From patchwork Thu Sep 17 12:50:53 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Dumitru Ceara X-Patchwork-Id: 1366103 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=openvswitch.org (client-ip=140.211.166.133; helo=hemlock.osuosl.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Authentication-Results: ozlabs.org; dkim=fail reason="signature verification failed" (1024-bit key; unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256 header.s=mimecast20190719 header.b=KEBNCEmc; dkim-atps=neutral Received: from hemlock.osuosl.org (smtp2.osuosl.org [140.211.166.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 4BscKN0vhxz9sSC for ; Thu, 17 Sep 2020 22:51:19 +1000 (AEST) Received: from localhost (localhost [127.0.0.1]) by hemlock.osuosl.org (Postfix) with ESMTP id 20FAA87604; Thu, 17 Sep 2020 12:51:17 +0000 (UTC) X-Virus-Scanned: amavisd-new at osuosl.org Received: from hemlock.osuosl.org ([127.0.0.1]) by localhost (.osuosl.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Z8LTr9-hkIaP; Thu, 17 Sep 2020 12:51:15 +0000 (UTC) Received: from lists.linuxfoundation.org (lf-lists.osuosl.org [140.211.9.56]) by hemlock.osuosl.org (Postfix) with ESMTP id BA12A875D5; Thu, 17 Sep 2020 12:51:15 +0000 (UTC) Received: from lf-lists.osuosl.org (localhost [127.0.0.1]) by lists.linuxfoundation.org (Postfix) with ESMTP id 96F15C0890; Thu, 17 Sep 2020 12:51:15 +0000 (UTC) X-Original-To: dev@openvswitch.org Delivered-To: ovs-dev@lists.linuxfoundation.org Received: from whitealder.osuosl.org (smtp1.osuosl.org [140.211.166.138]) by lists.linuxfoundation.org (Postfix) with ESMTP id 187F1C0051 for ; Thu, 17 Sep 2020 12:51:15 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by whitealder.osuosl.org (Postfix) with ESMTP id 084F8878A4 for ; Thu, 17 Sep 2020 12:51:15 +0000 (UTC) X-Virus-Scanned: amavisd-new at osuosl.org Received: from whitealder.osuosl.org ([127.0.0.1]) by localhost (.osuosl.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id x+yBIA61Tjpg for ; Thu, 17 Sep 2020 12:51:13 +0000 (UTC) X-Greylist: domain auto-whitelisted by SQLgrey-1.7.6 Received: from us-smtp-delivery-1.mimecast.com (us-smtp-1.mimecast.com [205.139.110.61]) by whitealder.osuosl.org (Postfix) with ESMTPS id 9264A87869 for ; Thu, 17 Sep 2020 12:51:13 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1600347072; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=wU8ZZ13TPTULQpdT0PkNBySqUgAera/Lipra2rVrp3s=; b=KEBNCEmc/X8uv8a6t2pAZ++KFsUzpoc5Bw9swry7S7zSwd2yn/ns1t5xLhihtMKmhkgJ/r UKRUZnpZWJK+OnoB2IXOQG/3obUL0ymP83yu89d12ssa7kH+uRvdMUdqgZKAvkC+v++7og VR6UnzFFScralppxhgMrsJ9Xq1DX9gQ= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-366-y47T71aZPn655mBz-QA9Rw-1; Thu, 17 Sep 2020 08:50:57 -0400 X-MC-Unique: y47T71aZPn655mBz-QA9Rw-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id ED54E800400 for ; Thu, 17 Sep 2020 12:50:56 +0000 (UTC) Received: from dceara.remote.csb (ovpn-113-244.ams2.redhat.com [10.36.113.244]) by smtp.corp.redhat.com (Postfix) with ESMTP id 52A1C2C31E for ; Thu, 17 Sep 2020 12:50:56 +0000 (UTC) From: Dumitru Ceara To: dev@openvswitch.org Date: Thu, 17 Sep 2020 14:50:53 +0200 Message-Id: <20200917125050.19729.41529.stgit@dceara.remote.csb> In-Reply-To: <20200917125025.19729.19409.stgit@dceara.remote.csb> References: <20200917125025.19729.19409.stgit@dceara.remote.csb> User-Agent: StGit/0.17.1-dirty MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 Authentication-Results: relay.mimecast.com; auth=pass smtp.auth=CUSA124A263 smtp.mailfrom=dceara@redhat.com X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com Subject: [ovs-dev] [PATCH v3 ovn 1/4] ovn-northd: Drop IP packets destined to router owned IPs (after NAT). X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: ovs-dev-bounces@openvswitch.org Sender: "dev" OVN was dropping IP packets destined to IPs owned by logical routers but only if those IPs are not used for SNAT rules. However, if a packet doesn't match an existing NAT session and its destination is still a router owned IP, it can be safely dropped. Otherwise it will trigger an unnecessary packet-in in stage lr_in_arp_request. To achieve that we add flows that drop traffic to router owned SNAT IPs in table lr_in_arp_resolve. Reported-by: Tim Rozet Reported-at: https://bugzilla.redhat.com/1876174 Signed-off-by: Dumitru Ceara Acked-by: Han Zhou --- northd/ovn-northd.8.xml | 24 ++++++ northd/ovn-northd.c | 194 +++++++++++++++++++++++++++-------------------- tests/ovn.at | 88 +++++++++++++++++++++ 3 files changed, 225 insertions(+), 81 deletions(-) diff --git a/northd/ovn-northd.8.xml b/northd/ovn-northd.8.xml index bd42105..f1c7c9b 100644 --- a/northd/ovn-northd.8.xml +++ b/northd/ovn-northd.8.xml @@ -3089,6 +3089,30 @@ outport = P;

  • + Traffic with IP destination an address owned by the router should be + dropped. Such traffic is normally dropped in ingress table + IP Input except for IPs that are also shared with SNAT + rules. However, if there was no unSNAT operation that happened + successfully until this point in the pipeline and the destination IP + of the packet is still a router owned IP, the packets can be safely + dropped. +

    + +

    + A priority-1 logical flow with match ip4.dst = {..} + matches on traffic destined to router owned IPv4 addresses which are + also SNAT IPs. This flow has action drop;. +

    + +

    + A priority-1 logical flow with match ip6.dst = {..} + matches on traffic destined to router owned IPv6 addresses which are + also SNAT IPs. This flow has action drop;. +

    +
    • + +
  • +

    Dynamic MAC bindings. These flows resolve MAC-to-IP bindings that have become known dynamically through ARP or neighbor discovery. (The ingress table ARP Request will diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c index cfec6a2..d5d7631 100644 --- a/northd/ovn-northd.c +++ b/northd/ovn-northd.c @@ -623,6 +623,9 @@ struct ovn_datapath { /* NAT entries configured on the router. */ struct ovn_nat *nat_entries; + /* SNAT IPs used by the router. */ + struct sset snat_ips; + struct ovn_port **localnet_ports; size_t n_localnet_ports; @@ -641,6 +644,10 @@ struct ovn_nat { struct lport_addresses ext_addrs; }; +static bool +get_force_snat_ip(struct ovn_datapath *od, const char *key_type, + struct lport_addresses *laddrs); + /* Returns true if a 'nat_entry' is valid, i.e.: * - parsing was successful. * - the string yielded exactly one IPv4 address or exactly one IPv6 address. @@ -663,7 +670,35 @@ nat_entry_is_v6(const struct ovn_nat *nat_entry) static void init_nat_entries(struct ovn_datapath *od) { - if (!od->nbr || od->nbr->n_nat == 0) { + struct lport_addresses snat_addrs; + + if (!od->nbr) { + return; + } + + sset_init(&od->snat_ips); + if (get_force_snat_ip(od, "dnat", &snat_addrs)) { + if (snat_addrs.n_ipv4_addrs) { + sset_add(&od->snat_ips, snat_addrs.ipv4_addrs[0].addr_s); + } + if (snat_addrs.n_ipv6_addrs) { + sset_add(&od->snat_ips, snat_addrs.ipv6_addrs[0].addr_s); + } + destroy_lport_addresses(&snat_addrs); + } + + memset(&snat_addrs, 0, sizeof(snat_addrs)); + if (get_force_snat_ip(od, "lb", &snat_addrs)) { + if (snat_addrs.n_ipv4_addrs) { + sset_add(&od->snat_ips, snat_addrs.ipv4_addrs[0].addr_s); + } + if (snat_addrs.n_ipv6_addrs) { + sset_add(&od->snat_ips, snat_addrs.ipv6_addrs[0].addr_s); + } + destroy_lport_addresses(&snat_addrs); + } + + if (!od->nbr->n_nat) { return; } @@ -682,6 +717,13 @@ init_nat_entries(struct ovn_datapath *od) VLOG_WARN_RL(&rl, "Bad ip address %s in nat configuration " "for router %s", nat->external_ip, od->nbr->name); + continue; + } + + if (!nat_entry_is_v6(nat_entry)) { + sset_add(&od->snat_ips, nat_entry->ext_addrs.ipv4_addrs[0].addr_s); + } else { + sset_add(&od->snat_ips, nat_entry->ext_addrs.ipv6_addrs[0].addr_s); } } } @@ -693,6 +735,7 @@ destroy_nat_entries(struct ovn_datapath *od) return; } + sset_destroy(&od->snat_ips); for (size_t i = 0; i < od->nbr->n_nat; i++) { destroy_lport_addresses(&od->nat_entries[i].ext_addrs); } @@ -8744,6 +8787,68 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports, op, lflows, &match, &actions); } + /* Drop IP traffic destined to router owned IPs. Part of it is dropped + * in stage "lr_in_ip_input" but traffic that could have been unSNATed + * but didn't match any existing session might still end up here. + */ + HMAP_FOR_EACH (op, key_node, ports) { + if (!op->nbrp) { + continue; + } + + if (op->lrp_networks.n_ipv4_addrs) { + ds_clear(&match); + for (size_t i = 0; i < op->lrp_networks.n_ipv4_addrs; i++) { + if (!sset_find(&op->od->snat_ips, + op->lrp_networks.ipv4_addrs[i].addr_s)) { + continue; + } + ds_put_format(&match, "%s, ", + op->lrp_networks.ipv4_addrs[i].addr_s); + } + + if (ds_last(&match) != EOF) { + ds_chomp(&match, ' '); + ds_chomp(&match, ','); + + char *drop_match = xasprintf("ip4.dst == {%s}", + ds_cstr(&match)); + /* Drop traffic with IP.dest == router-ip. */ + ovn_lflow_add_with_hint(lflows, op->od, + S_ROUTER_IN_ARP_RESOLVE, 1, + drop_match, "drop;", + &op->nbrp->header_); + free(drop_match); + } + } + + if (op->lrp_networks.n_ipv6_addrs) { + ds_clear(&match); + for (size_t i = 0; i < op->lrp_networks.n_ipv6_addrs; i++) { + if (!sset_find(&op->od->snat_ips, + op->lrp_networks.ipv6_addrs[i].addr_s)) { + continue; + } + ds_put_format(&match, "%s, ", + op->lrp_networks.ipv6_addrs[i].addr_s); + } + + if (ds_last(&match) != EOF) { + ds_chomp(&match, ' '); + ds_chomp(&match, ','); + + char *drop_match = xasprintf("ip6.dst == {%s}", + ds_cstr(&match)); + /* Drop traffic with IP.dest == router-ip. */ + ovn_lflow_add_with_hint(lflows, op->od, + S_ROUTER_IN_ARP_RESOLVE, 1, + drop_match, "drop;", + &op->nbrp->header_); + free(drop_match); + } + } + } + HMAP_FOR_EACH (od, key_node, datapaths) { if (!od->nbr) { continue; @@ -9035,77 +9140,15 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports, } /* A gateway router can have 4 SNAT IP addresses