As the number of flows grows so does the length of time taken to update the
statistics of all facets. It has been obvserved that this becomes a performance
problem. This patch mitiages this effect by only dumping a limited number of
facets at a time.
Signed-off-by: Simon Horman
---
Since the la
On Wed, Dec 21, 2011 at 12:10:45PM -0800, Ben Pfaff wrote:
> On Wed, Dec 21, 2011 at 05:54:18PM +0900, Simon Horman wrote:
> > On Mon, Dec 12, 2011 at 09:46:44AM -0800, Ben Pfaff wrote:
> > > On Sun, Dec 11, 2011 at 11:20:50AM +0900, Simon Horman wrote:
> > > > sorry for the long delay. I finally h
On Tue, Jan 24, 2012 at 07:17:12PM -0800, Ethan Jackson wrote:
> Here's an incremental.
The incremental looks good.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Tue, Jan 24, 2012 at 06:11:02PM -0800, Ethan Jackson wrote:
> There are no known implementations of the "min-rate" QoS algorithm
> documented on ovs-vswitchd.conf.db. This patch removes its
> documentation pending an implementation.
>
> Signed-off-by: Ethan Jackson
Thanks, this looks good.
_
Here's an incremental.
---
lib/stp.c | 22 --
vswitchd/ovs-vswitchd.8.in |6 --
2 files changed, 20 insertions(+), 8 deletions(-)
diff --git a/lib/stp.c b/lib/stp.c
index 67a93d0..ba27306 100644
--- a/lib/stp.c
+++ b/lib/stp.c
@@ -210,7 +210,8 @@ sta
Signed-off-by: Ethan Jackson
---
Here's another version.
---
lib/stp.c | 16
lib/stp.h |1 +
ofproto/ofproto-dpif.c |4
3 files changed, 21 insertions(+), 0 deletions(-)
diff --git a/lib/stp.c b/lib/stp.c
index 3e9a5b6..f7fe53c 100644
---
There are no known implementations of the "min-rate" QoS algorithm
documented on ovs-vswitchd.conf.db. This patch removes its
documentation pending an implementation.
Signed-off-by: Ethan Jackson
---
vswitchd/vswitch.xml | 13 -
1 files changed, 0 insertions(+), 13 deletions(-)
d
On Mon, Jan 23, 2012 at 01:38:04PM -0800, Ethan Jackson wrote:
> > +/* Removes 'old' from its oftable then, if 'new' is nonnull, inserts
> > 'new'. */
> > +static void
> > +oftable_substitute_rule(struct rule *old, struct rule *new)
> > +{
> > + oftable_remove_rule(old);
> > + if (new) {
> >
> You can write strlen(args) more cheaply as args[0] != '\0'.
Good point.
> I assume that you will also put this in one fashion or another on
> branch-1.5?
branch-1.5 cherry picks cleanly.
Ethan
___
dev mailing list
dev@openvswitch.org
http://openvswi
On Tue, Jan 24, 2012 at 03:04:32PM -0800, Ethan Jackson wrote:
> Signed-off-by: Ethan Jackson
Seems fine.
You can write strlen(args) more cheaply as args[0] != '\0'.
I assume that you will also put this in one fashion or another on
branch-1.5?
Thanks,
Ben.
On Tue, Jan 24, 2012 at 02:03:47PM -0800, Ethan Jackson wrote:
> Looks good,
>
> In heap_raw_insert() why not use x2nrealloc() instead of xrealloc()?
Because we need heap->allocated + 1 elements instead of just
heap->allocated elements.
> 1) You've ifdef'd out some code. Any reason not to simpl
Signed-off-by: Ethan Jackson
---
ofproto/ofproto-dpif.c | 25 -
vswitchd/ovs-vswitchd.8.in |5 +++--
2 files changed, 19 insertions(+), 11 deletions(-)
diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
index d5ada4c..1e387ba 100644
--- a/ofproto/ofpro
> Good point, here's a fixed version:
New version looks good to me.
Ethan
>
> /* Parses a subfield from the beginning of '*sp' into 'sf'. If successful,
> * returns NULL and advances '*sp' to the first byte following the parsed
> * string. On failure, returns a malloc()'d error message, does
On Tue, Jan 24, 2012 at 02:41:29PM -0800, Ethan Jackson wrote:
> Signed-off-by: Ethan Jackson
Looks good to me.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Mon, Jan 23, 2012 at 01:13:03PM -0800, Ethan Jackson wrote:
> In meta-flow.c
>
> In two places you mention ofp_mkerr() which no longer exists.
Fixed. (It existed when I wrote the code.)
> In mf_parse_subfield__() you modify 's' even if there's an error which
> would be fine if the comments d
Signed-off-by: Ethan Jackson
---
ofproto/ofproto-dpif.c | 23 +++
vswitchd/ovs-vswitchd.8.in |5 +++--
2 files changed, 18 insertions(+), 10 deletions(-)
diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
index 471ba64..ea5ed1e 100644
--- a/ofproto/ofproto
Processing commands for cont...@bugs.debian.org:
> # Update Debian BTS.
