Hey Joe,
Great to see this patch, one comment below:
diff --git a/ofproto/ofproto-dpif-upcall.c b/ofproto/ofproto-dpif-upcall.c
> index 416cfdc..906bf17 100644
> --- a/ofproto/ofproto-dpif-upcall.c
> +++ b/ofproto/ofproto-dpif-upcall.c
> @@ -45,6 +45,8 @@
>
> VLOG_DEFINE_THIS_MODULE(ofproto_dpi
Though we've discussed it, at this point there is no public bug
tracker for the OVS project. We just moved to github, so there may be
one there, but to date we haven't used it.
If you want to see what went into the 1.10 release, you may consider
checking the 1.10 git log for the 1.10 branch. We
My understanding of the implementation of patch ports is that they
are rather special, being handled as a special case inside
compose_output_action__() and do not exist in the datapath.
A side effect of the implementation of patch ports (though perhaps not the
portion mentioned above) is that the
When the datapath flow table is modified while a flow dump operation is
in progress, it is possible for the same flow to be dumped twice. In
such cases, revalidators may perform redundant work, or attempt to
delete the same flow twice.
This was causing intermittent testsuite failures for test #670
When the datapath flow table is modified while a flow dump operation is
in progress, it is possible for the same flow to be dumped twice. In
such cases, revalidators may perform redundant work, or attempt to
delete the same flow twice.
This was causing intermittent testsuite failures for test #670
Hi All,
Please guide me to find out the openvswitch 1.10.0 bugs list and their
fixes.
--
Thanks & Regards
Neelakantam Gaddam
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
OK, thanks.
There might already be a way to figure out when OVS restarts. Have you
looked through the existing configuration schema to try to find one?
On Apr 22, 2014 7:18 PM, "Ryan Wilson 76511" wrote:
> I agree, this makes much more sense. We could add this restart tag to the
> Open_vSwitch t
I agree, this makes much more sense. We could add this restart tag to the
Open_vSwitch table instead, so the controller team could still do event-based
triggering on database changes. Also, this would clearly separate the cases of
a port being deleted and re-added (uuid changing in Interface tab
It would make a lot more sense to have a global indication whether OVS was
restarted. This is a property of OVS, not of a netdev. Otherwise we'll get
subsequent requests to add such a random tag to every single table in the
database.
You can already tell if a port is deleted and readded by noticin
> I don't understand why, if you need to know when vswitchd restarts (why do
> you?), it should be tied to netdevs.
So if you're monitoring a netdev for packet stats, you need to know if
vswitchd restarted so that you can take account of the fact the stats
were reset. I suppose it needs to be tie
I understand now. Thank you. This patch looks good, then.
Please add your explanation to the commit message.
On Apr 22, 2014 5:28 PM, "YAMAMOTO Takashi" wrote:
> > On Tue, Apr 22, 2014 at 01:47:32PM +0900, YAMAMOTO Takashi wrote:
> >> Improve random distribution for an hmap with a small number o
> On Tue, Apr 22, 2014 at 01:47:32PM +0900, YAMAMOTO Takashi wrote:
>> Improve random distribution for an hmap with a small number of nodes
>> with the expense of the increased cpu cost.
>> It would be a fair trade-off because the situation is rather common
>> for bond, which is currently the only
I don't understand why, if you need to know when vswitchd restarts (why do
you?), it should be tied to netdevs.
On Apr 22, 2014 4:09 PM, "Ryan Wilson 76511" wrote:
> netdev: Add random tag to struct netdev.
>
>
> Before, there would be no way to tell if ovs-vswitchd had
> been restarted or killed
On Wed, Apr 23, 2014 at 09:36:30AM +1200, Joe Stringer wrote:
> On 23 April 2014 06:19, Ben Pfaff wrote:
> >
> > +Please read SubmittingPatches and CodingStyle in the top of the source
> >
>
> SubmittingPatches -> CONTRIBUTING
>
>
> +OpenFlow Group Bucket Stats
> > +---
netdev: Add random tag to struct netdev.
Before, there would be no way to tell if ovs-vswitchd had
been restarted or killed via ovsdb logging. Now, a random
tag will be generated upon initialization of the netdev
struct which happens when ovs-vswitchd is restarted. A
change in tag value in
This code path currently does not initialize
netdev_class_mutex.
dummy_enable
->netdev_dummy_register
->netdev_register_provider
->ovs_mutex_lock(&netdev_class_mutex)
ovsdb-server on windows crashes without it.
