[sfc-dev] Service Chain Last Hop

Ricardo Noriega De Soto ricardo.noriega.de.soto at ericsson.com
Wed Aug 26 15:46:10 UTC 2015


Hello guys,

Something similar is happening to me:

    ubuntu at sff:~$ sudo ovs-ofctl -O OpenFlow13 dump-flows ovs-br2

    OFPST_FLOW reply (OF1.3) (xid=0x2):

    cookie=0x14, duration=3880.253s, table=0, n_packets=42,
    n_bytes=2982, priority=250,ip actions=goto_table:3

    cookie=0x14, duration=3880.253s, table=0, n_packets=0, n_bytes=0,
    priority=5 actions=drop

    cookie=0x14, duration=3880.253s, table=1, n_packets=0, n_bytes=0,
    priority=5 actions=goto_table:2

    cookie=0x14, duration=3880.253s, table=2, n_packets=0, n_bytes=0,
    priority=5 actions=goto_table:3

    cookie=0x14, duration=3880.253s, table=3, n_packets=20,
    n_bytes=1417, priority=5 actions=goto_table:10 →PACKETS COMING FROM
    THE SF

    cookie=0x14, duration=3879.873s, table=3, n_packets=22,
    n_bytes=1565, priority=540,nsp=1,nsi=255
    actions=load:0xc0a8010c->NXM_NX_TUN_IPV4_DST[],goto_table:10
    →PACKETS GOING TO THE SF

    cookie=0x14, duration=3880.253s, table=10, n_packets=0, n_bytes=0,
    priority=5 actions=drop

      cookie=0xba5eba11ba5eba11, duration=3879.725s, table=10,
    n_packets=22, n_bytes=1565, priority=640,nsp=1,nsi=255
    actions=move:NXM_NX_NSH_C1[]->NXM_NX_NSH_C1[],move:NXM_NX_NSH_C2[]->NXM_NX_NSH_C2[],move:NXM_NX_TUN_ID[0..31]->NXM_NX_TUN_ID[0..31],IN_PORT

    * cookie=0xba5eba11ba5eba11, duration=3879.683s, table=10,
    n_packets=0, n_bytes=0, priority=640,nsp=1,nsi=254
    actions=move:NXM_NX_NSI[]->NXM_NX_NSI[],move:NXM_NX_NSP[]->NXM_NX_NSP[],move:NXM_NX_NSH_C1[]->NXM_NX_TUN_IPV4_DST[],move:NXM_NX_NSH_C2[]->NXM_NX_TUN_ID[0..31],move:NXM_NX_TUN_ID[0..31]->NXM_NX_TUN_ID[0..31],IN_PORT*

    **

    *cookie=0x14, duration=3879.666s, table=10, n_packets=20,
    n_bytes=1417, priority=660,nsp=1,nsi=254,nshc1=0 actions=IN_PORT*

I am struggling now to understand which rule in table=10 the packet 
coming from the SF should match, because there are two rules for 
table=10 and nsi=254 (last 2 rules); and when the field nshc1=0 (in the 
latest rule) should be used. Do you have any idea about this??


@Reinaldo: How can I configure the SFF to decapsulate the traffic as 
last hop, and send it to the original destination IP? Could you send me 
any example or anything? THANKS!

BR
Ricky


On 08/26/2015 11:50 AM, João Silva wrote:
> Hi all!
>
> It seems that none of the options are in place at the moment.
>
> With the current sfc implementation in ODL, using ovs as the SFF, when 
> the packet from the last sf is received in the SFF it is dropped.
> Here are the flows(in this example I have 2 service functions):
>
> table 0:
>
> cookie=0x14, duration=45.006s, table=0, n_packets=0, n_bytes=0, 
> priority=250,ip actions=goto_table:3
>
>   * Matches all VxLAN packets
>
> table 3:
>
>  cookie=0x14, duration=45.222s, table=3, n_packets=0, n_bytes=0, 
> priority=540,nsp=3,nsi=255 
> actions=load:0xc0a83204->NXM_NX_TUN_IPV4_DST[],goto_table:10
>
>   * Matches nsi=255, the packet coming from the classifier to be sent
>     to sf1
>
>  cookie=0x14, duration=44.597s, table=3, n_packets=0, n_bytes=0, 
> priority=540,nsp=3,nsi=254 
> actions=load:0xc0a83205->NXM_NX_TUN_IPV4_DST[],goto_table:10
>
>   * Matches nsi=254, the packet coming from the sf1 to be sent to sf2
>
> Shouldn't there be a rule for nsi=253? Because since there isn't once 
> the packet with nsi=253 reaches table 3 matches nowhere and is 
> dropped. Is this intended behavior? Please keep in mind that I'm using 
> an OVS as SFF and not the SFF agent.
>
>
> Best regards,
>
> João
>
>
>
> _______________________________________________
> sfc-dev mailing list
> sfc-dev at lists.opendaylight.org
> https://lists.opendaylight.org/mailman/listinfo/sfc-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opendaylight.org/pipermail/sfc-dev/attachments/20150826/4e558e67/attachment.html>


More information about the sfc-dev mailing list