to force DNATed and - * LBed traffic respectively to be SNATed. In addition, there can be - * a number of SNAT rules in the NAT table. */ - struct v46_ip *snat_ips = xmalloc(sizeof *snat_ips - * (op->od->nbr->n_nat + 4)); - size_t n_snat_ips = 0; - struct lport_addresses snat_addrs; - - if (get_force_snat_ip(op->od, "dnat", &snat_addrs)) { - if (snat_addrs.n_ipv4_addrs) { - snat_ips[n_snat_ips].family = AF_INET; - snat_ips[n_snat_ips++].ipv4 = snat_addrs.ipv4_addrs[0].addr; - } - if (snat_addrs.n_ipv6_addrs) { - snat_ips[n_snat_ips].family = AF_INET6; - snat_ips[n_snat_ips++].ipv6 = snat_addrs.ipv6_addrs[0].addr; - } - destroy_lport_addresses(&snat_addrs); - } - - memset(&snat_addrs, 0, sizeof(snat_addrs)); - if (get_force_snat_ip(op->od, "lb", &snat_addrs)) { - if (snat_addrs.n_ipv4_addrs) { - snat_ips[n_snat_ips].family = AF_INET; - snat_ips[n_snat_ips++].ipv4 = snat_addrs.ipv4_addrs[0].addr; - } - if (snat_addrs.n_ipv6_addrs) { - snat_ips[n_snat_ips].family = AF_INET6; - snat_ips[n_snat_ips++].ipv6 = snat_addrs.ipv6_addrs[0].addr; - } - destroy_lport_addresses(&snat_addrs); - } - - for (size_t i = 0; i < op->od->nbr->n_nat; i++) { - struct ovn_nat *nat_entry = &op->od->nat_entries[i]; - const struct nbrec_nat *nat = nat_entry->nb; - - /* Skip entries we failed to parse. */ - if (!nat_entry_is_valid(nat_entry)) { - continue; - } - - if (!strcmp(nat->type, "snat")) { - if (nat_entry_is_v6(nat_entry)) { - struct in6_addr *ipv6 = - &nat_entry->ext_addrs.ipv6_addrs[0].addr; - - snat_ips[n_snat_ips].family = AF_INET6; - snat_ips[n_snat_ips++].ipv6 = *ipv6; - } else { - ovs_be32 ip = nat_entry->ext_addrs.ipv4_addrs[0].addr; - snat_ips[n_snat_ips].family = AF_INET; - snat_ips[n_snat_ips++].ipv4 = ip; - } - } - } - + * LBed traffic respectively to be SNATed. In addition, there can be + * a number of SNAT rules in the NAT table. + * Skip all of them for drop flows. */ ds_clear(&match); ds_put_cstr(&match, "ip4.dst == {"); bool has_drop_ips = false; for (int i = 0; i < op->lrp_networks.n_ipv4_addrs; i++) { - bool snat_ip_is_router_ip = false; - for (int j = 0; j < n_snat_ips; j++) { - /* Packets to SNAT IPs should not be dropped. */ - if (snat_ips[j].family == AF_INET - && op->lrp_networks.ipv4_addrs[i].addr - == snat_ips[j].ipv4) { - snat_ip_is_router_ip = true; - break; - } - } - if (snat_ip_is_router_ip) { + if (sset_find(&op->od->snat_ips, + op->lrp_networks.ipv4_addrs[i].addr_s)) { continue; } ds_put_format(&match, "%s, ", @@ -9122,17 +9165,8 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports, } for (int i = 0; i < op->lrp_networks.n_ipv6_addrs; i++) { - bool snat_ip_is_router_ip = false; - for (int j = 0; j < n_snat_ips; j++) { - /* Packets to SNAT IPs should not be dropped. */ - if (snat_ips[j].family == AF_INET6 - && !memcmp(&op->lrp_networks.ipv6_addrs[i].addr, - &snat_ips[j].ipv6, sizeof snat_ips[j].ipv6)) { - snat_ip_is_router_ip = true; - break; - } - } - if (snat_ip_is_router_ip) { + if (sset_find(&op->od->snat_ips, + op->lrp_networks.ipv6_addrs[i].addr_s)) { continue; } ds_put_format(&match, "%s, ", @@ -9151,8 +9185,6 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports, &op->nbrp->header_); } - free(snat_ips); - /* ARP/NS packets are taken care of per router. The only exception * is on the l3dgw_port where we might