Commit Graph

65 Commits

Author SHA1 Message Date
Brad Fitzpatrick 91c9c33036 cmd/tailscaled: don't block ipnserver startup behind engine init on Windows
With this change, the ipnserver's safesocket.Listen (the localhost
tcp.Listen) happens right away, before any synchronous
TUN/DNS/Engine/etc setup work, which might be slow, especially on
early boot on Windows.

Because the safesocket.Listen starts up early, that means localhost
TCP dials (the safesocket.Connect from the GUI) complete successfully
and thus the GUI avoids the MessageBox error. (I verified that
pacifies it, even without a Listener.Accept; I'd feared that Windows
localhost was maybe special and avoided the normal listener backlog).

Once the GUI can then connect immediately without errors, the various
timeouts then matter less, because the backend is no longer trying to
race against the GUI's timeout. So keep retrying on errors for a
minute, or 10 minutes if the system just booted in the past 10
minutes.

This should fix the problem with Windows 10 desktops auto-logging in
and starting the Tailscale frontend which was then showing a
MessageBox error about failing to connect to tailscaled, which was
slow coming up because the Windows networking stack wasn't up
yet. Fingers crossed.

Fixes #1313 (previously #1187, etc)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-04-20 22:26:27 -07:00
Brad Fitzpatrick f99e63bb17 ipn: don't Logout when Windows GUI disconnects
Logout used to be a no-op, so the ipnserver previously synthensized a Logout
on disconnect. Now that Logout actually invalidates the node key that was
forcing all GUI closes to log people out.

Instead, add a method to LocalBackend to specifically mean "the
Windows GUI closed, please forget all the state".

Fixes tailscale/corp#1591 (ignoring the notification issues, tracked elsewhere)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-04-20 13:14:10 -07:00
Brad Fitzpatrick 8f3e453356 ipn, cmd/tailscale/cli: add pref to configure sudo-free operator user
From discussion with @danderson.

Fixes #1684 (in a different way)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-04-19 10:12:41 -07:00
Brad Fitzpatrick 950fc28887 ipn, paths, cmd/tailscaled: remove LegacyConfigPath, relaynode migration
It is time.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-04-07 10:15:45 -07:00
Maisem Ali db13b2d0c8 cmd/tailscale, ipn/localapi: add "tailscale bugreport" subcommand
Adding a subcommand which prints and logs a log marker. This should help
diagnose any issues that users face.

Fixes #1466

Signed-off-by: Maisem Ali <maisem@tailscale.com>
2021-03-31 15:19:51 -07:00
Brad Fitzpatrick 27c4dd9a97 Revert "cmd/tailscaled, ipn/{ipnlocal,ipnserver}: let netstack get access to LocalBackend"
This reverts commit 2bc518dcb2.

@namansood didn't end up needing it in his 770aa71ffb.
2021-03-16 12:33:13 -07:00
Brad Fitzpatrick 2bc518dcb2 cmd/tailscaled, ipn/{ipnlocal,ipnserver}: let netstack get access to LocalBackend
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-03-15 09:31:25 -07:00
Brad Fitzpatrick 43b30e463c ipn/ipnserver: refactor permissions checks a bit, document more, fix Windows
Windows was only running the localapi on the debug port which was a
stopgap at the time while doing peercreds work. Removed that, and
wired it up correctly, with some more docs.

More clean-up to do after 1.6, moving the localhost TCP auth code into
the peercreds package. But that's too much for now, so the docs will
have to suffice, even if it's at a bit of an awkward stage with the
newly-renamed "NotWindows" field, which still isn't named well, but
it's better than its old name of "Unknown" which hasn't been accurate
since unix sock peercreds work anyway.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-03-05 21:46:32 -08:00
Brad Fitzpatrick 1ca3e739f7 ipn/ipnserver: set PermitWrite on localapi handler
The TODO was easy now with peerCreds and the isReadonlyConn func.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-03-05 12:14:24 -08:00
Brad Fitzpatrick 15b6969a95 ipn/ipnserver: grant client r/w access if peer uid matches tailscaled
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-03-02 12:34:11 -08:00
Brad Fitzpatrick be779b3587 safesocket, ipn/ipnserver: unify peercred info, fix bug on FreeBSD etc
FreeBSD wasn't able to run "tailscale up" since the recent peercred
refactoring.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-03-02 11:23:26 -08:00
Brad Fitzpatrick 38dc6fe758 cmd/tailscaled, wgengine: remove --fake, replace with netstack
And add a --socks5-server flag.

