Hi, Thanks for the extra information, that was useful!
I found a problem. Inside the classifier1, packets hit the classification rule: table=0, n_packets=49, n_bytes=3626, priority=1000,tcp,in_port=1,nw_src=192.168. 2.0/24,nw_dst=192.168.2.0/24,tp_dst=80 actions=push_nsh,load:0x1->NXM_NX_NSH_MDTY PE[],load:0x3->NXM_NX_NSH_NP[],load:0x7f->NXM_NX_NSP[0..23],load:0xff->NXM_NX_NSI [],load:0x1->NXM_NX_NSH_C1[],load:0x2->NXM_NX_NSH_C2[],load:0x3->NXM_NX_NSH_C3[], load:0x4->NXM_NX_NSH_C4[],load:0x4->NXM_NX_TUN_GPE_NP[],load:0xc0a83c14->NXM_NX_T UN_IPV4_DST[],output:1 However, output:1 is 1(veth-br) which is not a vxlan-gpe port and thus the packet is not encapsulated with vxlan-gpe + nsh. Therefore, the packets which leave the OVS have the original IP header: 17:36:21.879104 IP 192.168.2.1.43460 > 192.168.2.2.http: Flags [S], seq 1480997927, win 27200, options [mss 1360,sackOK,TS val 413304 ecr 0,nop,wscale 7], length 0 Eth0 does not know where is 192.168.2.2 and its ARP request does not get answer. I changed that rule to have output:2 which is the vxlan-gpe port. Packets flow but they don't come back from SF1. I think it might be a problem with the vxlan tool because I don't see logs coming in /home/vagrant/vxlan_tool.log even though I can see the packets with tcpdump. I will try to continue tomorrow as I don't have more time today :(!! Regards, Manuel -----Original Message----- From: Yang, Yi Y [mailto:[email protected]] Sent: Saturday, November 12, 2016 3:47 AM To: Manuel Buil <[email protected]>; [email protected] Subject: RE: Welcome you to try sfc104 demo Manuel, good progress, "sudo" is unnecessary. This indicated sfc didn't create correct flows to SFF1 and SFF2 You can run the below cmds to check every vagrant VMs. $ source ./env.sh $ vagrant ssh classifier1 $ vagrant ssh sff1 ... In classifier1, sff1, sff2 and classifier2 vagrant VMs, you can run the below cmds to check ovs ports, flows. $ sudo ovs-vsctl show $ sudo ovs-ofctl -Oopenflow13 dump-flows br-sfc In sf1 and sf2 VMs, you can run the below cmds to check if SF is running correctly $ sudo ps aux | grep vxlan_tool $ cat /home/vagrant/vxlan_tool.log In order to check it further, you can run ping and wget in classifier1 and check where the packet can arrive. $ vagrant ssh classifier1 $ sudo ip netns exec app ping 192.168.2.2 How do you start your ODL SFC? Another possible reason you run this demo before SFC features aren't completely started although "feature:list -i | grep sfc" can see them. I suggest you can make sure ODL SFC is really ready before run this demo. I'll add such check in run_demo.sh but now it isn't there. -----Original Message----- From: Manuel Buil [mailto:[email protected]] Sent: Saturday, November 12, 2016 1:40 AM To: Yang, Yi Y <[email protected]>; [email protected] Subject: RE: Welcome you to try sfc104 demo Hi Yi yang, I am having some problems with your demo. I execute " sudo ./run_demo.sh ovs ". The VMs seems to be created: $ sudo VBoxManage list runningvms "sfc104_classifier1_1478856929727_28276" {e41e8e3e-0049-458b-a5f0-e427bdfd6646} "sfc104_sff1_1478856985216_33607" {4c181ea9-0c6f-4a59-a7ef-7a836dc25d79} "sfc104_sf1_1478857042513_36221" {f05e13ca-901b-4cac-9446-8da42b2ea5df} "sfc104_sf2_1478857100191_91163" {c9232419-9fac-4229-b6f2-7dde8afff04e} "sfc104_sff2_1478857157560_59263" {1bf0ac70-500a-4afa-bc04-84883c098a85} "sfc104_classifier2_1478857215119_78365" {8affee62-c7e3-483d-bdee-8ac7b4464753} However, at some point of the script it checks the connectivity and it does not work: PING 192.168.2.2 (192.168.2.2) 56(84) bytes of data. 64 bytes from 192.168.2.2: icmp_seq=1 ttl=64 time=11.8 ms 