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>
Previously it was both metadata about the class of warnable item as
well as the value.
Now it's only metadata and the value is per-Tracker.
Updates #11874
Updates #4136
Change-Id: Ia1ed1b6c95d34bc5aae36cffdb04279e6ba77015
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This moves most of the health package global variables to a new
`health.Tracker` type.
But then rather than plumbing the Tracker in tsd.System everywhere,
this only goes halfway and makes one new global Tracker
(`health.Global`) that all the existing callers now use.
A future change will eliminate that global.
Updates #11874
Updates #4136
Change-Id: I6ee27e0b2e35f68cb38fecdb3b2dc4c3f2e09d68
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
If this happens, it results in us pessimistically closing more
connections than might be necessary, but is more correct since we won't
"miss" a change to the default route interface and keep trying to send
data over a nonexistent interface, or one that can't reach the internet.
Updates tailscale/corp#19124
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: Ia0b8b04cb8cdcb0da0155fd08751c9dccba62c1a
-Move Android impl into interfaces_android.go
-Instead of using ip route to get the interface name, use the one passed in by Android (ip route is restricted in Android 13+ per termux/termux-app#2993)
Follow-up will be to do the same for router
Fixestailscale/corp#19215Fixestailscale/corp#19124
Signed-off-by: kari-ts <kari@tailscale.com>
This ensures that we close the underlying connection(s) when a major
link change happens. If we don't do this, on mobile platforms switching
between WiFi and cellular can result in leftover connections in the
http.Client's connection pool which are bound to the "wrong" interface.
Updates #10821
Updates tailscale/corp#19124
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: Ibd51ce2efcaf4bd68e14f6fdeded61d4e99f9a01
This wasn't previously handling the case where an interface in s2 was
removed and not present in s1, and would cause the Equal method to
incorrectly return that the states were equal.
Updates tailscale/corp#19124
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I3af22bc631015d1ddd0a1d01bfdf312161b9532d
It should've been deleted in 11ece02f52.
Updates #9040
Change-Id: If8a136bdb6c82804af658c9d2b0a8c63ce02d509
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
At least in userspace-networking mode.
Fixes#11361
Change-Id: I78d33f0f7e05fe9e9ee95b97c99b593f8fe498f2
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
We were being too aggressive when deciding whether to write our NRPT rules
to the local registry key or the group policy registry key.
After once again reviewing the document which calls itself a spec
(see issue), it is clear that the presence of the DnsPolicyConfig subkey
is the important part, not the presence of values set in the DNSClient
subkey. Furthermore, a footnote indicates that the presence of
DnsPolicyConfig in the GPO key will always override its counterpart in
the local key. The implication of this is important: we may unconditionally
write our NRPT rules to the local key. We copy our rules to the policy
key only when it contains NRPT rules belonging to somebody other than us.
Fixes https://github.com/tailscale/corp/issues/19071
Signed-off-by: Aaron Klotz <aaron@tailscale.com>
This removes a potentially increased boot delay for certain boot
topologies where they block on ExecStartPre that may have socket
activation dependencies on other system services (such as
systemd-resolved and NetworkManager).
Also rename cleanup to clean up in affected/immediately nearby places
per code review commentary.
Fixes#11599
Signed-off-by: James Tucker <james@tailscale.com>
At least in the case of dialing a Tailscale IP.
Updates #4529
Change-Id: I9fd667d088a14aec4a56e23aabc2b1ffddafa3fe
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Updates #7946
[@bradfitz fixed up version of #8417]
Change-Id: I1dbf6fa8d525b25c0d7ad5c559a7f937c3cd142a
Signed-off-by: alexelisenko <39712468+alexelisenko@users.noreply.github.com>
Signed-off-by: Alex Paguis <alex@windscribe.com>
The netcheck package and the magicksock package coordinate via the
health package, but both sides have time based heuristics through
indirect dependencies. These were misaligned, so the implemented
heuristic aimed at reducing DERP moves while there is active traffic
were non-operational about 3/5ths of the time.
It is problematic to setup a good test for this integration presently,
so instead I added comment breadcrumbs along with the initial fix.
