tailscale/ipn/ipnlocal/dnsconfig_test.go

410 lines
11 KiB
Go
Raw Normal View History

// Copyright (c) 2021 Tailscale Inc & AUTHORS All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package ipnlocal
import (
"encoding/json"
"reflect"
"testing"
"inet.af/netaddr"
"tailscale.com/ipn"
"tailscale.com/net/dns"
"tailscale.com/tailcfg"
"tailscale.com/tstest"
"tailscale.com/types/dnstype"
"tailscale.com/types/netmap"
ipn/ipnlocal, net/dns*, util/cloudenv: specialize DNS config on Google Cloud This does three things: * If you're on GCP, it adds a *.internal DNS split route to the metadata server, so we never break GCP DNS names. This lets people have some Tailscale nodes on GCP and some not (e.g. laptops at home) without having to add a Tailnet-wide *.internal DNS route. If you already have such a route, though, it won't overwrite it. * If the 100.100.100.100 DNS forwarder has nowhere to forward to, it forwards it to the GCP metadata IP, which forwards to 8.8.8.8. This means there are never errNoUpstreams ("upstream nameservers not set") errors on GCP due to e.g. mangled /etc/resolv.conf (GCP default VMs don't have systemd-resolved, so it's likely a DNS supremacy fight) * makes the DNS fallback mechanism use the GCP metadata IP as a fallback before our hosted HTTP-based fallbacks I created a default GCP VM from their web wizard. It has no systemd-resolved. I then made its /etc/resolv.conf be empty and deleted its GCP hostnames in /etc/hosts. I then logged in to a tailnet with no global DNS settings. With this, tailscaled writes /etc/resolv.conf (direct mode, as no systemd-resolved) and sets it to 100.100.100.100, which then has regular DNS via the metadata IP and *.internal DNS via the metadata IP as well. If the tailnet configures explicit DNS servers, those are used instead, except for *.internal. This also adds a new util/cloudenv package based on version/distro where the cloud type is only detected once. We'll likely expand it in the future for other clouds, doing variants of this change for other popular cloud environments. Fixes #4911 RELNOTES=Google Cloud DNS improvements Change-Id: I19f3c2075983669b2b2c0f29a548da8de373c7cf Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2022-06-29 20:19:34 +00:00
"tailscale.com/util/cloudenv"
"tailscale.com/util/dnsname"
)
func ipps(ippStrs ...string) (ipps []netaddr.IPPrefix) {
for _, s := range ippStrs {
if ip, err := netaddr.ParseIP(s); err == nil {
ipps = append(ipps, netaddr.IPPrefixFrom(ip, ip.BitLen()))
continue
}
ipps = append(ipps, netaddr.MustParseIPPrefix(s))
}
return
}
func ips(ss ...string) (ips []netaddr.IP) {
for _, s := range ss {
ips = append(ips, netaddr.MustParseIP(s))
}
return
}
func TestDNSConfigForNetmap(t *testing.T) {
tests := []struct {
name string
nm *netmap.NetworkMap
os string // version.OS value; empty means linux
