Upcoming work on incremental netmap change handling will require some
replumbing of which subsystems get notified about what. Done naively,
it could break "tailscale status --json" visibility later. To make sure
I understood the flow of all the updates I was rereading the status code
and realized parts of ipnstate.Status were being populated by the wrong
subsystems.
The engine (wireguard) and magicsock (data plane, NAT traveral) should
only populate the stuff that they uniquely know. The WireGuard bits
were fine but magicsock was populating stuff stuff that LocalBackend
could've better handled, so move it there.
Updates #1909
Change-Id: I6d1b95d19a2d1b70fbb3c875fac8ea1e169e8cb0
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
If we don't have the ICMP hint available, such as on Android, we can use
the signal of rx traffic to bias toward a particular endpoint.
We don't want to stick to a particular endpoint for a very long time
without any signals, so the sticky time is reduced to 1 second, which is
large enough to avoid excessive packet reordering in the common case,
but should be small enough that either rx provides a strong signal, or
we rotate in a user-interactive schedule to another endpoint, improving
the feel of failover to other endpoints.
Updates #8999
Co-authored-by: Charlotte Brandhorst-Satzkorn <charlotte@tailscale.com>
Signed-off-by: James Tucker <james@tailscale.com>
Signed-off-by: Charlotte Brandhorst-Satzkorn <charlotte@tailscale.com>
If a node is flapping or otherwise generating lots of STUN endpoints, we
can end up caching a ton of useless values and sending them to peers.
Instead, let's apply a fixed per-Addr limit of endpoints that we cache,
so that we're only sending peers up to the N most recent.
Updates tailscale/corp#13890
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I8079a05b44220c46da55016c0e5fc96dd2135ef8
This change removes the noV4/noV6 check from addrForSendWireGuardLocked.
On Android, the client panics when reaching `rand.Intn()`, likely due to
the candidates list being containing no candidates. The suspicion is
that the `noV4` and the `noV6` are both being triggered causing the
loop to continue.
Updates tailscale/corp#12938
Updates #7826
Signed-off-by: Charlotte Brandhorst-Satzkorn <charlotte@tailscale.com>
Switch our best address selection to use a scoring-based approach, where
we boost each address based on whether it's a private IP or IPv6.
For users in cloud environments, this biases endpoint selection towards
using an endpoint that is less likely to cost the user money, and should
be less surprising to users.
This also involves updating the tests to not use private IPv4 addresses;
other than that change, the behaviour should be identical for existing
endpoints.
Updates #8097
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I069e3b399daea28be66b81f7e44fc27b2943d8af
This change introduces address selection for wireguard only endpoints.
If a endpoint has not been used before, an address is randomly selected
to be used based on information we know about, such as if they are able
to use IPv4 or IPv6. When an address is initially selected, we also
initiate a new ICMP ping to the endpoints addresses to determine which
endpoint offers the best latency. This information is then used to
update which endpoint we should be using based on the best possible
route. If the latency is the same for a IPv4 and an IPv6 address, IPv6
will be used.
Updates #7826
Signed-off-by: Charlotte Brandhorst-Satzkorn <charlotte@tailscale.com>
Avoid selecting an endpoint as "better" than the current endpoint if the
total latency improvement is less than 1%. This adds some hysteresis to
avoid flapping between endpoints for a minimal improvement in latency.
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: If8312e1768ea65c4b4d4e13d8de284b3825d7a73
Previously, when updating endpoints we would immediately stop
advertising any endpoint that wasn't discovered during
determineEndpoints. This could result in, for example, a case where we
performed an incremental netcheck, didn't get any of our three STUN
packets back, and then dropped our STUN endpoint from the set of
advertised endpoints... which would result in clients falling back to a
DERP connection until the next call to determineEndpoints.
Instead, let's cache endpoints that we've discovered and continue
reporting them to clients until a timeout expires. In the above case
where we temporarily don't have a discovered STUN endpoint, we would
continue reporting the old value, then re-discover the STUN endpoint
again and continue reporting it as normal, so clients never see a
withdrawal.
Updates tailscale/coral#108
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: I42de72e7418ab328a6c732bdefc74549708cf8b9
So we're staying within the netip.Addr/AddrPort consistently and
avoiding allocs/conversions to the legacy net addr types.
Updates #5162
Change-Id: I59feba60d3de39f773e68292d759766bac98c917
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
These tests are passing locally and on CI. They had failed earlier in
the day when first fixing up CI, and it is not immediately clear why. I
have cycled IPv6 support locally, but this should not have a substantial
effect.
