Looks Good.
Ethan
On Wed, Jun 8, 2011 at 14:04, Ben Pfaff wrote:
> The kernel now provides the entire flow key for a packet sent up to
> userspace, but dpif_recv() would only log the in_port. This change makes
> userspace log the entire flow key.
>
> This would have made a bug that I recently l
Looks Good.
Ethan
On Wed, Jun 8, 2011 at 13:41, Ben Pfaff wrote:
> It's reasonable to want to dump a flow table with "dump-flows" and then
> later use that output as input to other commands that read flows, but until
> now that hasn't been directly possible because "dump-flows" adds extra
> fiel
Looks Good to me.
I wonder if it make sense to change table_id to table in the internal
data structures as well so things are consistent?
Ethan
On Wed, Jun 8, 2011 at 13:41, Ben Pfaff wrote:
> Flow dumps printed the OpenFlow table ID under the name "table_id", but
> the flow parser only accepte
---
tests/ovs-ofctl.at |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/tests/ovs-ofctl.at b/tests/ovs-ofctl.at
index a7d3b84..8b15f9c 100644
--- a/tests/ovs-ofctl.at
+++ b/tests/ovs-ofctl.at
@@ -121,6 +121,7 @@ actions=note:41.42.43,note:00.01.02.03.04.05.06.07,note
tun_
Looks Good.
Ethan
On Wed, Jun 8, 2011 at 13:41, Ben Pfaff wrote:
> Until now, the flow parser has supported dl_vlan and dl_vlan_pcp but not
> the fully maskable vlan_tci that NXM allows. This adds that support.
> ---
> lib/ofp-parse.c | 36
> tes
---
lib/vlog.c | 13 -
1 files changed, 12 insertions(+), 1 deletions(-)
diff --git a/lib/vlog.c b/lib/vlog.c
index 4f6523f..265e9c6 100644
--- a/lib/vlog.c
+++ b/lib/vlog.c
@@ -29,6 +29,7 @@
#include "dirs.h"
#include "dynamic-string.h"
#include "sat-math.h"
+#include "svec.h"
PMC生产计划与物料控制高级研修班
时间地点：2011年6月25-26日深圳
2011年7月14-15日苏州
2011年7月23-24日深圳
2011年7月28-29日北京
主办单位：智源培训(www.zhiyuanpx.com)
学员对象：生产总监、生产经理、生产计划相关人员、车间主任、物料主管、采
购主管及相关人员.
费 用：2800/人（含授课费、资料费、会务费、午餐费）
联系电话：
上海：021 -- 5109 3280 周小姐钱先生
深圳
On Wed, Jun 08, 2011 at 02:04:11PM -0700, Jesse Gross wrote:
> On Wed, Jun 8, 2011 at 1:47 PM, Ben Pfaff wrote:
> > On Wed, Jun 08, 2011 at 01:05:45PM -0700, Jesse Gross wrote:
> >> On Mon, Jun 6, 2011 at 12:41 PM, Ben Pfaff wrote:
> >> > ovs-vsctl is carefully written to avoid races in database
On Wed, Jun 8, 2011 at 1:51 PM, Ben Pfaff wrote:
> On Wed, Jun 08, 2011 at 01:21:41PM -0700, Jesse Gross wrote:
>> If we can't parse a header because it is invalid or not present due to
>> fragmentation, we still need to include the length of that header when
>> comparing the flow key. The value
On Wed, Jun 8, 2011 at 1:47 PM, Ben Pfaff wrote:
> On Wed, Jun 08, 2011 at 01:05:45PM -0700, Jesse Gross wrote:
>> On Mon, Jun 6, 2011 at 12:41 PM, Ben Pfaff wrote:
>> > ovs-vsctl is carefully written to avoid races in database access. ??It is
>> > much simpler to just call it than to reimplement
The kernel now provides the entire flow key for a packet sent up to
userspace, but dpif_recv() would only log the in_port. This change makes
userspace log the entire flow key.
This would have made a bug that I recently looked at a bit easier to
investigate.
