath thr
Implement helper function to retrieve the process status. This will be
used by keepalive monitoring thread to detect false alarms and to show
PMD thread state in future commits.
Signed-off-by: Bhanuprakash Bodireddy
---
lib/process.c | 73 +
Keepalive feature is aimed at achieving Fastpath Service Assurance
in OVS-DPDK deployments. It adds support for monitoring the packet
processing cores(PMD thread cores) by dispatching heartbeats at regular
intervals. Incase of heartbeat misses additional health checks are
enabled on the PMD thread
Introduce helper functions in 'dpdk' module that are needed for
DPDK keepalive functionality. Also add dummy functions in 'dpdk-stub' module
that are needed when DPDK datapath is not available.
Signed-off-by: Bhanuprakash Bodireddy
---
lib/dpdk-stub.c | 24
lib/dpdk.c
Found by valgrind.
Signed-off-by: Ben Pfaff
---
lib/ofp-util.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/ofp-util.c b/lib/ofp-util.c
index 601a719797e5..da171cdf6fe5 100644
--- a/lib/ofp-util.c
+++ b/lib/ofp-util.c
@@ -159,6 +159,7 @@ ofputil_match_from_ofp10_match(const struct ofp
On Fri, Jun 09, 2017 at 10:46:56PM -0700, nickcooper-zhangtonghao wrote:
> When I test OvS openflow, I confuse the 'in_port' with
> 'output:in_port'. I doc the openflow.rst and it may
> help others to avoid it.
>
> Signed-off-by: nickcooper-zhangtonghao
> ---
> Documentation/faq/openflow.rst | 4
It was being misinterpreted as output:NXM_OF_IN_PORT[], which is subtly
different and doesn't do anything useful.
CC: Jarno Rajahalme
Fixes: 21b2fa617126 ("ofp-parse: Allow match field names in actions and
brackets in matches.")
Reported-by: nickcooper-zhangtonghao
Signed-off-by: Ben Pfaff
---
On 06/09/2017 07:47 PM, Ben Pfaff wrote:
Signed-off-by: Ben Pfaff
---
lib/nx-match.c | 4 ++--
lib/ofp-actions.c| 2 +-
ofproto/ofproto-dpif-xlate.c | 2 +-
tests/ofproto-dpif.at| 8
4 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/l
On 09/06/2017 23:04, Flavio Leitner wrote:
On Thu, Jun 08, 2017 at 02:46:35PM +0300, Roi Dayan wrote:
From: Paul Blakey
Currently only tunnel offload is supported.
Signed-off-by: Paul Blakey
Reviewed-by: Roi Dayan
Reviewed-by: Simon Horman
---
Happy that there was no reason to hardcode
On 08/06/2017 17:30, Simon Horman wrote:
On Thu, Jun 08, 2017 at 02:46:19PM +0300, Roi Dayan wrote:
Add tc module to expose tc operations to be used by other modules.
Move some tc related functions from netdev-linux.c to tc.c
This patch doesn't change any functionality.
Signed-off-by: Paul Bl
> Some pmd driver(e.g: vNIC thunderx PMD) want mbuf_size to be multiple of
> cache_line_size. With out this fix, Netdev-dpdk initialization would fail
> for those PMD.
>
> Signed-off-by: Santosh Shukla
> Acked-by: Mark Kavanagh
> Tested-by: Mark Kavanagh
> ---
> v1 --> v2
> - Removed mtu to dm
On 09/06/2017 21:37, Flavio Leitner wrote:
On Thu, Jun 08, 2017 at 02:46:22PM +0300, Roi Dayan wrote:
Add tc helper functions to query and manipulate the flower classifier.
Signed-off-by: Paul Blakey
Co-authored-by: Roi Dayan
Signed-off-by: Roi Dayan
Acked-by: Flavio Leitner
Not sure w
On 08/06/2017 17:24, Simon Horman wrote:
On Thu, Jun 08, 2017 at 02:46:18PM +0300, Roi Dayan wrote:
Refactor tc_make_request and tc_add_del_ingress_qdisc to accept
ifindex instead of netdev struct.
