[ovs-dev] [patch v2] ovn: Fix gateway load balancing.

Han Zhou zhouhan at gmail.com
Thu Jun 28 18:44:07 UTC 2018


Hi Mark, what I meant is the test for the feature of LB in Gateway. If we
had a test case, the problem would have been noticed when Lorenzo is
working on ICMP feature.

On Thu, Jun 28, 2018 at 11:29 AM, Mark Michelson <mmichels at redhat.com>
wrote:

> Lorenzo actually brought up in today's OVN IRC meeting that writing ICMP
> tests for IPv6 is problematic right now. IPv4 tests have an M4 macro called
> OVN_POPULATE_ARP that they call. This manually populates tables in OVS so
> that there are no ARP requests being sent during the test. It does this by
> calling `ovs-appct tnl/neigh/set` for each IP address-MAC pair.
>
> IPv6 does not have an equivalent. The result is that tests will sometimes
> succeed, but sometimes will fail because IPv6 ND requests will be sent
> during the test. I haven't looked into this myself, but Lorenzo mentioned
> that there would need to be changes made to OVS itself in order to have an
> IPv6 version of a working OVN_POPULATE_ARP.
>
> On 06/28/2018 02:16 PM, Han Zhou wrote:
>
>> Agree with Mark that ICMP is still needed for non VIP on gateway. But LB
>> feature on GW is also important.
>> Shall we have a test case to cover this particular scenario?
>>
>> On Thu, Jun 28, 2018 at 9:41 AM, Mark Michelson <mmichels at redhat.com
>> <mailto:mmichels at redhat.com>> wrote:
>>
>>     Hi Darrell,
>>
>>     I'm not 100% sure this is correct. I think this change would only be
>>     correct if the router's IP addresses are also load balancer VIPs. In
>>     the case where the VIPs do not overlap with the router IP addresses,
>>     I think we'd want to have the ICMP responses still in place.
>>
>>
>>     On 06/28/2018 01:15 AM, Darrell Ball wrote:
>>
>>         Non-distributed and distributed gateway load balancing is broken.
>>         Recent changes for port unreachable handling broke the associated
>>         unsnat functionality.  The fix approach is check for gateway
>>         contexts and accept packets directed to gateway router IPs.
>>
>>         Fixes:  86558ac2e476 ("OVN: add UDP port unreachable support to
>>         OVN logical router.")
>>         Fixes:  159932c9e4ea ("OVN: add TCP port unreachable support to
>>         OVN logical router.")
>>         Fixes:  0e858e05f76b ("OVN: add protocol unreachable support to
>>         OVN router ports.")
>>         CC: Lorenzo Bianconi <lorenzo.bianconi at redhat.com
>>         <mailto:lorenzo.bianconi at redhat.com>>
>>         Signed-off-by: Darrell Ball <dlu998 at gmail.com
>>         <mailto:dlu998 at gmail.com>>
>>
>>         ---
>>            ovn/northd/ovn-northd.8.xml |  17 ++++---
>>            ovn/northd/ovn-northd.c     | 106
>>         ++++++++++++++++++++++----------------------
>>            2 files changed, 64 insertions(+), 59 deletions(-)
>>
>>         diff --git a/ovn/northd/ovn-northd.8.xml
>>         b/ovn/northd/ovn-northd.8.xml
>>         index cfd3511..280efd0 100644
>>         --- a/ovn/northd/ovn-northd.8.xml
>>         +++ b/ovn/northd/ovn-northd.8.xml
>>         @@ -1310,8 +1310,9 @@ nd_na {
>>                    <p>
>>                      UDP port unreachable.  Priority-80 flows generate
>>         ICMP port
>>                      unreachable messages in reply to UDP datagrams
>>         directed to the
>>         -          router's IP address.  The logical router doesn't
>>         accept any UDP
>>         -          traffic so it always generates such a reply.
>>         +          router's IP address, except in the special case of
>>         gateways,
>>         +          which accept traffic directed to a router IP for load
>>         balancing
>>         +          purposes.
>>                    </p>
>>                      <p>
>>         @@ -1321,10 +1322,10 @@ nd_na {
>>                    <li>
>>                    <p>
>>         -          TCP reset.  Priority-80 flows generate TCP reset
>>         messages in reply to
>>         -          TCP datagrams directed to the router's IP address.
>>      The logical
>>         -          router doesn't accept any TCP traffic so it always
>>         generates such a
>>         -          reply.
>>         +          TCP reset.  Priority-80 flows generate TCP reset
>>         messages in reply
