Palo Alto reported interpreting hairpin probes as LAND attacks, and the
firewalls may be responding to this by shutting down otherwise in use NAT sessions
prematurely. We don't currently make use of the outcome of the hairpin
probes, and they contribute to other user confusion with e.g. the
AirPort Extreme hairpin session workaround. We decided in response to
remove the whole probe feature as a result.
Updates #188
Updates tailscale/corp#19106
Updates tailscale/corp#19116
Signed-off-by: James Tucker <james@tailscale.com>
After some analysis, stateful filtering is only necessary in tailnets
that use `autogroup:danger-all` in `src` in ACLs. And in those cases
users explicitly specify that hosts outside of the tailnet should be
able to reach their nodes. To fix local DNS breakage in containers, we
disable stateful filtering by default.
Updates #12108
Signed-off-by: Andrew Lytvynov <awly@tailscale.com>
It was requested by the first customer 4-5 years ago and only used
for a brief moment of time. We later added netmap visibility trimming
which removes the need for this.
It's been hidden by the CLI for quite some time and never documented
anywhere else.
This keeps the CLI flag, though, out of caution. It just returns an
error if it's set to anything but true (its default).
Fixes#12058
Change-Id: I7514ba572e7b82519b04ed603ff9f3bdbaecfda7
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Palo Alto firewalls have a typically hard NAT, but also have a mode
called Persistent DIPP that is supposed to provide consistent port
mapping suitable for STUN resolution of public ports. Persistent DIPP
works initially on most Palo Alto firewalls, but some models/software
versions have a bug which this works around.
The bug symptom presents as follows:
- STUN sessions resolve a consistent public IP:port to start with
- Much later netchecks report the same IP:Port for a subset of
sessions, most often the users active DERP, and/or the port related
to sustained traffic.
- The broader set of DERPs in a full netcheck will now consistently
observe a new IP:Port.
- After this point of observation, new inbound connections will only
succeed to the new IP:Port observed, and existing/old sessions will
only work to the old binding.
In this patch we now advertise the lowest latency global endpoint
discovered as we always have, but in addition any global endpoints that
are observed more than once in a single netcheck report. This should
provide viable endpoints for potential connection establishment across
a NAT with this behavior.
Updates tailscale/corp#19106
Signed-off-by: James Tucker <james@tailscale.com>
Fixestailscale/tailscale#10393Fixestailscale/corp#15412Fixestailscale/corp#19808
On Apple platforms, exit nodes and subnet routers have been unable to relay pings from Tailscale devices to non-Tailscale devices due to sandbox restrictions imposed on our network extensions by Apple. The sandbox prevented the code in netstack.go from spawning the `ping` process which we were using.
Replace that exec call with logic to send an ICMP echo request directly, which appears to work in userspace, and not trigger a sandbox violation in the syslog.
Signed-off-by: Andrea Gottardo <andrea@gottardo.me>
This adds a new `UserLogf` field to the `Server` struct.
When set this any logs generated by Server are logged using
`UserLogf` and all spammy backend logs are logged to `Logf`.
If it `UserLogf` is unset, we default to `log.Printf` and
if `Logf` is unset we discard all the spammy logs.
Fixes#12094
Signed-off-by: Maisem Ali <maisem@tailscale.com>
Turn off stateful filtering for egress proxies to allow cluster
traffic to be forwarded to tailnet.
Allow configuring stateful filter via tailscaled config file.
Deprecate EXPERIMENTAL_TS_CONFIGFILE_PATH env var and introduce a new
TS_EXPERIMENTAL_VERSIONED_CONFIG env var that can be used to provide
containerboot a directory that should contain one or more
tailscaled config files named cap-<tailscaled-cap-version>.hujson.
Containerboot will pick the one with the newest capability version
that is not newer than its current capability version.
Proxies with this change will not work with older Tailscale
Kubernetes operator versions - users must ensure that
the deployed operator is at the same version or newer (up to
4 version skew) than the proxies.
Updates tailscale/tailscale#12061
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
Co-authored-by: Maisem Ali <maisem@tailscale.com>
We were missing `snat-subnet-routes`, `stateful-filtering`
and `netfilter-mode`. Add those to set too.
Fixes#12061
Signed-off-by: Maisem Ali <maisem@tailscale.com>
We are now publishing nameserver images to tailscale/k8s-nameserver,
so we can start defaulting the images if users haven't set
them explicitly, same as we already do with proxy images.
The nameserver images are currently only published for unstable
track, so we have to use the static 'unstable' tag.
Once we start publishing to stable, we can make the operator
default to its own tag (because then we'll know that for each
operator tag X there is also a nameserver tag X as we always
cut all images for a given tag.