Updates #7876
Signed-off-by: James Tucker <jftucker@gmail.com>
We accidentally switched to ./tool/go in
4022796484 which resulted in no longer
running Windows builds, as this is attempting to run a bash script.
I was unable to quickly fix the various tests that have regressed, so
instead I've added skips referencing #7876, which we need to back and
fix.
Updates #7262
Updates #7876
Signed-off-by: James Tucker <james@tailscale.com>
This makes `omitempty` actually work, and saves bytes in each map response.
Updates tailscale/corp#8020
Signed-off-by: Maisem Ali <maisem@tailscale.com>
A peer can have IsWireGuardOnly, which means it will not support DERP or
Disco, and it must have Endpoints filled in order to be usable.
In the present implementation only the first Endpoint will be used as
the bestAddr.
Updates tailscale/corp#10351
Co-authored-by: Charlotte Brandhorst-Satzkorn <charlotte@tailscale.com>
Co-authored-by: James Tucker <james@tailscale.com>
Signed-off-by: James Tucker <james@tailscale.com>
Identified in review in #7821 endpoint.discoKey and endpoint.discoShort
are often accessed without first taking endpoint.mu. The arrangement
with endpoint.mu is inconvenient for a good number of those call-sites,
so it is instead replaced with an atomic pointer to carry both pieces of
disco info. This will also help with #7821 that wants to add explicit
checks/guards to disable disco behaviors when disco keys are missing
which is necessarily implicitly mostly covered by this change.
Updates #7821
Signed-off-by: James Tucker <james@tailscale.com>
This commit implements UDP offloading for Linux. GSO size is passed to
and from the kernel via socket control messages. Support is probed at
runtime.
UDP GSO is dependent on checksum offload support on the egress netdev.
UDP GSO will be disabled in the event sendmmsg() returns EIO, which is
a strong signal that the egress netdev does not support checksum
offload.
Updates tailscale/corp#8734
Signed-off-by: Jordan Whited <jordan@tailscale.com>
This adds the util/sysresources package, which currently only contains a
function to return the total memory size of the current system.
Then, we modify magicsock to scale the number of buffered DERP messages
based on the system's available memory, ensuring that we never use a
value lower than the previous constant of 32.
Signed-off-by: Andrew Dunham <andrew@du.nham.ca>
Change-Id: Ib763c877de4d0d4ee88869078e7d512f6a3a148d
This updates all source files to use a new standard header for copyright
and license declaration. Notably, copyright no longer includes a date,
and we now use the standard SPDX-License-Identifier header.
This commit was done almost entirely mechanically with perl, and then
some minimal manual fixes.
Updates #6865
Signed-off-by: Will Norris <will@tailscale.com>
The Tailscale logging service has a hard limit on the maximum
log message size that can be accepted.
We want to ensure that netlog messages never exceed
this limit otherwise a client cannot transmit logs.
Move the goroutine for periodically dumping netlog messages
from wgengine/netlog to net/connstats.
This allows net/connstats to manage when it dumps messages,
either based on time or by size.
Updates tailscale/corp#8427
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
This is temporary while we work to upstream performance work in
https://github.com/WireGuard/wireguard-go/pull/64. A replace directive
is less ideal as it breaks dependent code without duplication of the
directive.
Signed-off-by: Jordan Whited <jordan@tailscale.com>
This commit updates the wireguard-go dependency and implements the
necessary changes to the tun.Device and conn.Bind implementations to
support passing vectors of packets in tailscaled. This significantly
improves throughput performance on Linux.
Updates #414
Signed-off-by: Jordan Whited <jordan@tailscale.com>
Signed-off-by: James Tucker <james@tailscale.com>
Co-authored-by: James Tucker <james@tailscale.com>
We would replace the existing real implementation of nettype.PacketConn
with a blockForeverConn, but that violates the contract of atomic.Value
(where the type cannot change). Fix by switching to a pointer value
(atomic.Pointer[nettype.PacketConn]).
A longstanding issue, but became more prevalent when we started binding
connections to interfaces on macOS and iOS (#6566), which could lead to
the bind call failing if the interface was no longer available.
Fixes#6641
Signed-off-by: Mihai Parparita <mihai@tailscale.com>
Previously, tstun.Wrapper and magicsock.Conn managed their
own statistics data structure and relied on an external call to
Extract to extract (and reset) the statistics.
This makes it difficult to ensure a maximum size on the statistics
as the caller has no introspection into whether the number
of unique connections is getting too large.