ipn/ipnlocal, net/dns*, util/cloudenv: specialize DNS config on Google Cloud This does three things: * If you're on GCP, it adds a *.internal DNS split route to the metadata server, so we never break GCP DNS names. This lets people have some Tailscale nodes on GCP and some not (e.g. laptops at home) without having to add a Tailnet-wide *.internal DNS route. If you already have such a route, though, it won't overwrite it. * If the 100.100.100.100 DNS forwarder has nowhere to forward to, it forwards it to the GCP metadata IP, which forwards to 8.8.8.8. This means there are never errNoUpstreams ("upstream nameservers not set") errors on GCP due to e.g. mangled /etc/resolv.conf (GCP default VMs don't have systemd-resolved, so it's likely a DNS supremacy fight) * makes the DNS fallback mechanism use the GCP metadata IP as a fallback before our hosted HTTP-based fallbacks I created a default GCP VM from their web wizard. It has no systemd-resolved. I then made its /etc/resolv.conf be empty and deleted its GCP hostnames in /etc/hosts. I then logged in to a tailnet with no global DNS settings. With this, tailscaled writes /etc/resolv.conf (direct mode, as no systemd-resolved) and sets it to 100.100.100.100, which then has regular DNS via the metadata IP and *.internal DNS via the metadata IP as well. If the tailnet configures explicit DNS servers, those are used instead, except for *.internal. This also adds a new util/cloudenv package based on version/distro where the cloud type is only detected once. We'll likely expand it in the future for other clouds, doing variants of this change for other popular cloud environments. Fixes #4911 RELNOTES=Google Cloud DNS improvements Change-Id: I19f3c2075983669b2b2c0f29a548da8de373c7cf Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2022-06-29 20:19:34 +00:00
cloud cloudenv.Cloud
prefs *ipn.Prefs
want *dns.Config
wantLog string
}{
{
name: "empty",
nm: &netmap.NetworkMap{},
prefs: &ipn.Prefs{},
want: &dns.Config{
Routes: map[dnsname.FQDN][]*dnstype.Resolver{},
Hosts: map[dnsname.FQDN][]netaddr.IP{},
},
},
{
name: "self_name_and_peers",
nm: &netmap.NetworkMap{
Name: "myname.net",
Addresses: ipps("100.101.101.101"),
Peers: []*tailcfg.Node{
{
Name: "peera.net",
Addresses: ipps("100.102.0.1", "100.102.0.2", "fe75::1001", "fe75::1002"),
},
{
Name: "b.net",
Addresses: ipps("100.102.0.1", "100.102.0.2", "fe75::2"),
},
{
Name: "v6-only.net",
Addresses: ipps("fe75::3"), // no IPv4, so we don't ignore IPv6
},
},
},
prefs: &ipn.Prefs{},
want: &dns.Config{
Routes: map[dnsname.FQDN][]*dnstype.Resolver{},
Hosts: map[dnsname.FQDN][]netaddr.IP{
"b.net.": ips("100.102.0.1", "100.102.0.2"),
"myname.net.": ips("100.101.101.101"),
"peera.net.": ips("100.102.0.1", "100.102.0.2"),
"v6-only.net.": ips("fe75::3"),
},
},
},
{
// An ephemeral node with only an IPv6 address
// should get IPv6 records for all its peers,
// even if they have IPv4.
name: "v6_only_self",
nm: &netmap.NetworkMap{
Name: "myname.net",
Addresses: ipps("fe75::1"),
Peers: []*tailcfg.Node{
{
Name: "peera.net",
Addresses: ipps("100.102.0.1", "100.102.0.2", "fe75::1001"),
},
{
Name: "b.net",
Addresses: ipps("100.102.0.1", "100.102.0.2", "fe75::2"),
},
{
Name: "v6-only.net",
Addresses: ipps("fe75::3"), // no IPv4, so we don't ignore IPv6
},
},
},
prefs: &ipn.Prefs{},
want: &dns.Config{
OnlyIPv6: true,
Routes: map[dnsname.FQDN][]*dnstype.Resolver{},
Hosts: map[dnsname.FQDN][]netaddr.IP{
"b.net.": ips("fe75::2"),
"myname.net.": ips("fe75::1"),
"peera.net.": ips("fe75::1001"),
"v6-only.net.": ips("fe75::3"),