This commit adds a new initialization function.
Signed-off-by: Gurucharan Sh
Currently, we are calling guarded_list_destroy()
to destroy a mutex and then go ahead and use it through
delete_group
->delete_group__
->handle_flow_mod__
->run_rule_executes
->guarded_list_pop_all
The group related unit tests cause ovs-vswitchd to crash
because of this (on windows).
On Tue, Apr 22, 2014 at 01:40:55PM -0700, Gurucharan Shetty wrote:
> > Should OVS chdir to the root (on all drives?) on Windows
> > also?
> Looks like chdir("/") on windows takes me to C:/ . So doing that on
> windows should be good. I will re-spin this.
OK.
If Windows works like it used to, ther
On Tue, Apr 22, 2014 at 01:08:10PM -0700, Gurucharan Shetty wrote:
> On Tue, Apr 22, 2014 at 9:35 AM, Ben Pfaff wrote:
> > On Fri, Apr 18, 2014 at 11:04:14AM -0700, Gurucharan Shetty wrote:
> >> The option won't have any effect on the running of the daemon.
> >> Recognizing the option lets us avoi
Yes, thanks.
Acked-by: Ben Pfaff
On Tue, Apr 22, 2014 at 12:54:31PM -0700, Gurucharan Shetty wrote:
> Something like this, right?
>
> diff --git a/tests/atlocal.in b/tests/atlocal.in
> index 06e7384..60f3dbf 100644
> --- a/tests/atlocal.in
> +++ b/tests/atlocal.in
> @@ -78,3 +78,13 @@ FreeBSD|
On Tue, Apr 22, 2014 at 06:27:18PM +0100, Zoltan Kiss wrote:
> This function returns true when 's' is negative or greater than UINT_MAX.
> Also,
> the representation of 'int' and 'unsigned int' is implementation dependent, so
> converting [INT_MAX..UINT_MAX] values with str_to_int is fragile.
> In
On 23 April 2014 06:19, Ben Pfaff wrote:
>
> +Please read SubmittingPatches and CodingStyle in the top of the source
>
SubmittingPatches -> CONTRIBUTING
+OpenFlow Group Bucket Stats
> +---
> +
> +When OpenFlow group support was added, we forgot to support statistics
> +f
Signed-off-by: Pritesh Kothari
---
v4: rebase for changes in datapath/actions.c and datapath/vport.c,
split patch in two for seperating out skb_clear_hash related stuff.
v3: rebase for changes in datapath/vport.c.
v2: Use OVS_GREP_IFELSE instead of upstream kernel versions as Thomas suggested.
Signed-off-by: Pritesh Kothari
---
acinclude.m4 | 1 +
datapath/actions.c | 10 +-
datapath/linux/compat/include/linux/skbuff.h | 4
datapath/linux/compat/ip_tunnels_core.c | 4 ++--
4 files changed, 12 insertions(+),
> Should OVS chdir to the root (on all drives?) on Windows
> also?
Looks like chdir("/") on windows takes me to C:/ . So doing that on
windows should be good. I will re-spin this.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/lis
On Tue, Apr 22, 2014 at 9:35 AM, Ben Pfaff wrote:
> On Fri, Apr 18, 2014 at 11:04:14AM -0700, Gurucharan Shetty wrote:
>> The option won't have any effect on the running of the daemon.
>> Recognizing the option lets us avoid if else conditions in unit
>> tests.
>>
>> Signed-off-by: Gurucharan Shet
On Tue, Apr 22, 2014 at 9:20 AM, Ben Pfaff wrote:
> On Fri, Apr 18, 2014 at 11:04:12AM -0700, Gurucharan Shetty wrote:
>> When "--detach" is specified, a daemon will create a new
>> process with the same command line options as the parent.
>> Additionally, an undocumented command line option "--pi
Something like this, right?
diff --git a/tests/atlocal.in b/tests/atlocal.in
index 06e7384..60f3dbf 100644
--- a/tests/atlocal.in
+++ b/tests/atlocal.in
@@ -78,3 +78,13 @@ FreeBSD|NetBSD)
LOOPBACK_INTERFACE=lo0
;;
esac
+
+# Check for MINGW platform.
+case `uname` in
+MINGW*)
+IS_WIN
From: Zoltan Kiss
commit 36d5fe6a000790f56039afe26834265db0a3ad4c upstream.
skb_zerocopy can copy elements of the frags array between skbs, but it doesn't
orphan them. Also, it doesn't handle errors, so this patch takes care of that
as well, and modify the callers accordingly. skb_tx_error() is
Also fix some tests that can now properly print packets
with the new protocol version.
