Re: [ovs-discuss] VXLAN offload not work on openvswitch 2.6.0 + 3.10.45

2016-10-24 Thread
vxlan receive offload feature not work on these two ovs versions desired for this kernel? 2016-10-25 10:39 GMT+08:00 张东亚 : > Hi List, > > We have now testing openvswitch 2.6.0 on kernel 3.10.45, however we > noticed there is performance drop than openvswitch 2.5.0 on the same > kern

[ovs-discuss] VXLAN offload not work on openvswitch 2.6.0 + 3.10.45

2016-10-24 Thread
Hi List, We have now testing openvswitch 2.6.0 on kernel 3.10.45, however we noticed there is performance drop than openvswitch 2.5.0 on the same kernel, which iperf can only have a 2G bandwidth for a pair of 10G links. After check the code, it seems on openvswitch 2.5.0, HAVE_METADATA_DST macro

Re: [ovs-discuss] Roadmap for DPDK usespace NAT

2016-09-19 Thread
Hi Darrel Ball, Glad to hear that, thank a lot for provide the information. 2016-09-20 7:41 GMT+08:00 Darrell Ball : > I am working on DPDK NAT support and targeting 2.7. > > > > Darrell > > > > > > > > *From: *discuss on behalf of 张东亚 < > fortitude

Re: [ovs-discuss] Roadmap for DPDK usespace NAT

2016-09-18 Thread
Hi Justin, Thanks a lot for your prompt reply, we will try the ct feature first. 2016-09-19 13:56 GMT+08:00 Justin Pettit : > > > On Sep 18, 2016, at 7:21 PM, 张东亚 wrote: > > > > Hi Ben, > > > > So that means the only thing we can know is NAT for DPDK is not &

Re: [ovs-discuss] Roadmap for DPDK usespace NAT

2016-09-18 Thread
7PM +0800, 张东亚 wrote: > > From ovs 2.6 branch NEWS file, it have stated the following sentence: > > > > Basic connection tracking for the userspace datapath (no ALG, > > fragmentation or NAT support yet) > > > > I am wonder what's the developmenent schedule for

[ovs-discuss] Roadmap for DPDK usespace NAT

2016-09-18 Thread
Hi List, >From ovs 2.6 branch NEWS file, it have stated the following sentence: Basic connection tracking for the userspace datapath (no ALG, fragmentation or NAT support yet) I am wonder what's the developmenent schedule for userspace NAT based on DPDK, will it be released with first OVS 2.6 re

Re: [ovs-discuss] Enable sFlow on br-tun cause OVS 2.5.0 crashed

2016-08-29 Thread
rts, just their representation in the sFlow > agent. I believe the intention was to make sure that all internal > agent state was reset cleanly on a configuration change. For > example, the sequence number for packet and counter samples is > required to reset to 0. > > Rega

Re: [ovs-discuss] Enable sFlow on br-tun cause OVS 2.5.0 crashed

2016-08-29 Thread
Hi Neil, There is another question regarding the issue, seems dpif_sflow_add_port will call dpif_sflow_del_port port every time, does this is intended to remove previously added dsp since dp_port might be same for different tunnel port? 2016-08-29 11:16 GMT+08:00 张东亚 : > Hi Neil, > &

Re: [ovs-discuss] Enable sFlow on br-tun cause OVS 2.5.0 crashed

2016-08-28 Thread
pe) : 0; > > Neil > > > -- > Neil McKee > InMon Corp. > http://www.inmon.com > > On Fri, Aug 26, 2016 at 2:30 AM, 张东亚 wrote: > >> Hi List, >> >> Recently we are testing sFlow on OVS 2.5.0, since OVS decide whether do >> sample based on ingres

[ovs-discuss] Enable sFlow on br-tun cause OVS 2.5.0 crashed

2016-08-26 Thread
Hi List, Recently we are testing sFlow on OVS 2.5.0, since OVS decide whether do sample based on ingress bridge, we then enable sFlow on br-tun, and encounter following crash: #0 0x00434ca8 in dpif_sflow_received (ds=0x4266100, packet=packet@entry=0x7f3cb3fc8e18, flow=flow@entry=0x7f3cb3

Re: [ovs-discuss] skb_warn_bad_offload+0xc8/0xd3() kernel warning on linux v3.14.57+ovs 2.5.0

2016-07-13 Thread
Hi Jesse, Thanks a lot for your information, seems the commit have been merged to v4.6+ kernel which may not be easily for us to upgrade. We will try to upgrade to lastest OVS and do test again. 2016-07-13 22:48 GMT+08:00 Jesse Gross : > On Wed, Jul 13, 2016 at 4:05 AM, 张东亚 wrote: >

[ovs-discuss] skb_warn_bad_offload+0xc8/0xd3() kernel warning on linux v3.14.57+ovs 2.5.0

2016-07-13 Thread
Hi, Recently we upgraded to ovs 2.5 and use the datapath in the ovs code(means we do not use kernel datapath), however, when we test vxlan performance between two VMs on different compute node, we observed the bandwidth can only achieve 5Gbps on the 10Gbps intel nic pair, and we are still using ke

Re: [ovs-discuss] Slow ovsdb-server performance issue with lots of tunnel ports on switch with embedded SoC.

2014-09-24 Thread
Hi Ben, Thanks a lot for your advice, yes, it's fast if we add 1000 tunnels in one transaction, however, our user create/delete tunnels reactivelly according to the traffic between the two VTEPS. I also want to precreate 1000 tunnels and let user modify the tunnels as they needed, since the

[ovs-discuss] Slow ovsdb-server performance issue with lots of tunnel ports on switch with embedded SoC.

2014-09-24 Thread
Hi all, We have used ovsdb-server/openvswitch for a long time on our switch which use an freescale SoC p1010 with cpu core frequency set to 533Mhz, recently we want to add 1000 GRE tunnel ports support to our switch, however, we have observed significant slow down when we trying to configure more

[ovs-discuss] Segment fault when running ovs-ofctl snoop br0 using ovs master

2013-01-25 Thread
Hi, I have built an ovs (master-bfe834e) and found and segment fault error when I do the following operation: 1. start ovs-vswitchd and ovsdb-server 2. set and controller that does not exist, ovs-vsctl set-controller br0 "tcp:127.0.0.1"(which will lead connect refused error when ovs trying t