},
},
},
{
name: "extra_records",
nm: &netmap.NetworkMap{
Name: "myname.net",
Addresses: ipps("100.101.101.101"),
DNS: tailcfg.DNSConfig{
ExtraRecords: []tailcfg.DNSRecord{
{Name: "foo.com", Value: "1.2.3.4"},
{Name: "bar.com", Value: "1::6"},
{Name: "sdlfkjsdklfj", Type: "IGNORE"},
},
},
},
prefs: &ipn.Prefs{},
want: &dns.Config{
Routes: map[dnsname.FQDN][]*dnstype.Resolver{},
Hosts: map[dnsname.FQDN][]netaddr.IP{
"myname.net.": ips("100.101.101.101"),
"foo.com.": ips("1.2.3.4"),
"bar.com.": ips("1::6"),
},
},
},
{
name: "corp_dns_misc",
nm: &netmap.NetworkMap{
Name: "host.some.domain.net.",
DNS: tailcfg.DNSConfig{
Proxied: true,
Domains: []string{"foo.com", "bar.com"},
},
},
prefs: &ipn.Prefs{
CorpDNS: true,
},
want: &dns.Config{
Hosts: map[dnsname.FQDN][]netaddr.IP{},
Routes: map[dnsname.FQDN][]*dnstype.Resolver{
"0.e.1.a.c.5.1.1.a.7.d.f.ip6.arpa.": nil,
"100.100.in-addr.arpa.": nil,
"101.100.in-addr.arpa.": nil,
"102.100.in-addr.arpa.": nil,
"103.100.in-addr.arpa.": nil,
"104.100.in-addr.arpa.": nil,
"105.100.in-addr.arpa.": nil,
"106.100.in-addr.arpa.": nil,
"107.100.in-addr.arpa.": nil,
"108.100.in-addr.arpa.": nil,
"109.100.in-addr.arpa.": nil,
"110.100.in-addr.arpa.": nil,
"111.100.in-addr.arpa.": nil,
"112.100.in-addr.arpa.": nil,
"113.100.in-addr.arpa.": nil,
"114.100.in-addr.arpa.": nil,
"115.100.in-addr.arpa.": nil,
"116.100.in-addr.arpa.": nil,
"117.100.in-addr.arpa.": nil,
"118.100.in-addr.arpa.": nil,
"119.100.in-addr.arpa.": nil,
"120.100.in-addr.arpa.": nil,
"121.100.in-addr.arpa.": nil,
"122.100.in-addr.arpa.": nil,
"123.100.in-addr.arpa.": nil,
"124.100.in-addr.arpa.": nil,
"125.100.in-addr.arpa.": nil,
"126.100.in-addr.arpa.": nil,
"127.100.in-addr.arpa.": nil,
"64.100.in-addr.arpa.": nil,
"65.100.in-addr.arpa.": nil,
"66.100.in-addr.arpa.": nil,
"67.100.in-addr.arpa.": nil,
"68.100.in-addr.arpa.": nil,
"69.100.in-addr.arpa.": nil,
"70.100.in-addr.arpa.": nil,
"71.100.in-addr.arpa.": nil,
"72.100.in-addr.arpa.": nil,
"73.100.in-addr.arpa.": nil,
"74.100.in-addr.arpa.": nil,
"75.100.in-addr.arpa.": nil,
"76.100.in-addr.arpa.": nil,
"77.100.in-addr.arpa.": nil,
"78.100.in-addr.arpa.": nil,
"79.100.in-addr.arpa.": nil,
"80.100.in-addr.arpa.": nil,
"81.100.in-addr.arpa.": nil,
"82.100.in-addr.arpa.": nil,
"83.100.in-addr.arpa.": nil,
"84.100.in-addr.arpa.": nil,
"85.100.in-addr.arpa.": nil,
"86.100.in-addr.arpa.": nil,
"87.100.in-addr.arpa.": nil,
"88.100.in-addr.arpa.": nil,
"89.100.in-addr.arpa.": nil,
"90.100.in-addr.arpa.": nil,
"91.100.in-addr.arpa.": nil,
"92.100.in-addr.arpa.": nil,
"93.100.in-addr.arpa.": nil,
"94.100.in-addr.arpa.": nil,
"95.100.in-addr.arpa.": nil,
"96.100.in-addr.arpa.": nil,
"97.100.in-addr.arpa.": nil,
"98.100.in-addr.arpa.": nil,
"99.100.in-addr.arpa.": nil,
"some.domain.net.": nil,
},
SearchDomains: []dnsname.FQDN{
"foo.com.",
"bar.com.",
},
},
},
{
// Prior to fixing https://github.com/tailscale/tailscale/issues/2116,
// Android had cases where it needed FallbackResolvers. This was the
// negative test for the case where Override-local-DNS was set, so the
// fallback resolvers did not need to be used. This test is still valid
// so we keep it, but the fallback test has been removed.