Updates tailscale/tailscale#10499
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
The CLI's "up" is kinda chaotic and LocalBackend.Start is kinda
chaotic and they both need to be redone/deleted (respectively), but
this fixes some buggy behavior meanwhile. We were previously calling
StartLoginInteractive (to start the controlclient's RegisterRequest)
redundantly in some cases, causing test flakes depending on timing and
up's weird state machine.
We only need to call StartLoginInteractive in the client if Start itself
doesn't. But Start doesn't tell us that. So cheat a bit and a put the
information about whether there's a current NodeKey in the ipn.Status.
It used to be accessible over LocalAPI via GetPrefs as a private key but
we removed that for security. But a bool is fine.
So then only call StartLoginInteractive if that bool is false and don't
do it in the WatchIPNBus loop.
Fixes#12028
Updates #12042
Change-Id: I0923c3f704a9d6afd825a858eb9a63ca7c1df294
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
There was a small window in ipnserver after we assigned a LocalBackend
to the ipnserver's atomic but before we Start'ed it where our
initalization Start could conflict with API calls from the LocalAPI.
Simplify that a bit and lay out the rules in the docs.
Updates #12028
Change-Id: Ic5f5e4861e26340599184e20e308e709edec68b1
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This way the default gets populated on first start, when no existing
state exists to migrate. Also fix `ipn.PrefsFromBytes` to preserve empty
fields, rather than layering `NewPrefs` values on top.
Updates https://github.com/tailscale/corp/issues/19623
Signed-off-by: Andrew Lytvynov <awly@tailscale.com>
The CLI "up" command is a historical mess, both on the CLI side and
the LocalBackend side. We're getting closer to cleaning it up, but in
the meantime it was again implicated in flaky tests.
In this case, the background goroutine running WatchIPNBus was very
occasionally running enough to get to its StartLoginInteractive call
before the original goroutine did its Start call. That meant
integration tests were very rarely but sometimes logging in with the
default control plane URL out on the internet
(controlplane.tailscale.com) instead of the localhost control server
for tests.
This also might've affected new Headscale etc users on initial "up".
Fixes#11960Fixes#11962
Change-Id: I36f8817b69267a99271b5ee78cb7dbf0fcc0bd34
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
We'd like to use tsdial.Dialer.UserDial instead of SystemDial for DNS over TCP.
This is primarily necessary to properly dial internal DNS servers accessible
over Tailscale and subnet routes. However, to avoid issues when switching
between Wi-Fi and cellular, we need to ensure that we don't retain connections
to any external addresses on the old interface. Therefore, we need to determine
which dialer to use internally based on the configured routes.
This plumbs routes and localRoutes from router.Config to tsdial.Dialer,
and updates UserDial to use either the peer dialer or the system dialer,
depending on the network address and the configured routes.
Updates tailscale/corp#18725
Fixes#4529
Signed-off-by: Nick Khyl <nickk@tailscale.com>
The netcheck client, when no UDP is available, probes distance using
HTTPS.
Several problems:
* It probes using /derp/latency-check.
* But cmd/derper serves the handler at /derp/probe
* Despite the difference, it work by accident until c8f4dfc8c0
which made netcheck's probe require a 2xx status code.
* in tests, we only use derphttp.Handler, so the cmd/derper-installed
mux routes aren't preesnt, so there's no probe. That breaks
tests in airplane mode. netcheck.Client then reports "unexpected
HTTP status 426" (Upgrade Required)
This makes derp handle both /derp/probe and /derp/latency-check
equivalently, and in both cmd/derper and derphttp.Handler standalone
modes.
I notice this when wgengine/magicsock TestActiveDiscovery was failing
in airplane mode (no wifi). It still doesn't pass, but it gets
further.
Fixes#11989
Change-Id: I45213d4bd137e0f29aac8bd4a9ac92091065113f
Not buying wifi on a short flight is a good way to find tests
that require network. Whoops.
Updates #cleanup
Change-Id: Ibe678e9c755d27269ad7206413ffe9971f07d298
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
cmd/k8s-operator: optionally update dnsrecords Configmap with DNS records for proxies.
This commit adds functionality to automatically populate
DNS records for the in-cluster ts.net nameserver
to allow cluster workloads to resolve MagicDNS names
associated with operator's proxies.
The records are created as follows:
* For tailscale Ingress proxies there will be
a record mapping the MagicDNS name of the Ingress
device and each proxy Pod's IP address.
* For cluster egress proxies, configured via
tailscale.com/tailnet-fqdn annotation, there will be
a record for each proxy Pod, mapping
the MagicDNS name of the exposed
tailnet workload to the proxy Pod's IP.