Updates #8603
Signed-off-by: James Tucker <james@tailscale.com>
We now allow some more ICMP errors to flow, specifically:
- ICMP parameter problem in both IPv4 and IPv6 (corrupt headers)
- ICMP Packet Too Big (for IPv6 PMTU)
Updates #311
Updates #8102
Updates #11002
Signed-off-by: James Tucker <james@tailscale.com>
This implementation uses less memory than tempfork/device,
which helps avoid OOM conditions in the iOS VPN extension when
switching to a Tailnet with ExitNode routing enabled.
Updates tailscale/corp#18514
Signed-off-by: Percy Wegmann <percy@tailscale.com>
I was running all tests while preparing a recent stable release, and
this was failing because my computer is connected to a fairly large
tailnet.
```
--- FAIL: TestGetRouteTable (0.01s)
routetable_linux_test.go:32: expected at least one default route;
...
```
```
$ ip route show table 52 | wc -l
1051
```
Updates #cleanup
Signed-off-by: Anton Tolchanov <anton@tailscale.com>
This test could hang because the subprocess was blocked on writing to
the stdout pipe if we find the address we're looking for early in the
output.
Updates #cleanup
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I68d82c22a5d782098187ae6d8577e43063b72573
The `stack.PacketBufferPtr` type no longer exists; replace it with
`*stack.PacketBuffer` instead.
Updates #8043
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: Ib56ceff09166a042aa3d9b80f50b2aa2d34b3683
When reverse path filtering is in strict mode on Linux, using an exit
node blocks all network connectivity. This change adds a warning about
this to `tailscale status` and the logs.
Example in `tailscale status`:
```
- not connected to home DERP region 22
- The following issues on your machine will likely make usage of exit nodes impossible: [interface "eth0" has strict reverse-path filtering enabled], please set rp_filter=2 instead of rp_filter=1; see https://github.com/tailscale/tailscale/issues/3310
```
Example in the logs:
```
2024/02/21 21:17:07 health("overall"): error: multiple errors:
not in map poll
The following issues on your machine will likely make usage of exit nodes impossible: [interface "eth0" has strict reverse-path filtering enabled], please set rp_filter=2 instead of rp_filter=1; see https://github.com/tailscale/tailscale/issues/3310
```
Updates #3310
Signed-off-by: Anton Tolchanov <anton@tailscale.com>
Tailscaled becomes inoperative if the Tailscale Tunnel wintun adapter is abruptly removed.
wireguard-go closes the device in case of a read error, but tailscaled keeps running.
This adds detection of a closed WireGuard device, triggering a graceful shutdown of tailscaled.
It is then restarted by the tailscaled watchdog service process.
Fixes#11222
Signed-off-by: Nick Khyl <nickk@tailscale.com>
The WinTun adapter may have been removed by the time we're closing
the dns.windowsManager, and its associated interface registry key might
also have been deleted. We shouldn't use winutil.OpenKeyWait and wait
for the interface key to appear when performing a cleanup as a part of
the windowsManager shutdown.
Updates #11222
Signed-off-by: Nick Khyl <nickk@tailscale.com>
If a client socket is remotely lost but the client is not sent an RST in
response to the next request, the socket might sit in RTO for extended
lengths of time, resulting in "no internet" for users. Instead, timeout
after 10s, which will close the underlying socket, recovering from the
situation more promptly.
Updates #10967
Signed-off-by: James Tucker <james@tailscale.com>
We don't need a log line every time defaultRoute is read in the good
case, and we now only log default interface updates that are actually
changes.
Updates #3363
Signed-off-by: James Tucker <james@tailscale.com>
Setting a user timeout will be a more practical tuning knob for a number
of endpoints, this provides a way to set it.
Updates tailscale/corp#17587
Signed-off-by: James Tucker <james@tailscale.com>
This fixes an infinite loop caused by the configuration of
systemd-resolved on Amazon Linux 2023 and how that interacts with
Tailscale's "direct" mode. We now drop the Tailscale service IP from the
OS's "base configuration" when we detect this configuration.
Updates #7816
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I73a4ea8e65571eb368c7e179f36af2c049a588ee
These are functionally the same as the "urn:schemas-upnp-org" services
with a few minor changes, and are still used by older devices. Support
them to improve our ability to obtain an external IP on such networks.