name: "android_does_NOT_need_fallbacks",
os: "android",
nm: &netmap.NetworkMap{
DNS: tailcfg.DNSConfig{
Resolvers: []*dnstype.Resolver{
{Addr: "8.8.8.8"},
},
FallbackResolvers: []*dnstype.Resolver{
{Addr: "8.8.4.4"},
},
Routes: map[string][]*dnstype.Resolver{
"foo.com.": {{Addr: "1.2.3.4"}},
},
},
},
prefs: &ipn.Prefs{
CorpDNS: true,
},
want: &dns.Config{
Hosts: map[dnsname.FQDN][]netaddr.IP{},
DefaultResolvers: []*dnstype.Resolver{
{Addr: "8.8.8.8"},
},
Routes: map[dnsname.FQDN][]*dnstype.Resolver{
"foo.com.": {{Addr: "1.2.3.4"}},
},
},
},
{
name: "exit_nodes_need_fallbacks",
nm: &netmap.NetworkMap{
DNS: tailcfg.DNSConfig{
FallbackResolvers: []*dnstype.Resolver{
{Addr: "8.8.4.4"},
},
},
},
prefs: &ipn.Prefs{
CorpDNS: true,
ExitNodeID: "some-id",
},
want: &dns.Config{
Hosts: map[dnsname.FQDN][]netaddr.IP{},
Routes: map[dnsname.FQDN][]*dnstype.Resolver{},
DefaultResolvers: []*dnstype.Resolver{
{Addr: "8.8.4.4"},
},
},
},
{
name: "not_exit_node_NOT_need_fallbacks",
nm: &netmap.NetworkMap{
DNS: tailcfg.DNSConfig{
FallbackResolvers: []*dnstype.Resolver{
{Addr: "8.8.4.4"},
},
},
},
prefs: &ipn.Prefs{
CorpDNS: true,
},
want: &dns.Config{
Hosts: map[dnsname.FQDN][]netaddr.IP{},
Routes: map[dnsname.FQDN][]*dnstype.Resolver{},
},
},
ipn/ipnlocal, net/dns*, util/cloudenv: specialize DNS config on Google Cloud This does three things: * If you're on GCP, it adds a *.internal DNS split route to the metadata server, so we never break GCP DNS names. This lets people have some Tailscale nodes on GCP and some not (e.g. laptops at home) without having to add a Tailnet-wide *.internal DNS route. If you already have such a route, though, it won't overwrite it. * If the 100.100.100.100 DNS forwarder has nowhere to forward to, it forwards it to the GCP metadata IP, which forwards to 8.8.8.8. This means there are never errNoUpstreams ("upstream nameservers not set") errors on GCP due to e.g. mangled /etc/resolv.conf (GCP default VMs don't have systemd-resolved, so it's likely a DNS supremacy fight) * makes the DNS fallback mechanism use the GCP metadata IP as a fallback before our hosted HTTP-based fallbacks I created a default GCP VM from their web wizard. It has no systemd-resolved. I then made its /etc/resolv.conf be empty and deleted its GCP hostnames in /etc/hosts. I then logged in to a tailnet with no global DNS settings. With this, tailscaled writes /etc/resolv.conf (direct mode, as no systemd-resolved) and sets it to 100.100.100.100, which then has regular DNS via the metadata IP and *.internal DNS via the metadata IP as well. If the tailnet configures explicit DNS servers, those are used instead, except for *.internal. This also adds a new util/cloudenv package based on version/distro where the cloud type is only detected once. We'll likely expand it in the future for other clouds, doing variants of this change for other popular cloud environments. Fixes #4911 RELNOTES=Google Cloud DNS improvements Change-Id: I19f3c2075983669b2b2c0f29a548da8de373c7cf Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2022-06-29 20:19:34 +00:00
{
name: "google_cloud",
nm: &netmap.NetworkMap{
DNS: tailcfg.DNSConfig{},
},
cloud: cloudenv.GCP,
prefs: &ipn.Prefs{
CorpDNS: true,
},
want: &dns.Config{
Hosts: map[dnsname.FQDN][]netaddr.IP{},
Routes: map[dnsname.FQDN][]*dnstype.Resolver{
"internal.": []*dnstype.Resolver{{Addr: cloudenv.GoogleMetadataAndDNSIP}},
},
},
},
{
name: "google_cloud_with_exiting_internal",
nm: &netmap.NetworkMap{
DNS: tailcfg.DNSConfig{