No records will be created for any other proxy types.
Records will only be created if users have configured
the operator to deploy an in-cluster ts.net nameserver
by applying tailscale.com/v1alpha1.DNSConfig.
It is user's responsibility to add the ts.net nameserver
as a stub nameserver for ts.net DNS names.
https://kubernetes.io/docs/tasks/administer-cluster/dns-custom-nameservers/#configuration-of-stub-domain-and-upstream-nameserver-using-corednshttps://cloud.google.com/kubernetes-engine/docs/how-to/kube-dns#upstream_nameservers
See also https://github.com/tailscale/tailscale/pull/11017
Updates tailscale/tailscale#10499
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
This configures localClient correctly during flag parsing, so that the --socket
option is effective when generating tab-completion results. For example, the
following would not connect to the system Tailscale for tab-completion results:
tailscale --socket=/tmp/tailscaled.socket switch <TAB>
Updates #3793
Signed-off-by: Paul Scott <paul@tailscale.com>
cmd/k8s-operator/deploy/chart: allow users to configure additional labels for the operator's Pod via Helm chart values.
Fixes#11947
Signed-off-by: Gabe Gorelick <gabe@hightouch.io>
* cmd/k8s-nameserver,k8s-operator: add a nameserver that can resolve ts.net DNS names in cluster.
Adds a simple nameserver that can respond to A record queries for ts.net DNS names.
It can respond to queries from in-memory records, populated from a ConfigMap
mounted at /config. It dynamically updates its records as the ConfigMap
contents changes.
It will respond with NXDOMAIN to queries for any other record types
(AAAA to be implemented in the future).
It can respond to queries over UDP or TCP. It runs a miekg/dns
DNS server with a single registered handler for ts.net domain names.
Queries for other domain names will be refused.
The intended use of this is:
1) to allow non-tailnet cluster workloads to talk to HTTPS tailnet
services exposed via Tailscale operator egress over HTTPS
2) to allow non-tailnet cluster workloads to talk to workloads in
the same cluster that have been exposed to tailnet over their
MagicDNS names but on their cluster IPs.
DNSConfig CRD can be used to configure
the operator to deploy kube nameserver (./cmd/k8s-nameserver) to cluster.
Updates tailscale/tailscale#10499
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
Before attempting to enable IPv6 forwarding in the proxy init container
check if the relevant module is found, else the container crashes
on hosts that don't have it.
Updates#11860
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
cmd/containerboot,kube,ipn/store/kubestore: allow interactive login and empty state Secrets, check perms
* Allow users to pre-create empty state Secrets
* Add a fake internal kube client, test functionality that has dependencies on kube client operations.
* Fix an issue where interactive login was not allowed in an edge case where state Secret does not exist
* Make the CheckSecretPermissions method report whether we have permissions to create/patch a Secret if it's determined that these operations will be needed
Updates tailscale/tailscale#11170
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
In prep for most of the package funcs in net/interfaces to become
methods in a long-lived netmon.Monitor that can cache things. (Many
of the funcs are very heavy to call regularly, whereas the long-lived
netmon.Monitor can subscribe to things from the OS and remember
answers to questions it's asked regularly later)
Updates tailscale/corp#10910
Updates tailscale/corp#18960
Updates #7967
Updates #3299
Change-Id: Ie4e8dedb70136af2d611b990b865a822cd1797e5
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
... in prep for merging the net/interfaces package into net/netmon.
This is a no-op change that updates a bunch of the API signatures ahead of
a future change to actually move things (and remove the type alias)
Updates tailscale/corp#10910
Updates tailscale/corp#18960
Updates #7967
Updates #3299
Change-Id: I477613388f09389214db0d77ccf24a65bff2199c
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Modifies containerboot to wait on tailscaled process
only, not on any child process of containerboot.
Waiting on any subprocess was racing with Go's
exec.Cmd.Run, used to run iptables commands and
that starts its own subprocesses and waits on them.
Containerboot itself does not run anything else
except for tailscaled, so there shouldn't be a need
to wait on anything else.
Updates tailscale/tailscale#11593
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
The goal is to move more network state accessors to netmon.Monitor
where they can be cheaper/cached. But first (this change and others)
we need to make sure the one netmon.Monitor is plumbed everywhere.
Some notable bits:
* tsdial.NewDialer is added, taking a now-required netmon
* because a tsdial.Dialer always has a netmon, anything taking both
a Dialer and a NetMon is now redundant; take only the Dialer and
get the NetMon from that if/when needed.