And fix a race in SOCKS5 replies where the response header was written
concurrently with the copy from the backend.

Co-authored with Naman Sood.

Updates #707
Updates #504

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-03-01 11:09:19 -08:00
Brad Fitzpatrick f11952ad7f ipn/ipnserver: fix Windows connection auth regression
Regression from code movement in d3efe8caf6

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-02-23 13:27:04 -08:00
Brad Fitzpatrick 7038c09bc9 ipn/ipnserver: on darwin, let users who are admins use CLI without sudo
Tangentially related to #987, #177, #594, #925, #505

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-02-16 21:09:27 -08:00
Brad Fitzpatrick d3efe8caf6 safesocket, ipn/ipnserver: look up peer creds on Darwin
And open up socket permissions like Linux, now that we know who
connections are from.

This uses the new inet.af/peercred that supports Linux and Darwin at
the moment.

Fixes #1347
Fixes #1348

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-02-16 20:38:44 -08:00
Brad Fitzpatrick fdac0387a7 ipn/ipnserver, ipn/ipnlocal: move whois handler to new localapi package 2021-02-15 10:46:22 -08:00
Brad Fitzpatrick d76334d2f0 ipn: split LocalBackend off into new ipn/ipnlocal package
And move a couple other types down into leafier packages.

Now cmd/tailscale doesn't bring in netlink, magicsock, wgengine, etc.

Fixes #1181

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-02-04 14:04:23 -08:00
Brad Fitzpatrick 006a224f50 ipn/ipnserver, cmd/hello: do whois over unix socket, not debug http
Start of a local HTTP API. Not a stable interface yet.
2021-01-29 13:23:13 -08:00
Brad Fitzpatrick c611d8480b cmd/tailscaled: add whois/identd-ish debug handler 2021-01-28 15:31:52 -08:00
Brad Fitzpatrick 4d3c09ced4 ipn/ipnserver: on Windows in unattended mode, wait for Engine forever
Updates #1187
2021-01-25 15:32:13 -08:00
Brad Fitzpatrick 5611f290eb ipn, ipnserver: only require sudo on Linux for mutable CLI actions
This partially reverts d6e9fb1df0, which modified the permissions
on the tailscaled Unix socket and thus required "sudo tailscale" even
for "tailscale status".

Instead, open the permissions back up (on Linux only) but have the
server look at the peer creds and only permit read-only actions unless
you're root.

In the future we'll also have a group that can do mutable actions.

On OpenBSD and FreeBSD, the permissions on the socket remain locked
down to 0600 from d6e9fb1df0.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2021-01-15 10:13:00 -08:00
Brad Fitzpatrick 5aa5db89d6 cmd/tailscaled, wgengine/netstack: add start of gvisor userspace netstack work
Not usefully functional yet (mostly a proof of concept), but getting
it submitted for some work @namansood is going to do atop this.

Updates #707
Updates #634
Updates #48
Updates #835
2021-01-11 09:31:14 -08:00
Brad Fitzpatrick 053a1d1340 all: annotate log verbosity levels on most egregiously spammy log prints
Fixes #924
Fixes #282

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-12-21 12:59:33 -08:00
Christine Dodrill 7ea809897d ipn/ipnserver: enable systemd-notify support
Addresses #964

Still to be done:
- Figure out the correct logging lines in util/systemd
- Figure out if we need to slip the systemd.Status function anywhere
  else
- Log util/systemd errors? (most of the errors are of the "you cannot do
  anything about this, but it might be a bad idea to crash the program if
  it errors" kind)

Assistance in getting this over the finish line would help a lot.