64 bytes from 192.168.2.2: icmp_seq=2 ttl=64 time=8.71 ms 64 bytes from 192.168.2.2: icmp_seq=3 ttl=64 time=11.1 ms 64 bytes from 192.168.2.2: icmp_seq=4 ttl=64 time=9.34 ms 64 bytes from 192.168.2.2: icmp_seq=5 ttl=64 time=9.60 ms --- 192.168.2.2 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4008ms rtt min/avg/max/mdev = 8.710/10.138/11.853/1.186 ms Connection to 127.0.0.1 closed. --2016-11-11 16:58:01-- http://192.168.2.2/ Connecting to 192.168.2.2:80... failed: Connection timed out. Retrying. --2016-11-11 17:00:09-- (try: 2) http://192.168.2.2/ Connecting to 192.168.2.2:80... failed: Connection timed out. Retrying. --2016-11-11 17:02:19-- (try: 3) http://192.168.2.2/ Connecting to 192.168.2.2:80... failed: Connection timed out. Retrying. --2016-11-11 17:04:29-- (try: 4) http://192.168.2.2/ Connecting to 192.168.2.2:80... failed: Connection timed out. Retrying. --2016-11-11 17:06:40-- (try: 5) http://192.168.2.2/ Connecting to 192.168.2.2:80... Do you know what might be happening? Thanks, Manuel -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Yang, Yi Y Sent: Friday, November 11, 2016 11:47 AM To: [email protected] Subject: [sfc-dev] Welcome you to try sfc104 demo Hi, folks Brady has merged sfc104 demo into sfc master, so each of you can try sfc104 demo, please let me know if you encounter any issue. $ cat sfc-demo/sfc104/README.md SFC104 Demo =========== Overview -------- SFC103 demo is docker container based, so you have no way to run it if you will use OVS DPDK or VPP, so we have to create SFC104 demo, SFC104 demo actually includes three demos, they are for OVS, OVS DPDK and VPP, respectively. Your host machine needs 32GB memory at least because DPDK or VPP uses hugepages which must be allocated in advance, 6 vagrant VMs will occupy 24 GB memory. Too less memory will result in vagant hang or some other weird behaviors. Topology ------- +-----------------+ | Host (ODL SFC) | | 192.168.60.1 | +-----------------+ / | | \ / | | \ / | | \ +---------------+ +--------------+ +--------------+ +---------------+ | classifier1 | | sff1 | | sff2 | | classifier2 | | 192.168.60.10 | |192.168.60.20 | |192.168.60.50 | | 192.168.60.60 | +---------------+ +--------------+ +--------------+ +---------------+ | | | | +---------------+ +--------------+ | sf1(DPI-1) | | sf2(FW-1) | |192.168.60.30 | |192.168.60.40 | +---------------+ +--------------+ Setup Demo ---------- 1. Install virtualbox & vagrant 2. Start ODL SFC in host machine and install necessary features 1) For ovs or ovs_dpdk feature:install odl-sfc-scf-openflow odl-sfc-openflow-renderer odl-sfc-ui 2) For vpp feature:install odl-sfc-vpp-renderer odl-sfc-ui Notice: please do stop, clean up, then restart ODL SFC when you run this demo in order that demo can run successfully. opendaylight-user@root>shutdown -f opendaylight-user@root> $ rm -rf data snapshots journal instances $ ./bin/karaf 3. Run demo SFC 104 demos will download Ubuntu trusty x86_64 vagrant image and install all the necessary packages into host and vagrant VMs, so please make sure to export http_proxy and http_proxy environment variables if you have proxy behind your network before run demo script, demo script will inject these proxy settings to vagrant VMS. 1) For OVS demo $ ./run_demo.sh ovs 2) For OVS DPDK demo $ ./run_demo.sh ovs_dpdk 3) For VPP demo $ ./run_demo.sh vpp _______________________________________________ sfc-dev mailing list [email protected] https://lists.opendaylight.org/mailman/listinfo/sfc-dev _______________________________________________ sfc-dev mailing list [email protected] https://lists.opendaylight.org/mailman/listinfo/sfc-dev