Routes: map[string][]*dnstype.Resolver{
".internal": []*dnstype.Resolver{{Addr: "1.2.3.4"}},
},
},
},
cloud: cloudenv.GCP,
prefs: &ipn.Prefs{
CorpDNS: true,
},
want: &dns.Config{
Hosts: map[dnsname.FQDN][]netaddr.IP{},
Routes: map[dnsname.FQDN][]*dnstype.Resolver{
"internal.": []*dnstype.Resolver{{Addr: "1.2.3.4"}},
},
},
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
verOS := tt.os
if verOS == "" {
verOS = "linux"
}
var log tstest.MemLogger
ipn/ipnlocal, net/dns*, util/cloudenv: specialize DNS config on Google Cloud This does three things: * If you're on GCP, it adds a *.internal DNS split route to the metadata server, so we never break GCP DNS names. This lets people have some Tailscale nodes on GCP and some not (e.g. laptops at home) without having to add a Tailnet-wide *.internal DNS route. If you already have such a route, though, it won't overwrite it. * If the 100.100.100.100 DNS forwarder has nowhere to forward to, it forwards it to the GCP metadata IP, which forwards to 8.8.8.8. This means there are never errNoUpstreams ("upstream nameservers not set") errors on GCP due to e.g. mangled /etc/resolv.conf (GCP default VMs don't have systemd-resolved, so it's likely a DNS supremacy fight) * makes the DNS fallback mechanism use the GCP metadata IP as a fallback before our hosted HTTP-based fallbacks I created a default GCP VM from their web wizard. It has no systemd-resolved. I then made its /etc/resolv.conf be empty and deleted its GCP hostnames in /etc/hosts. I then logged in to a tailnet with no global DNS settings. With this, tailscaled writes /etc/resolv.conf (direct mode, as no systemd-resolved) and sets it to 100.100.100.100, which then has regular DNS via the metadata IP and *.internal DNS via the metadata IP as well. If the tailnet configures explicit DNS servers, those are used instead, except for *.internal. This also adds a new util/cloudenv package based on version/distro where the cloud type is only detected once. We'll likely expand it in the future for other clouds, doing variants of this change for other popular cloud environments. Fixes #4911 RELNOTES=Google Cloud DNS improvements Change-Id: I19f3c2075983669b2b2c0f29a548da8de373c7cf Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2022-06-29 20:19:34 +00:00
got := dnsConfigForNetmap(tt.nm, tt.prefs, log.Logf, verOS, tt.cloud)
if !reflect.DeepEqual(got, tt.want) {
gotj, _ := json.MarshalIndent(got, "", "\t")
wantj, _ := json.MarshalIndent(tt.want, "", "\t")
t.Errorf("wrong\n got: %s\n\nwant: %s\n", gotj, wantj)
}
if got := log.String(); got != tt.wantLog {
t.Errorf("log output wrong\n got: %q\nwant: %q\n", got, tt.wantLog)
}
})
}
}
func TestAllowExitNodeDNSProxyToServeName(t *testing.T) {
b := &LocalBackend{}
if b.allowExitNodeDNSProxyToServeName("google.com") {
t.Fatal("unexpected true on backend with nil NetMap")
}
b.netMap = &netmap.NetworkMap{
DNS: tailcfg.DNSConfig{
ExitNodeFilteredSet: []string{
".ts.net",
"some.exact.bad",
},
},
}
tests := []struct {
name string
want bool
}{
// Allow by default:
{"google.com", true},
{"GOOGLE.com", true},
// Rejected by suffix:
{"foo.TS.NET", false},
{"foo.ts.net", false},
// Suffix doesn't match
{"ts.net", true},
// Rejected by exact match:
{"some.exact.bad", false},
{"SOME.EXACT.BAD", false},
// But a prefix is okay.
{"prefix-okay.some.exact.bad", true},
}
for _, tt := range tests {
got := b.allowExitNodeDNSProxyToServeName(tt.name)
if got != tt.want {
t.Errorf("for %q = %v; want %v", tt.name, got, tt.want)
}
}
}