Signed-off-by: Alexandru Copot
Cc: Daniel Baluta
---
v4: moved version print change here
---
lib/ofp-print.c| 3 +++
tests/ofp-print.at | 6 +++---
tests/ofp-util.at | 2 +-
3 files changed, 7 insertions
This is only the communication part of the bundles functionality.
The actual message pre-validation and commits are not implemented.
Signed-off-by: Alexandru Copot
Cc: Daniel Baluta
---
v4:
* adjust copyright
* fix style issues
* remove locking in ofconn
* delete bundles in ofconn_destro
Signed-off-by: Alexandru Copot
Cc: Daniel Baluta
---
v4: rebase
v3: small fix
v2: rebase
---
tests/ofp-print.at | 125 ++
tests/ofproto-macros.at | 4 +-
tests/ofproto.at| 234
3 files changed, 361 insertions
This series adds the basic communication handlers for the bundles
functionality. The actual message pre-validation and commits are
not implemented.
Signed-off-by: Alexandru Copot
Cc: Daniel Baluta
Alexandru Copot (3):
Add basic implementation for OpenFlow 1.4 bundles
tests: add some OpenFlo
I've had a couple of requests for an updated project list, so this commit
adds it to the tree.
Signed-off-by: Ben Pfaff
---
Makefile.am |1 +
TODO| 290 +++
2 files changed, 291 insertions(+)
create mode 100644 TODO
diff --gi
Thx, applied,
On Tue, Apr 22, 2014 at 11:05 AM, Ben Pfaff wrote:
> On Tue, Apr 15, 2014 at 12:23:14PM -0700, Alex Wang wrote:
> > The ovs-rcu module adds a new thread for checking the grace period.
> > Since the thread name is not set, it will inherit the name of the
> > thread that creates it.
On Tue, Apr 15, 2014 at 12:23:14PM -0700, Alex Wang wrote:
> The ovs-rcu module adds a new thread for checking the grace period.
> Since the thread name is not set, it will inherit the name of the
> thread that creates it. This makes the 'top' output quite confusing.
>
> This commit names the thr
On 22/04/14 16:38, Ben Hutchings wrote:
On Mon, 2014-04-21 at 12:26 +0100, Luis Henriques wrote:
Hi David,
On Thu, Mar 27, 2014 at 03:29:56PM -0400, David Miller wrote:
From: Zoltan Kiss
Date: Wed, 26 Mar 2014 22:37:45 +
skb_zerocopy can copy elements of the frags array between skbs, bu
This function returns true when 's' is negative or greater than UINT_MAX. Also,
the representation of 'int' and 'unsigned int' is implementation dependent, so
converting [INT_MAX..UINT_MAX] values with str_to_int is fragile.
Instead, we should convert straight to 'long long' and do a boundary check
On 18/04/14 02:42, Jarno Rajahalme wrote:
On Apr 17, 2014, at 11:26 AM, Zoltan Kiss wrote:
On 16/04/14 18:00, Justin Pettit wrote:
On April 16, 2014 at 9:00:15 AM, Zoltan Kiss (zoltan.k...@citrix.com) wrote:
My actual problem is that an important rule gets deleted:
cookie=0x0, duration=15
On Tue, Apr 22, 2014 at 10:03:00PM +0530, Shivanker Goel wrote:
> But what I've been trying to do for a while now, is interacting with the
> controller with respect to this. The controller should be able to enable
> eviction in a table using an OFP_TABLE_MOD packet, it should be able to
> modify
On Fri, Apr 18, 2014 at 11:04:17AM -0700, Gurucharan Shetty wrote:
> We use kill to cleanup processes from pidfiles.
> Windows has a 'taskkill' which does something similar.
> We can check if the process with a PID exists with
> 'tasklist'. Both tasklist and taskkill return 0 for
> both success and
On Fri, Apr 18, 2014 at 11:04:18AM -0700, Gurucharan Shetty wrote:
> In unit tests, we compare text written in logs or stdout/stderr
> to figure out the success or failure of tests. In Windows,
> since new line is represented by CR+LF, autoconf tests run in
> MinGW environment fail.
>
> Asking dif
On Fri, Apr 18, 2014 at 11:04:16AM -0700, Gurucharan Shetty wrote:
> On MinGW, "pwd -W" gives the present working directory
> in the form of windows path (i.e C:/temp instead of /c/temp).