Updates #10911
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I05501fad9d6f0a3b8cf19fc95eee80e7d16cc2cf
This no longer results in a nil pointer exception when we get a valid
UPnP response with no supported clients.
Updates #10911
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I6e3715a49a193ff5261013871ad7fff197a4d77e
The prefix has space for 32-bit site IDs, but the validateViaPrefix
function would previously have disallowed site IDs greater than 255.
Fixestailscale/corp#16470
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I4cdb0711dafb577fae72d86c4014cf623fa538ef
Add a standalone server for STUN that can be hosted independently of the
derper, and factor that back into the derper.
Fixes#8434Closes#8435Closes#10745
Signed-off-by: James Tucker <james@tailscale.com>
To make it easier to correlate the starting/ending log messages.
Updates #cleanup
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I2802d53ad98e19bc8914bc58f8c04d4443227b26
Updates #8022
Updates #6075
On iOS, we currently rely on delegated interface information to figure out the default route interface. The NetworkExtension framework in iOS seems to set the delegate interface only once, upon the *creation* of the VPN tunnel. If a network transition (e.g. from Wi-Fi to Cellular) happens while the tunnel is connected, it will be ignored and we will still try to set Wi-Fi as the default route because the delegated interface is not getting updated as connectivity transitions.
Here we work around this on the Swift side with a NWPathMonitor instance that observes the interface name of the first currently satisfied network path. Our Swift code will call into `UpdateLastKnownDefaultRouteInterface`, so we can rely on that when it is set.
If for any reason the Swift machinery didn't work and we don't get any updates, here we also have some fallback logic: we try finding a hardcoded Wi-Fi interface called en0. If en0 is down, we fall back to cellular (pdp_ip0) as a last resort. This doesn't handle all edge cases like USB-Ethernet adapters or multiple Ethernet interfaces, but it is good enough to ensure connectivity isn't broken.
I tested this on iPhones and iPads running iOS 17.1 and it appears to work. Switching between different cellular plans on a dual SIM configuration also works (the interface name remains pdp_ip0).
Signed-off-by: Andrea Gottardo <andrea@tailscale.com>
If the epoch that we see during a Probe is less than the existing epoch,
it means that the gateway has either restarted or reset its
configuration, and an existing mapping is no longer valid. Reset any
saved mapping(s) if we detect this case so that a future
createOrGetMapping will not attempt to re-use it.
Updates #10597
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: Ie3cddaf625cb94a29885f7a1eeea25dbf6b97b47
When the portable Monitor creates a winMon via newOSMon, we register
address and route change callbacks with Windows. Once a callback is hit,
it starts a goroutine that attempts to send the event into messagec and returns.
The newly started goroutine then blocks until it can send to the channel.
However, if the monitor is never started and winMon.Receive is never called,
the goroutines remain indefinitely blocked, leading to goroutine leaks and
significant memory consumption in the tailscaled service process on Windows.
Unlike the tailscaled subprocess, the service process creates but never starts
a Monitor.
This PR adds a check within the callbacks to confirm the monitor's active status,
and exits immediately if the monitor hasn't started.
Updates #9864
Signed-off-by: Nick Khyl <nickk@tailscale.com>
Run `staticcheck` with `U1000` to find unused code. This cleans up about
a half of it. I'll do the other half separately to keep PRs manageable.
Updates #cleanup
Signed-off-by: Andrew Lytvynov <awly@tailscale.com>
This logs additional information about what mapping(s) are obtained
during the creation process, including whether we return an existing
cached mapping.
Updates #10597
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I9ff25071f064c91691db9ab0b9365ccc5f948d6e
Currently, we get the "likely home router" gateway IP and then iterate
through all IPs for all interfaces trying to match IPs to determine the
source IP. However, on many platforms we know what interface the gateway
is through, and thus we don't need to iterate through all interfaces
checking IPs. Instead, use the IP address of the associated interface.
This better handles the case where we have multiple interfaces on a
system all connected to the same gateway, and where the first interface
that we visit (as iterated by ForeachInterfaceAddress) isn't also the
default internet route.
Updates #8992
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I8632f577f1136930f4ec60c76376527a19a47d1f