Signed-off-by: Christine Dodrill <me@christine.website>

util/systemd: rename the nonlinux file to appease the magic

Signed-off-by: Christine Dodrill <me@christine.website>

util/systemd: fix package name

Signed-off-by: Christine Dodrill <me@christine.website>

util/systemd: fix review feedback from @mdlayher

Signed-off-by: Christine Dodrill <me@christine.website>

cmd/tailscale{,d}: update depaware manifests

Signed-off-by: Christine Dodrill <me@christine.website>

util/systemd: use sync.Once instead of func init

Signed-off-by: Christine Dodrill <me@christine.website>

control/controlclient: minor review feedback fixes

Signed-off-by: Christine Dodrill <me@christine.website>

{control,ipn,systemd}: fix review feedback

Signed-off-by: Christine Dodrill <me@christine.website>

review feedback fixes

Signed-off-by: Christine Dodrill <me@christine.website>

ipn: fix sprintf call

Signed-off-by: Christine Dodrill <me@christine.website>

ipn: make staticcheck less sad

Signed-off-by: Christine Dodrill <me@christine.website>

ipn: print IP address in connected status

Signed-off-by: Christine Dodrill <me@christine.website>

ipn: review feedback

Signed-off-by: Christine Dodrill <me@christine.website>

final fixups

Signed-off-by: Christine Dodrill <me@christine.website>
2020-12-15 08:39:06 -05:00
Brad Fitzpatrick acade77c86 ipn/ipnserver: add knob to disable babysitter 2020-11-17 15:26:39 -08:00
Brad Fitzpatrick 07b6ffd55c ipn: only use Prefs, not computed stateKey, to determine server mode
When the service was running without a client (e.g. after a reboot)
and then the owner logs in and the GUI attaches, the computed state
key changed to "" (driven by frontend prefs), and then it was falling
out of server mode, despite the GUI-provided prefs still saying it
wanted server mode.

Also add some logging. And remove a scary "Access denied" from a
user-visible error, making the two possible already-in-use error
messages consistent with each other.
2020-11-02 21:13:51 -08:00
Brad Fitzpatrick 20a357b386 ipn, ipn/ipnserver: add IPN state for server in use, handle explicitly
On Windows, we were previously treating a server used by different
users as a fatal error, which meant the second user (upon starting
Tailscale, explicitly or via Start Up programs) got an invasive error
message dialog.

Instead, give it its own IPN state and change the Notify.ErrMessage to
be details in that state. Then the Windows GUI can be less aggresive
about that happening.

Also,

* wait to close the IPN connection until the server ownership state
  changes so the GUI doesn't need to repeatedly reconnect to discover
  changes.

* fix a bug discovered during testing: on system reboot, the
  ipnserver's serverModeUser was getting cleared while the state
  transitioned from Unknown to Running. Instead, track 'inServerMode'
  explicitly and remove the old accessor method which was error prone.

* fix a rare bug where the client could start up and set the server
  mode prefs in its Start call and we wouldn't persist that to the
  StateStore storage's prefs start key. (Previously it was only via a
  prefs toggle at runtime)
2020-11-02 15:25:11 -08:00
Brad Fitzpatrick f3aa08de76 ipn/ipnserver: remove "Server mode" from a user-visible error message
That's an internal nickname.
2020-11-02 09:22:21 -08:00
Brad Fitzpatrick ff7ddd9d20 ipn/ipnserver: move Windows local disk logging up to the parent process
To capture panics, log.Printf writes to os.Stderr, etc.

Fixes #726
2020-10-29 15:02:04 -07:00
Brad Fitzpatrick c64718e9a0 ipn/ipnserver: work around os/user.LookupId failure on Windows
If we can't find the mapping from SID ("user ID") -> username, don't
treat that as a fatal. Apparently that happens in the wild for Reasons.
Ignore it for now. It's just a nice-to-have for error messages in the
rare multi-user case.

Updates #869

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-10-29 13:16:53 -07:00
David Anderson 54e6c3a290 version: use OSS repo's version when building.
When building with redo, also include the git commit hash
from the proprietary repo, so that we have a precise commit
that identifies all build info (including Go toolchain version).

Add a top-level build script demonstrating to downstream distros
how to burn the right information into builds.

Adjust `tailscale version` to print commit hashes when available.

Fixes #841.

Signed-off-by: David Anderson <danderson@tailscale.com>
2020-10-28 16:17:21 -07:00
Brad Fitzpatrick 515866d7c6 ipn, ipnserver, cmd/tailscale: add "server mode" support on Windows
This partially (but not yet fully) migrates Windows to tailscaled's
StateStore storage system.