>>         +          to TCP datagrams directed to the router's IP address,
>>         except in
>>         +          the special case of gateways, which accept traffic
>>         directed to a
>>         +          router IP for load balancing purposes.
>>                    </p>
>>                      <p>
>>         @@ -1336,7 +1337,9 @@ nd_na {
>>                    <p>
>>                      Protocol unreachable.  Priority-70 flows generate
>>         ICMP protocol
>>                      unreachable messages in reply to packets directed
>>         to the router's IP
>>         -          address on IP protocols other than UDP, TCP, and ICMP.
>>         +          address on IP protocols other than UDP, TCP, and
>>         ICMP, except in the
>>         +          special case of gateways, which accept traffic
>>         directed to a router
>>         +          IP for load balancing purposes.
>>                    </p>
>>                      <p>
>>         diff --git a/ovn/northd/ovn-northd.c b/ovn/northd/ovn-northd.c
>>         index 72fe4e7..7648bce 100644
>>         --- a/ovn/northd/ovn-northd.c
>>         +++ b/ovn/northd/ovn-northd.c
>>         @@ -5141,48 +5141,49 @@ build_lrouter_flows(struct hmap
>>         *datapaths, struct hmap *ports,
>>                                      ds_cstr(&match), ds_cstr(&actions));
>>                    }
>>            -        /* UDP/TCP port unreachable */
>>         -        for (int i = 0; i < op->lrp_networks.n_ipv4_addrs; i++) {
>>         -            const char *action;
>>         -
>>         -            ds_clear(&match);
>>         -            ds_put_format(&match,
>>         -                          "ip4 && ip4.dst == %s &&
>>         !ip.later_frag && udp",
>>         -                          op->lrp_networks.ipv4_addrs[i]
>> .addr_s);
>>         -            action = "icmp4 {"
>>         -                        "eth.dst <-> eth.src; "
>>         -                        "ip4.dst <-> ip4.src; "
>>         -                        "ip.ttl = 255; "
>>         -                        "icmp4.type = 3; "
>>         -                        "icmp4.code = 3; "
>>         -                        "next; };";
>>         -            ovn_lflow_add(lflows, op->od, S_ROUTER_IN_IP_INPUT,
>> 80,
>>         -                          ds_cstr(&match), action);
>>         +        if (!smap_get(&op->od->nbr->options, "chassis")
>>         +            && !op->od->l3dgw_port) {
>>         +            /* UDP/TCP port unreachable. */
>>         +            for (int i = 0; i < op->lrp_networks.n_ipv4_addrs;
>>         i++) {
>>         +                ds_clear(&match);
>>         +                ds_put_format(&match,
>>         +                              "ip4 && ip4.dst == %s &&
>>         !ip.later_frag && udp",
>>         +                                     op->lrp_networks.ipv4_addrs[
>> i].addr_s);
>>         +                const char *action = "icmp4 {"
>>         +                                     "eth.dst <-> eth.src; "
>>         +                                     "ip4.dst <-> ip4.src; "
>>         +                                     "ip.ttl = 255; "
>>         +                                     "icmp4.type = 3; "
>>         +                                     "icmp4.code = 3; "
>>         +                                     "next; };";
>>         +                ovn_lflow_add(lflows, op->od,
>>         S_ROUTER_IN_IP_INPUT, 80,
>>         +                              ds_cstr(&match), action);
>>            -            ds_clear(&match);
>>         -            ds_put_format(&match,
>>         -                          "ip4 && ip4.dst == %s &&
>>         !ip.later_frag && tcp",
>>         -                          op->lrp_networks.ipv4_addrs[i]
>> .addr_s);
>>         -            action = "tcp_reset {"
>>         -                        "eth.dst <-> eth.src; "
>>         -                        "ip4.dst <-> ip4.src; "
>>         -                        "next; };";
>>         -            ovn_lflow_add(lflows, op->od, S_ROUTER_IN_IP_INPUT,
>> 80,
>>         -                          ds_cstr(&match), action);
>>         +                ds_clear(&match);
>>         +                ds_put_format(&match,
>>         +                              "ip4 && ip4.dst == %s &&
>>         !ip.later_frag && tcp",