need to use a different * ETH address. diff --git a/tests/ovn.at b/tests/ovn.at index a6f1fb5..cb7e7cc 100644 --- a/tests/ovn.at +++ b/tests/ovn.at @@ -21659,6 +21659,94 @@ OVS_WAIT_UNTIL([test x$(as hv1 ovn-appctl -t ovn-controller debug/status) = "xru OVN_CLEANUP([hv1]) AT_CLEANUP +# Test dropping traffic destined to router owned IPs. +AT_SETUP([ovn -- gateway router drop traffic for own IPs]) +ovn_start + +ovn-nbctl lr-add r1 -- set logical_router r1 options:chassis=hv1 +ovn-nbctl ls-add s1 + +# Connnect r1 to s1. +ovn-nbctl lrp-add r1 lrp-r1-s1 00:00:00:00:01:01 10.0.1.1/24 +ovn-nbctl lsp-add s1 lsp-s1-r1 -- set Logical_Switch_Port lsp-s1-r1 type=router \ + options:router-port=lrp-r1-s1 addresses=router + +# Create logical port p1 in s1 +ovn-nbctl lsp-add s1 p1 \ +-- lsp-set-addresses p1 "f0:00:00:00:01:02 10.0.1.2" + +# Create two hypervisor and create OVS ports corresponding to logical ports. +net_add n1 + +sim_add hv1 +as hv1 +ovs-vsctl add-br br-phys +ovn_attach n1 br-phys 192.168.0.1 +ovs-vsctl -- add-port br-int hv1-vif1 -- \ + set interface hv1-vif1 external-ids:iface-id=p1 \ + options:tx_pcap=hv1/vif1-tx.pcap \ + options:rxq_pcap=hv1/vif1-rx.pcap \ + ofport-request=1 + +# Pre-populate the hypervisors' ARP tables so that we don't lose any +# packets for ARP resolution (native tunneling doesn't queue packets +# for ARP resolution). +OVN_POPULATE_ARP + +ovn-nbctl --wait=hv sync + +sw_key=$(ovn-sbctl --bare --columns tunnel_key list datapath_binding r1) + +AT_CHECK([ovn-sbctl lflow-list | grep lr_in_arp_resolve | grep 10.0.1.1], [1], []) + +ip_to_hex() { + printf "%02x%02x%02x%02x" "$@" +} + +# Send ip packets from p1 to lrp-r1-s1 +src_mac="f00000000102" +dst_mac="000000000101" +src_ip=`ip_to_hex 10 0 1 2` +dst_ip=`ip_to_hex 10 0 1 1` +packet=${dst_mac}${src_mac}08004500001c0000000040110000${src_ip}${dst_ip}0035111100080000 +as hv1 ovs-appctl netdev-dummy/receive hv1-vif1 $packet + +# No packet-ins should reach ovn-controller. +AT_CHECK([as hv1 ovs-ofctl dump-flows br-int | grep "actions=controller" | grep -v n_packets=0 -c], [1], [dnl +0 +]) + +# The packet should have been dropped in the lr_in_ip_input stage. +AT_CHECK([as hv1 ovs-ofctl dump-flows br-int | grep -E "table=11, n_packets=1,.* priority=60,ip,metadata=0x${sw_key},nw_dst=10.0.1.1 actions=drop" -c], [0], [dnl +1 +]) + +# Use the router IP as SNAT IP. +ovn-nbctl set logical_router r1 options:lb_force_snat_ip=10.0.1.1 +ovn-nbctl --wait=hv sync + +# Send ip packets from p1 to lrp-r1-s1 +src_mac="f00000000102" +dst_mac="000000000101" +src_ip=`ip_to_hex 10 0 1 2` +dst_ip=`ip_to_hex 10 0 1 1` +packet=${dst_mac}${src_mac}08004500001c0000000040110000${src_ip}${dst_ip}0035111100080000 +as hv1 ovs-appctl netdev-dummy/receive hv1-vif1 $packet + +# Even after configuring a router owned IP for SNAT, no packet-ins should +# reach ovn-controller. +AT_CHECK([as hv1 ovs-ofctl dump-flows br-int | grep "actions=controller" | grep -v n_packets=0 -c], [1], [dnl +0 +]) + +# The packet should've been dropped in the lr_in_arp_resolve stage. +AT_CHECK([as hv1 ovs-ofctl dump-flows br-int | grep -E "table=21, n_packets=1,.* priority=1,ip,metadata=0x${sw_key},nw_dst=10.0.1.1 actions=drop" -c], [0], [dnl +1 +]) + +OVN_CLEANUP([hv1]) +AT_CLEANUP + AT_SETUP([ovn -- nb_cfg timestamp]) ovn_start