> When we pass the directory path to daemons as arguments,
> we should be passing it in the form of windows pat
On Fri, Apr 18, 2014 at 11:04:14AM -0700, Gurucharan Shetty wrote:
> The option won't have any effect on the running of the daemon.
> Recognizing the option lets us avoid if else conditions in unit
> tests.
>
> Signed-off-by: Gurucharan Shetty
The reason that Unix daemons chdir to the root, by d
Hello everyone,
I was attempting to incorporate OF1.4 style eviction in OVS as a project.
By OF1.4 style eviction, I mean using the importance parameter of flows in
deciding which flow to evict. This parameter can be set by the controller
per flow.
I have defined the required enums and structs
On Fri, Apr 18, 2014 at 11:04:13AM -0700, Gurucharan Shetty wrote:
> In Windows, we cannot delete a file that has been opened.
> We use this feature to "lock" the pidfile.
>
> Signed-off-by: Gurucharan Shetty
I think that we are starting to accumulate some common code, which it
might be a good i
On Fri, Apr 18, 2014 at 11:04:12AM -0700, Gurucharan Shetty wrote:
> When "--detach" is specified, a daemon will create a new
> process with the same command line options as the parent.
> Additionally, an undocumented command line option "--pipe-handle"
> is passed to child. Once the child is ready
On Fri, Apr 18, 2014 at 11:04:11AM -0700, Gurucharan Shetty wrote:
> Sa far, we are using variable 'detach' to indicate whether the option
> "--service" has been set. We were using variable 'detached' to indicate that
> the daemon is being called from the Windows services manager.
>
> An upcoming
On April 22, 2014 at 7:33:38 AM, Neelakantam Gaddam (neelugad...@gmail.com)
wrote:
>
> I am using openvswitch-1.10.0.
>
> We are running openvswitch on MIPS based platform.
> While sending the traffic with 10K flows, I see lot of lost statistics in
> the dpctl command. I see not more than 1500
On Fri, Apr 18, 2014 at 11:04:10AM -0700, Gurucharan Shetty wrote:
> Now that we have a separate [p]stream_class for hiding
> kernel assigned TCP port inside a file meant for unix domain
> sockets in windows, use the new infrastructure provided.
>
> Signed-off-by: Gurucharan Shetty
I'd consider
On Fri, Apr 18, 2014 at 11:04:09AM -0700, Gurucharan Shetty wrote:
> On Linux, we heavily use --remote=punix:* to listen for
> connections through unix domain sockets. We also use, unix:*
> to connect to a daemon that is listening on unix domain sockets.
> Many times, we create default unix domain
On Fri, Apr 18, 2014 at 11:04:08AM -0700, Gurucharan Shetty wrote:
> Signed-off-by: Gurucharan Shetty
> ---
> tests/atlocal.in |3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/tests/atlocal.in b/tests/atlocal.in
> index 06e7384..8267554 100644
> --- a/tests/atlocal.in
> +++ b/tests
On Mon, 2014-04-21 at 12:26 +0100, Luis Henriques wrote:
> Hi David,
>
> On Thu, Mar 27, 2014 at 03:29:56PM -0400, David Miller wrote:
> > From: Zoltan Kiss
> > Date: Wed, 26 Mar 2014 22:37:45 +
> >
> > > skb_zerocopy can copy elements of the frags array between skbs, but it
> > > doesn't
>
On Tue, Apr 22, 2014 at 01:47:32PM +0900, YAMAMOTO Takashi wrote:
> Improve random distribution for an hmap with a small number of nodes
> with the expense of the increased cpu cost.
> It would be a fair trade-off because the situation is rather common
> for bond, which is currently the only consum
Hi All,
I am using openvswitch-1.10.0.
We are running openvswitch on MIPS based platform.
While sending the traffic with 10K flows, I see lot of lost statistics in
the dpctl command. I see not more than 1500 flows in the datapath.
>From my understanding, lost count is the number of packets desti
Hi Justin,
I knew that you wouldn’t accept the patch (same hooking place, same impact on
ovs-vswitchd), but I posted this patch anyway to get feedback on AXM concept
that has been demonstrated at the ONF forum (http://youtu.be/jkbkvX2B_kI): how
to use and present L4-L7 information at the OpenFl
With recirculation in place this should be safe.
XXX: What if recirculation is disabled???