This adds a new bool Pref, ForceDaemon, defined as:

// ForceDaemon specifies whether a platform that normally
// operates in "client mode" (that is, requires an active user
// logged in with the GUI app running) should keep running after the
// GUI ends and/or the user logs out.
//
// The only current applicable platform is Windows. This
// forced Windows to go into "server mode" where Tailscale is
// running even with no users logged in. This might also be
// used for macOS in the future. This setting has no effect
// for Linux/etc, which always operate in daemon mode.

Then, when ForceDaemon becomes true, we now write use the StateStore
to track which user started it in server mode, and store their prefs
under that key.

The ipnserver validates the connections/identities and informs that
LocalBackend which userid is currently in charge.

The GUI can then enable/disable server mode at runtime, without using
the CLI.

But the "tailscale up" CLI was also fixed, so Windows users can use
authkeys or ACL tags, etc.

Updates #275
2020-10-12 14:28:21 -07:00
Brad Fitzpatrick 638127530b ipn/ipnserver: prevent use by multiple Windows users, add HTML status page
It was previously possible for two different Windows users to connect
to the IPN server at once, but it didn't really work. They mostly
stepped on each other's toes and caused chaos.

Now only one can control it, but it can be active for everybody else.

Necessary dependency step for Windows server/headless mode (#275)

While here, finish wiring up the HTTP status page on Windows, now that
all the dependent pieces are available.
2020-10-09 12:20:47 -07:00
Josh Bleecher Snyder 585a0d8997 all: use testing.T.TempDir
Bit of Friday cleanup.

Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
2020-10-02 20:31:31 -07:00
Brad Fitzpatrick 4f71319f7c ipn/ipnserver: make ipnserver also be an HTTP server for localhost clients
For now it just says hello to show auth works. More later.
2020-09-11 15:11:28 -07:00
Brad Fitzpatrick dd97111d06 backoff: update to Go style, document a bit, make 30s explicit
Also, bit of behavior change: on non-nil err but expired context,
don't reset the consecutive failure count. I don't think the old
behavior was intentional.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-08-09 09:36:26 -07:00
Brad Fitzpatrick cbf71d5eba ipn/ipnserver: fix bug in earlier commit where conn can be stranded
If a connection causes getEngine to transition from broken to fixed,
that connection was getting lost.
2020-07-29 17:46:58 -07:00
Brad Fitzpatrick 9ff5b380cb ipn/ipnserver: staticcheck is not wrong
shamecube.gif
2020-07-29 15:15:05 -07:00
Brad Fitzpatrick 4aba86cc03 ipn/ipnserver: make Engine argument a func that tries again for each connection
So a backend in server-an-error state (as used by Windows) can try to
create a new Engine again each time somebody re-connects, relaunching
the GUI app.

(The proper fix is actually fixing Windows issues, but this makes things better
in the short term)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-07-29 14:33:33 -07:00
Brad Fitzpatrick 1f923124bf ipn/ipnserver: support simultaneous connections
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-07-15 21:39:09 -07:00
Dmytro Shynkevych 61abab999e
cmd/tailscaled: graceful shutdown (#534)
Signed-off-by: Dmytro Shynkevych <dmytro@tailscale.com>
2020-07-13 06:17:58 -04:00
Brad Fitzpatrick 3984f9be2f ipn, ipn/ipnserver: add support for serving in error-message-only mode
So Windows service failures can be propagated to the Windows UI client.
2020-07-08 14:20:01 -07:00
David Anderson 9cd4e65191 smallzstd: new package that constructs zstd small encoders/decoders.
It's just a config wrapper that passes "use less memory at the
expense of compression" parameters by default, so that we don't
accidentally construct resource-hungry (de)compressors.

Also includes a benchmark that measures the memory cost of the
small variants vs. the stock variants. The savings are significant
on both compressors (~8x less memory) and decompressors (~1.4x less,
not including the savings from the significantly smaller
window on the compression side - with those savings included it's
more like ~140x smaller).