We later want to move those outside netdev-linux module to be
used by other modules.
This patch
Some pmd driver(e.g: vNIC thunderx PMD) want mbuf_size to be multiple of
cache_line_size. With out this fix, Netdev-dpdk initialization would
fail for those PMD.
Signed-off-by: Santosh Shukla
Acked-by: Mark Kavanagh
Tested-by: Mark Kavanagh
---
v1 --> v2
- Removed mtu to dmp->mtu change in v2.
I think this patch adds a useful information to the PMD stats,
especially for debugging/tuning purposes.
In order to catch some performance degradation I've checked it with a
simple port-2-port bidirectional test. I've checked with 1 flow, 10,000
and 100,000 flows (Ixia streams). In all cases I co
Hi Mark,
On Monday 12 June 2017 07:28 PM, Kavanagh, Mark B wrote:
>> From: Stokes, Ian
>> Sent: Monday, June 12, 2017 1:41 PM
>> To: Santosh Shukla ; b...@ovn.org;
>> db...@vmware.com;
>> d...@openvswitch.org
>> Cc: ktray...@redhat.com; hemant.agra...@nxp.com;
>> jerin.ja...@caviumnetworks.com;
>From: Stokes, Ian
>Sent: Monday, June 12, 2017 1:41 PM
>To: Santosh Shukla ; b...@ovn.org;
>db...@vmware.com;
>d...@openvswitch.org
>Cc: ktray...@redhat.com; hemant.agra...@nxp.com;
>jerin.ja...@caviumnetworks.com; Kavanagh,
>Mark B
>Subject: RE: [PATCH v2] netdev-dpdk: round up mbuf_size to ca
Patch is marked as RFC because of DPDK 17.05 possible crash:
http://dpdk.org/dev/patchwork/patch/25012/
Upgrading to DPDK 17.05 adds new significant features
relevant to OVS, including, but not limited to:
- vhost pmd,
- tun/tap PMD,
- VFIO hotplug support,
- Generic flow API.
Following changes a
Hi Ian,
On Monday 12 June 2017 06:11 PM, Stokes, Ian wrote:
>> Subject: [PATCH v2] netdev-dpdk: round up mbuf_size to cache_line_size
>>
>> Some pmd driver(e.g: vNIC thunderx PMD) want mbuf_size to be multiple of
>> cache_line_size. With out this fix, Netdev-dpdk initialization would fail
>> for
> Subject: [PATCH v2] netdev-dpdk: round up mbuf_size to cache_line_size
>
> Some pmd driver(e.g: vNIC thunderx PMD) want mbuf_size to be multiple of
> cache_line_size. With out this fix, Netdev-dpdk initialization would fail
> for those PMD.
>
> Signed-off-by: Santosh Shukla
> ---
> v1 --> v2
>
On 08/06/2017 20:08, Joe Stringer wrote:
On 8 June 2017 at 05:05, Roi Dayan wrote:
On 31/05/2017 04:37, Joe Stringer wrote:
On 28 May 2017 at 04:59, Roi Dayan wrote:
From: Paul Blakey
Add a new API interface for offloading dpif flows to netdev.
The API consist on the following:
flo
DPDK provides an API to set the MTU of compatible physical devices -
rte_eth_dev_set_mtu(). Prior to DPDK v16.07 however, this API was not
implemented in some DPDK PMDs (i40e, specifically). To allow the use
of jumbo frames with affected NICs in OvS-DPDK, MTU configuration was
achieved by setting t
>From: Aaron Conole [mailto:acon...@redhat.com]
>Sent: Friday, June 9, 2017 8:26 PM
>To: Kavanagh, Mark B
>Cc: ovs-dev@openvswitch.org; Varghese, Vipin
>Subject: Re: [ovs-dev] [PATCH] netdev-dpdk: use rte_eth_dev_set_mtu
>
>Hi Mark,
>
>Mark Kavanagh writes:
>
>> DPDK provides an API to set the M
99 matches
Mail list logo