---
lib/dpif.c | 42
On Wed, Jun 08, 2011 at 11:19:36AM -0700, Jesse Gross wrote:
> On Wed, Jun 8, 2011 at 10:51 AM, Ben Pfaff wrote:
> > I accidentally discovered that on "master" every IP fragment ends up
> > being sent to userspace. ??My test case is "netperf -H 192.168.0.2 -t
> > UDP_STREAM" (which generates 64 kB
On Wed, Jun 08, 2011 at 01:21:42PM -0700, Jesse Gross wrote:
> The fields of the kernel flow key are now grouped by protocol rather
> than using generic names. The containing structures describe the
> category, so it is no longer necessary to use prefixes. Most of
> these prefixes have been remov
On Wed, Jun 08, 2011 at 01:21:41PM -0700, Jesse Gross wrote:
> If we can't parse a header because it is invalid or not present due to
> fragmentation, we still need to include the length of that header when
> comparing the flow key. The value of the field will be zero to
> indicate that header was
On Wed, Jun 08, 2011 at 01:05:45PM -0700, Jesse Gross wrote:
> On Mon, Jun 6, 2011 at 12:41 PM, Ben Pfaff wrote:
> > ovs-vsctl is carefully written to avoid races in database access. ??It is
> > much simpler to just call it than to reimplement its capabilities.
> >
> > This eliminates the requirem
Flow dumps printed the OpenFlow table ID under the name "table_id", but
the flow parser only accepted "table". This makes them consistent by
changing the output. (Another alternative would be to change the accepted
input name.)
[CCing people and groups who I know parse ovs-ofctl output.]
CC: Pau
Until now, the flow parser has supported dl_vlan and dl_vlan_pcp but not
the fully maskable vlan_tci that NXM allows. This adds that support.
---
lib/ofp-parse.c | 36
tests/ovs-ofctl.at |2 ++
utilities/ovs-ofctl.8.in | 37 +
It's reasonable to want to dump a flow table with "dump-flows" and then
later use that output as input to other commands that read flows, but until
now that hasn't been directly possible because "dump-flows" adds extra
fields that other commands reject on input. This changes the flow parser
to jus
The fields of the kernel flow key are now grouped by protocol rather
than using generic names. The containing structures describe the
category, so it is no longer necessary to use prefixes. Most of
these prefixes have been removed but nw_proto and nw_tos have
retained them. This renames the fiel
If we can't parse a header because it is invalid or not present due to
fragmentation, we still need to include the length of that header when
comparing the flow key. The value of the field will be zero to
indicate that header was not present, rather than effectively
wildcarding the value. However
On Mon, Jun 6, 2011 at 12:41 PM, Ben Pfaff wrote:
> ovs-vsctl is carefully written to avoid races in database access. It is
> much simpler to just call it than to reimplement its capabilities.
>
> This eliminates the requirement that bridges managed by ovs-brcompatd have
> no ports at ovs-brcompa
On Wed, Jun 8, 2011 at 10:51 AM, Ben Pfaff wrote:
> I accidentally discovered that on "master" every IP fragment ends up
> being sent to userspace. My test case is "netperf -H 192.168.0.2 -t
> UDP_STREAM" (which generates 64 kB UDP packets), then observing that
> the datapath "miss" counter goes
I accidentally discovered that on "master" every IP fragment ends up
being sent to userspace. My test case is "netperf -H 192.168.0.2 -t
UDP_STREAM" (which generates 64 kB UDP packets), then observing that
the datapath "miss" counter goes up quickly while the "hit" counter
stays put, as well as no
Thanks Ben.
:: psi
On Jun 7, 2011, at 4:54 PM, Ben Pfaff wrote:
> Thank you for the reviews. I pushed this series.
>
> On Tue, Jun 07, 2011 at 04:36:13PM -0700, Ethan Jackson wrote:
>> Looks Good.
>>
>> Ethan
>>
>> On Mon, Jun 6, 2011 at 14:28, Ben Pfaff wrote:
>>> This makes it possible to
On Tue, Jun 07, 2011 at 03:47:35PM -0700, Sajjad Lateef wrote:
> xenserver logs can be in any timezone, as configured by user
> during install time. Also, the timezone data is not currently
> recorded in the bugtool report. Therefore, it is currently not
> possible to easily correlate logs collecte
On Tue, Jun 07, 2011 at 06:22:03PM -0700, Ethan Jackson wrote:
> The patch seems fine. I had one thought which may or may not be valuable.
>
> In reading send_ifindex_reply() and get_bridge_intefaces() I thought
> that it might be valuable to implement a function which converts a "
> \t\r\n" deli
26 matches
Mail list logo