>>         +                                     op->lrp_networks.ipv4_addrs[
>> i].addr_s);
>>         +                action = "tcp_reset {"
>>         +                         "eth.dst <-> eth.src; "
>>         +                         "ip4.dst <-> ip4.src; "
>>         +                         "next; };";
>>         +                ovn_lflow_add(lflows, op->od,
>>         S_ROUTER_IN_IP_INPUT, 80,
>>         +                              ds_cstr(&match), action);
>>            -            ds_clear(&match);
>>         -            ds_put_format(&match,
>>         -                          "ip4 && ip4.dst == %s &&
>> !ip.later_frag",
>>         -                          op->lrp_networks.ipv4_addrs[i]
>> .addr_s);
>>         -            action = "icmp4 {"
>>         -                        "eth.dst <-> eth.src; "
>>         -                        "ip4.dst <-> ip4.src; "
>>         -                        "ip.ttl = 255; "
>>         -                        "icmp4.type = 3; "
>>         -                        "icmp4.code = 2; "
>>         -                        "next; };";
>>         -            ovn_lflow_add(lflows, op->od, S_ROUTER_IN_IP_INPUT,
>> 70,
>>         -                          ds_cstr(&match), action);
>>         +                ds_clear(&match);
>>         +                ds_put_format(&match,
>>         +                              "ip4 && ip4.dst == %s &&
>>         !ip.later_frag",
>>         +                                     op->lrp_networks.ipv4_addrs[
>> i].addr_s);
>>         +                action = "icmp4 {"
>>         +                         "eth.dst <-> eth.src; "
>>         +                         "ip4.dst <-> ip4.src; "
>>         +                         "ip.ttl = 255; "
>>         +                         "icmp4.type = 3; "
>>         +                         "icmp4.code = 2; "
>>         +                         "next; };";
>>         +                ovn_lflow_add(lflows, op->od,
>>         S_ROUTER_IN_IP_INPUT, 70,
>>         +                              ds_cstr(&match), action);
>>         +            }
>>                    }
>>                      ds_clear(&match);
>>         @@ -5306,19 +5307,20 @@ build_lrouter_flows(struct hmap
>>         *datapaths, struct hmap *ports,
>>                    }
>>                      /* TCP port unreachable */
>>         -        for (int i = 0; i < op->lrp_networks.n_ipv6_addrs; i++) {
>>         -            const char *action;
>>         -
>>         -            ds_clear(&match);
>>         -            ds_put_format(&match,
>>         -                          "ip6 && ip6.dst == %s &&
>>         !ip.later_frag && tcp",
>>         -                          op->lrp_networks.ipv6_addrs[i]
>> .addr_s);
>>         -            action = "tcp_reset {"
>>         -                        "eth.dst <-> eth.src; "
>>         -                        "ip6.dst <-> ip6.src; "
>>         -                        "next; };";
>>         -            ovn_lflow_add(lflows, op->od, S_ROUTER_IN_IP_INPUT,
>> 80,
>>         +        if (!smap_get(&op->od->nbr->options, "chassis")
>>         +            && !op->od->l3dgw_port) {
>>         +            for (int i = 0; i < op->lrp_networks.n_ipv6_addrs;
>>         i++) {
>>         +                ds_clear(&match);
>>         +                ds_put_format(&match,
>>         +                              "ip6 && ip6.dst == %s &&
>>         !ip.later_frag && tcp",
>>         +                                     op->lrp_networks.ipv6_addrs[
>> i].addr_s);
>>         +                const char *action = "tcp_reset {"
>>         +                                     "eth.dst <-> eth.src; "
>>         +                                     "ip6.dst <-> ip6.src; "
>>         +                                     "next; };";
>>         +                ovn_lflow_add(lflows, op->od,
>>         S_ROUTER_IN_IP_INPUT, 80,
>>                                      ds_cstr(&match), action);
>>         +            }
>>                    }
>>                }
>>
>>
>>     _______________________________________________
>>     dev mailing list
>>     dev at openvswitch.org <mailto:dev at openvswitch.org>
>>     https://mail.openvswitch.org/mailman/listinfo/ovs-dev
>>     <https://mail.openvswitch.org/mailman/listinfo/ovs-dev>
>>
>>
>>
>


More information about the dev mailing list