* netmon.NewStatic is added, primarily for tests
Updates tailscale/corp#10910
Updates tailscale/corp#18960
Updates #7967
Updates #3299
Change-Id: I877f9cb87618c4eb037cee098241d18da9c01691
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This has been a TODO for ages. Time to do it.
The goal is to move more network state accessors to netmon.Monitor
where they can be cheaper/cached.
Updates tailscale/corp#10910
Updates tailscale/corp#18960
Updates #7967
Updates #3299
Change-Id: I60fc6508cd2d8d079260bda371fc08b6318bcaf1
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Adds a new .spec.metrics field to ProxyClass to allow users to optionally serve
client metrics (tailscaled --debug) on <Pod-IP>:9001.
Metrics cannot currently be enabled for proxies that egress traffic to tailnet
and for Ingress proxies with tailscale.com/experimental-forward-cluster-traffic-via-ingress annotation
(because they currently forward all cluster traffic to their respective backends).
The assumption is that users will want to have these metrics enabled
continuously to be able to monitor proxy behaviour (as opposed to enabling
them temporarily for debugging). Hence we expose them on Pod IP to make it
easier to consume them i.e via Prometheus PodMonitor.
Updates tailscale/tailscale#11292
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
This adds a health.Tracker to tsd.System, accessible via
a new tsd.System.HealthTracker method.
In the future, that new method will return a tsd.System-specific
HealthTracker, so multiple tsnet.Servers in the same process are
isolated. For now, though, it just always returns the temporary
health.Global value. That permits incremental plumbing over a number
of changes. When the second to last health.Global reference is gone,
then the tsd.System.HealthTracker implementation can return a private
Tracker.
The primary plumbing this does is adding it to LocalBackend and its
dozen and change health calls. A few misc other callers are also
plumbed. Subsequent changes will flesh out other parts of the tree
(magicsock, controlclient, etc).
Updates #11874
Updates #4136
Change-Id: Id51e73cfc8a39110425b6dc19d18b3975eac75ce
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
* cmd/containerboot,util/linuxfw: support proxy backends specified by DNS name
Adds support for optionally configuring containerboot to proxy
traffic to backends configured by passing TS_EXPERIMENTAL_DEST_DNS_NAME env var
to containerboot.
Containerboot will periodically (every 10 minutes) attempt to resolve
the DNS name and ensure that all traffic sent to the node's
tailnet IP gets forwarded to the resolved backend IP addresses.
Currently:
- if the firewall mode is iptables, traffic will be load balanced
accross the backend IP addresses using round robin. There are
no health checks for whether the IPs are reachable.
- if the firewall mode is nftables traffic will only be forwarded
to the first IP address in the list. This is to be improved.
* cmd/k8s-operator: support ExternalName Services
Adds support for exposing endpoints, accessible from within
a cluster to the tailnet via DNS names using ExternalName Services.
This can be done by annotating the ExternalName Service with
tailscale.com/expose: "true" annotation.
The operator will deploy a proxy configured to route tailnet
traffic to the backend IPs that service.spec.externalName
resolves to. The backend IPs must be reachable from the operator's
namespace.
Updates tailscale/tailscale#10606
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
There is an undocumented 16KiB limit for text log messages.
However, the limit for JSON messages is 256KiB.
Even worse, logging JSON as text results in significant overhead
since each double quote needs to be escaped.
Instead, use logger.Logf.JSON to explicitly log the info as JSON.
We also modify osdiag to return the information as structured data
rather than implicitly have the package log on our behalf.
This gives more control to the caller on how to log.
Updates #7802
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
Creates new QNAP builder target, which builds go binaries then uses
docker to build into QNAP packages. Much of the docker/script code
here is pulled over from https://github.com/tailscale/tailscale-qpkg,
with adaptation into our builder structures.
The qnap/Tailscale folder contains static resources needed to build
Tailscale qpkg packages, and is an exact copy of the existing folder
in the tailscale-qpkg repo.
Builds can be run with:
```
sudo ./tool/go run ./cmd/dist build qnap
```
Updates tailscale/tailscale-qpkg#135
Signed-off-by: Sonia Appasamy <sonia@tailscale.com>
Kubernetes cluster domain defaults to 'cluster.local', but can also be customized.
We need to determine cluster domain to set up in-cluster forwarding to our egress proxies.
This was previously hardcoded to 'cluster.local', so was the egress proxies were not usable in clusters with custom domains.
This PR ensures that we attempt to determine the cluster domain by parsing /etc/resolv.conf.
In case the cluster domain cannot be determined from /etc/resolv.conf, we fall back to 'cluster.local'.
Updates tailscale/tailscale#10399,tailscale/tailscale#11445
Signed-off-by: Irbe Krumina <irbe@tailscale.com>