> tags 656448 + confirmed upstream fixed-upstream
Bug #656448 [openvswitch-pki] openvswitch-pki: unowned files after purge
(policy 6.8, 10.8)
Added tag(s) upstream, confirmed, and fixed-upstream.
> thanks
Stopping processing
# Update Debian BTS.
tags 656448 + confirmed upstream fixed-upstream
thanks
Thanks, I pushed this fix to master, branch-1.2, branch-1.3,
branch-1.4, branch-1.5. It will appear in the next Debian upload.
On Fri, Jan 20, 2012 at 03:32:17PM -0800, Ethan Jackson wrote:
> Looks fine to me, though I'
On Tue, Jan 24, 2012 at 10:50:34AM -0800, Justin Pettit wrote:
> On Jan 24, 2012, at 10:23 AM, Ben Pfaff wrote:
>
> > -* If XAPI is configured for a manger, configures the OVS
> > +* If XAPI is configured for a manager, configures the OVS
>
> Do you doubt the divine origins of XAP
Looks good,
In heap_raw_insert() why not use x2nrealloc() instead of xrealloc()?
I didn't read the testing code closely, but I have two minor questions:
1) You've ifdef'd out some code. Any reason not to simply delete it?
2) You've implemented print_heap() without using it. I suppose that's
fo
NICS-11.
Signed-off-by: Ben Pfaff
---
NEWS |2 ++
lib/learning-switch.c |6 --
lib/mac-learning.c | 41 ++---
lib/mac-learning.h | 10 +++---
ofproto/ofproto-dpif.c | 13 +++--
ofproto
When Cisco (and other?) routers are configured in high-availability modes,
they use two different MAC addresses. The router uses MAC 1 only for ARP
replies. The router uses MAC 2 for forwarding IP packets to end hosts.
When a MAC learning switch is attached to the router, therefore, it will
only
This is the "master" version of the vlan-maint patches posted yesterday.
Much had to be reworked.
I put the NEWS entry under 1.4.x because I intend to backport this to
1.4 and 1.5.
Ben Pfaff (2):
mac-learning: Increase MAC learning timeout to 300 seconds.
vswitchd: Make the MAC entry aging ti
On Jan 24, 2012, at 10:23 AM, Ben Pfaff wrote:
> -* If XAPI is configured for a manger, configures the OVS
> +* If XAPI is configured for a manager, configures the OVS
Do you doubt the divine origins of XAPI?
The series looks good to me.
--Justin
__
I expect that only Nicira internal builds want a build number, so it's
nice to avoid a complaint from RPM if the build_number macro is not
defined.
Signed-off-by: Ben Pfaff
---
xenserver/openvswitch-xen.spec |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/xenserver/open
The instructions in INSTALL.XenServer are more detailed and correct.
Signed-off-by: Ben Pfaff
---
xenserver/README | 11 ---
1 files changed, 0 insertions(+), 11 deletions(-)
diff --git a/xenserver/README b/xenserver/README
index 732a079..c626a92 100644
--- a/xenserver/README
+++ b/xe
---
INSTALL.XenServer |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/INSTALL.XenServer b/INSTALL.XenServer
index 1e3f35a..827a840 100644
--- a/INSTALL.XenServer
+++ b/INSTALL.XenServer
@@ -138,7 +138,7 @@ command. The plugin script does roughly the following:
On Mon, Jan 23, 2012 at 10:04:49PM -0800, Justin Pettit wrote:
> On Jan 23, 2012, at 3:53 PM, Ben Pfaff wrote:
>
> > +*)
> > +keep_files=$scripts
> > +replace_files=
> > +cat < > +
> > +The original XenServer scripts replaced by this package are not those
> > +of any su
US Healthcare Database
- 230,000 emails in 365 categories
Price until Friday: $99
CLICK HERE TO ORDER
File
Available for download in excel format
Just a few of the medical specialties available:
On Tue, Jan 24, 2012 at 08:29:29AM -0800, Gurucharan Shetty wrote:
> The current network-script, ifup-ovs does not work well if you
> enable DHCP on the OVS. It will work if we name the bridge
> alphabetically greater than the underlying physical interfaces.
> Even then, it will do multiple DHCP a
On Tue, Jan 24, 2012 at 12:08 AM, Ghe Rivero wrote:
> Package: openvswitch-datapath-dkms
> Version: 1.3.0-1
> Severity: normal
>
> Dear Maintainer,
> with kernel version 3.2, it fails to compile, openvswitch-datapath-dkms
> and source version too.
Open vSwitch 1.3 doesn't support Linux kernel
The current network-script, ifup-ovs does not work well if you
enable DHCP on the OVS. It will work if we name the bridge
alphabetically greater than the underlying physical interfaces.
Even then, it will do multiple DHCP attempts slowing down the boot
up process.
This patch is my attempt to allo
Package: openvswitch-datapath-dkms
Version: 1.3.0-1
Severity: normal
Dear Maintainer,
with kernel version 3.2, it fails to compile, openvswitch-datapath-dkms and
source version too.
The first error is:
CC [M] /var/lib/dkms/openvswitch/1.3.0/build/datapath/linux/brcompat.o
/var/lib/dkms/open
33 matches
Mail list logo