Signed-off-by: Simon Horman
---
lib/ofp-actions.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/lib/ofp-actions.c b/lib/ofp-actions.c
index ce14004..69d6505 100644
--- a/lib/ofp-actions.c
+++ b
In some cases an pop MPLS action changes a packet to be a non-mpls packet.
In this case subsequent any L3+ actions require access to portions
of the packet which were not decoded as they were opaque when the
packet was MPLS. Allow such actions to be translated by
first recirculating the packet.
Si
This is in preparation for using the same helper as part of support
for using recirculation in conjunction series of actions including
with MPLS actions that are currently not able to be translated.
In that scenario this helper will be used to test if load, move and
set_field actions require recir
This is in preparation for using the same helper as part of support
for using recirculation in conjunction series of actions including
with MPLS actions that are currently not able to be translated.
In that scenario the idle timeout will be used to expire internal
rules that are added to handle re
This is to allow a recirculation id to be associated with a rule
in the case that its actions cause recirculation.
In such a case if the recirc_id field is non-zero then that value should be
used, otherwise a value should be obtained using
ofproto_dpif_alloc_recirc_id and saved in recirc_id field.
This will be used by a subsequent patch to add support for
recirculation for MPLS.
Signed-off-by: Simon Horman
---
ofproto/ofproto-dpif.c | 10 +-
ofproto/ofproto-dpif.h | 2 ++
2 files changed, 11 insertions(+), 1 deletion(-)
diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.
This makes maintaining the controller test slightly easier
by splitting it in two.
Based on a similar patch by Joe Stringer.
Cc: Joe Stringer
Signed-off-by: Simon Horman
--
v3
* Rebase
v2
* No change
v1
* Taken from another patch-set and re-worked
---
tests/ofproto-dpif.at | 353 +++
This tests exercises execution of actions in ovs-vswitchd
in the case where a packet is processed due to a packet out message
with an in_port that doesn't exist in the datapath and translation
results in recirc actions due to bonding.
Signed-off-by: Simon Horman
---
tests/ofproto-dpif.at | 80 ++
The motivation of this series is to allow some sequences of actions
that include MPLS actions to be performed using recirculation.
Sequences of actions that could not previously be handled.
For example pop_mpls:0x0800,dec_ttl.
This uses recirculation after the pop_mpls action to allow access to
This is in preparation for using this value
in ofproto-dpif-xlate.c when composing recirculation
actions added as a result of processing (MPLS) actions.
Signed-off-by: Simon Horman
---
ofproto/bond.c | 1 -
ofproto/ofproto-dpif.h | 3 +++
2 files changed, 3 insertions(+), 1 deletion(-)
There does not seem to be a reason to expose
recirc members in struct xlate_out to move them
to struct xlate_ctx.
Signed-off-by: Simon Horman
---
ofproto/ofproto-dpif-xlate.c | 20
ofproto/ofproto-dpif-xlate.h | 3 ---
2 files changed, 12 insertions(+), 11 deletions(-)
dif
In the case where packet is being processed by
ofproto_dpif_execute_actions*() it may have an in_port
that does not exist in the datapath: e.g. a packet from a packet_out message
with the CONTROLLER as the in_port.
In such cases it seems to make little sense to perform execution
in the datapath if
If execution is performed using a helper then any recirc actions
encountered may be deferred. This change executes such actions.
Signed-off-by: Simon Horman
---
ofproto/ofproto-dpif.c | 81 +-
1 file changed, 73 insertions(+), 8 deletions(-)
diff
Rather then prohibiting recirc actions dpif_execute_helper_cb()
deffer their processing by saving the required state.
This is in preparation for allowing execution of recirc actions
in ovs-vswitchd.
Signed-off-by: Simon Horman
---
lib/dpif.c | 37 +++--
lib/dpif.
Rather then prohibiting hash actions dpif_execute_helper_cb()
execute them by passing them on to aux->dpif->dpif_class->execute()
and save the updated packet metadata afterwards. It is assumed
that if a hash is calculated it will be sored in the dp_hash field
of the metadata.
This is in preparatio
Add has_recirc field to struct xlate_out. It is set to
true during actions translation if the resulting actions include
a recirc action. It is false otherwise.
This will be used by a subsequent patch that executes
recirc actions for packet out in ovs-vswitchd.
Signed-off-by: Simon Horman
---
of
Add per-table counters. This resolves some short-comings
in the data provided in a table stats reply message.
* Lookups and matches are calculated based on table
accesses rather than datapath flow hits and misses.
* Lookups and matches are credited to the table where they
occurred rather than
73 matches
Mail list logo