Invert the control flow such that a *connstats.Statistics
is registered with tstun.Wrapper and magicsock.Conn.
Methods on non-nil *connstats.Statistics are called for every packet.
This allows the implementation of connstats.Statistics (in the future)
to better control when it needs to flush to ensure
bounds on maximum sizes.
The value registered into tstun.Wrapper and magicsock.Conn could
be an interface, but that has two performance detriments:
1. Method calls on interface values are more expensive since
they must go through a virtual method dispatch.
2. The implementation would need a sync.Mutex to protect the
statistics value instead of using an atomic.Pointer.
Given that methods on constats.Statistics are called for every packet,
we want reduce the CPU cost on this hot path.
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
There aren't any in the wild, other than one we ran on purpose to keep
us honest, but we can bump that one forward to 0.100.
Change-Id: I129e70724b2d3f8edf3b496dc01eba3ac5a2a907
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
There is utility in logging traffic statistics that occurs at the physical layer.
That is, in order to send packets virtually to a particular tailscale IP address,
what physical endpoints did we need to communicate with?
This functionality logs IP addresses identical to
what had always been logged in magicsock prior to #5823,
so there is no increase in PII being logged.
ExtractStatistics returns a mapping of connections to counts.
The source is always a Tailscale IP address (without port),
while the destination is some endpoint reachable on WAN or LAN.
As a special case, traffic routed through DERP will use 127.3.3.40
as the destination address with the port being the DERP region.
This entire feature is only enabled if data-plane audit logging
is enabled on the tailnet (by default it is disabled).
Example of type of information logged:
------------------------------------ Tx[P/s] Tx[B/s] Rx[P/s] Rx[B/s]
PhysicalTraffic: 25.80 3.39Ki 38.80 5.57Ki
100.1.2.3 -> 143.11.22.33:41641 15.40 2.00Ki 23.20 3.37Ki
100.4.5.6 -> 192.168.0.100:41641 10.20 1.38Ki 15.60 2.20Ki
100.7.8.9 -> 127.3.3.40:2 0.20 6.40 0.00 0.00
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
This field seems seldom used and the documentation is wrong.
It is simpler to just derive its original value dynamically
when endpoint.DstToString is called.
This method is potentially used by wireguard-go,
but not in any code path is performance sensitive.
All calls to it use it in conjunction with fmt.Printf,
which is going to be slow anyways since it uses Go reflection.
Signed-off-by: Joe Tsai <joetsai@digital-static.net>
The io/ioutil package has been deprecated as of Go 1.16 [1]. This commit
replaces the existing io/ioutil functions with their new definitions in
io and os packages.
Reference: https://golang.org/doc/go1.16#ioutil
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
This is entirely optional (i.e. failing in this code is non-fatal) and
only enabled on Linux for now. Additionally, this new behaviour can be
disabled by setting the TS_DEBUG_DISABLE_AF_PACKET environment variable.
Updates #3824
Replaces #5474
Co-authored-by: Andrew Dunham <andrew@du.nham.ca>
Signed-off-by: David Anderson <danderson@tailscale.com>
The version string changed slightly. Adapt.
And always check the current Go version to prevent future
accidental regressions. I would have missed this one had
I not explicitly manually checked it.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
The blockForeverConn was only using its sync.Cond one side. Looks like it
was just forgotten.
Fixes#3671
Change-Id: I4ed0191982cdd0bfd451f133139428a4fa48238c
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
And it updates the build tag style on a couple files.
Change-Id: I84478d822c8de3f84b56fa1176c99d2ea5083237
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
I believe that this should eliminate the flakiness.
If GitHub CI manages to be even slower that can be believed
(and I can believe a lot at this point),
then we should roll this back and make some more invasive changes.
Updates #654Fixes#3247 (I hope)
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
wgengine/wgcfg: introduce wgcfg.NewDevice helper to disable roaming
at all call sites (one real plus several tests).
Fixestailscale/corp#3016.
Signed-off-by: David Anderson <danderson@tailscale.com>
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
There are a few remaining uses of testing.AllocsPerRun:
Two in which we only log the number of allocations,
and one in which dynamically calculate the allocations
target based on a different AllocsPerRun run.
This also allows us to tighten the "no allocs"
test in wgengine/filter.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
Renames only; continuation of earlier 8049063d35
These kept confusing me while working on #3088
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
As more prep for removing the false assumption that you're able to
map from DiscoKey to a single peer, move the lastPingFrom and lastPingTime
fields from the endpoint type to a new discoInfo type, effectively upgrading
the old sharedDiscoKey map (which only held a *[32]byte nacl precomputed key
as its value) to discoInfo which then includes that naclbox key.