BenchmarkSmallEncoder-8            	   56174	     19354 ns/op	      31 B/op	       0 allocs/op
BenchmarkSmallEncoderWithBuild-8   	    2900	    382940 ns/op	 1746547 B/op	      36 allocs/op
BenchmarkStockEncoder-8            	   48921	     25761 ns/op	     286 B/op	       0 allocs/op
BenchmarkStockEncoderWithBuild-8   	     426	   2630241 ns/op	13843842 B/op	     124 allocs/op
BenchmarkSmallDecoder-8            	  123814	      9344 ns/op	       0 B/op	       0 allocs/op
BenchmarkSmallDecoderWithBuild-8   	   41547	     27455 ns/op	   27694 B/op	      31 allocs/op
BenchmarkStockDecoder-8            	  129832	      9417 ns/op	       1 B/op	       0 allocs/op
BenchmarkStockDecoderWithBuild-8   	   25561	     51751 ns/op	   39607 B/op	      92 allocs/op

Signed-off-by: David Anderson <danderson@tailscale.com>
2020-07-02 16:13:06 -07:00
David Anderson c71754eba2 ipn/ipnserver: revert decoder memory limit.
The zstd library treats that limit as a hard cap on decompressed
size, in the mode we're using it, rather than a window size.

Signed-off-by: David Anderson <danderson@tailscale.com>
2020-06-03 02:41:49 +00:00
Avery Pennarun db051fb013 ipnserver and logpolicy: configure zstd with low-memory settings.
The compressed blobs we send back and forth are small and infrequent,
which doesn't justify the 8MB * GOMAXPROCS memory that was being
allocated. This was the overwhelming majority of memory use in
tailscaled. On my system it goes from ~100M RSS to ~15M RSS (which is
still suspiciously high, but we can worry about that more later).

Signed-off-by: Avery Pennarun <apenwarr@tailscale.com>
2020-05-20 11:23:26 -04:00
Wendi Yu bb55694c95
wgengine: log node IDs when peers are added/removed (#381)
Also stop logging data sent/received from nodes we're not connected to (ie all those `x`s being logged in the `peers: ` line)
Signed-off-by: Wendi <wendi.yu@yahoo.ca>
2020-05-15 14:13:44 -06:00
Avery Pennarun 08acb502e5 Add tstest.PanicOnLog(), and fix various problems detected by this.
If a test calls log.Printf, 'go test' horrifyingly rearranges the
output to no longer be in chronological order, which makes debugging
virtually impossible. Let's stop that from happening by making
log.Printf panic if called from any module, no matter how deep, during
tests.

This required us to change the default error handler in at least one
http.Server, as well as plumbing a bunch of logf functions around,
especially in magicsock and wgengine, but also in logtail and backoff.

To add insult to injury, 'go test' also rearranges the output when a
parent test has multiple sub-tests (all the sub-test's t.Logf is always
printed after all the parent tests t.Logf), so we need to screw around
with a special Logf that can point at the "current" t (current_t.Logf)
in some places. Probably our entire way of using subtests is wrong,
since 'go test' would probably like to run them all in parallel if you
called t.Parallel(), but it definitely can't because the're all
manipulating the shared state created by the parent test. They should
probably all be separate toplevel tests instead, with common
setup/teardown logic. But that's a job for another time.

Signed-off-by: Avery Pennarun <apenwarr@tailscale.com>
2020-05-13 23:12:35 -04:00
Wendi Yu 0c69b4e00d
Implement rate limiting on log messages (#356)
Implement rate limiting on log messages

Addresses issue #317, where logs can get spammed with the same message
nonstop. Created a rate limiting closure on logging functions, which
limits the number of messages being logged per second based on format
string. To keep memory usage as constant as possible, the previous cache
purging at periodic time intervals has been replaced by an LRU that
discards the oldest string when the capacity of the cache is reached.


Signed-off-by: Wendi Yu <wendi.yu@yahoo.ca>
2020-05-08 13:21:36 -06:00
Brad Fitzpatrick 45f2b53aca all: remove unnecessary trailing newlines in format patterns for consistency
And document on logger.Logf that it's unnecessary.
2020-04-11 08:35:34 -07:00
Brad Fitzpatrick a4ef345737 cmd/tailscale: add status subcommand
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2020-03-27 20:34:36 -07:00