Then start plumbing it into handlePing in prep for removing the need
for handlePing to take an endpoint parameter.
Updates #3088
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
AFAICT this was always present, the log read mid-execution was never safe.
But it seems like the recent magicsock refactoring made the race much
more likely.
Signed-off-by: David Anderson <danderson@tailscale.com>
Over time, other magicsock refactors have made Start effectively a
no-op, except that some other functions choose to panic if called
before Start.
Signed-off-by: David Anderson <danderson@tailscale.com>
Our prod code doesn't eagerly handshake, because our disco layer enables
on-demand handshaking. Configuring both peers to eagerly handshake leads
to WireGuard handshake races that make TestTwoDevicePing flaky.
Signed-off-by: David Anderson <danderson@tailscale.com>
It only existed to override one test-only behavior with a
different test-only behavior, in both cases working around
an annoying feature of our CI environments. Instead, handle
that weirdness entirely in the test code, with a tweaked
TestOnlyPacketListener that gets injected.
Signed-off-by: David Anderson <danderson@tailscale.com>
The docstring said it was meant for use in tests, but it's specifically a
special codepath that is _only_ used in tests, so make the claim stronger.
Signed-off-by: David Anderson <danderson@tailscale.com>
Instead of using the legacy codepath, teach discoEndpoint to handle
peers that have a home DERP, but no disco key. We can still communicate
with them, but only over DERP.
Signed-off-by: David Anderson <danderson@tailscale.com>
magicsock makes multiple calls to Now per packet.
Move to mono.Now. Changing some of the calls to
use package mono has a cascading effect,
causing non-per-packet call sites to also switch.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
logBufWriter had no serialization.
It just so happens that none of its users currently ever log concurrently.
Make it safe for concurrent use.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
The DERPTestPort int meant two things before: which port to use, and
whether to disable TLS verification. Users would like to set the port
without disabling TLS, so break it into two options.
Updates #1264
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
After allowing for custom DERP maps, it's convenient to be able to see their latency in
netcheck. This adds a query to the local tailscaled for the current DERPMap.
Updates #1264
Signed-off-by: julianknodt <julianknodt@gmail.com>
Pull in the latest version of wireguard-windows.
Switch to upstream wireguard-go.
This requires reverting all of our import paths.
Unfortunately, this has to happen at the same time.
The wireguard-go change is very low risk,
as that commit matches our fork almost exactly.
(The only changes are import paths, CI files, and a go.mod entry.)
So if there are issues as a result of this commit,
the first place to look is wireguard-windows changes.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
magicsock.Conn.ParseEndpoint requires a peer's public key,
disco key, and legacy ip/ports in order to do its job.
We currently accomplish that by:
* adding the public key in our wireguard-go fork
* encoding the disco key as magic hostname
* using a bespoke comma-separated encoding
It's a bit messy.
Instead, switch to something simpler: use a json-encoded struct
containing exactly the information we need, in the form we use it.
Our wireguard-go fork still adds the public key to the
address when it passes it to ParseEndpoint, but now the code
compensating for that is just a couple of simple, well-commented lines.
Once this commit is in, we can remove that part of the fork
and remove the compensating code.
Signed-off-by: Josh Bleecher Snyder <josharian@gmail.com>
For historical reasons, we ended up with two near-duplicate
copies of curve25519 key types, one in the wireguard-go module
(wgcfg) and one in the tailscale module (types/wgkey).
Then we moved wgcfg to the tailscale module.
We can now remove the wgcfg key type in favor of wgkey.
Signed-off-by: Josh Bleecher Snyder <josharian@gmail.com>
Track endpoints internally with a new tailcfg.Endpoint type that
includes a typed netaddr.IPPort (instead of just a string) and
includes a type for how that endpoint was discovered (STUN, local,
etc).
Use []tailcfg.Endpoint instead of []string internally.
At the last second, send it to the control server as the existing
[]string for endpoints, but also include a new parallel
MapRequest.EndpointType []tailcfg.EndpointType, so the control server
can start filtering out less-important endpoint changes from
new-enough clients. Notably, STUN-discovered endpoints can be filtered
out from 1.6+ clients, as they can discover them amongst each other
via CallMeMaybe disco exchanges started over DERP. And STUN endpoints
change a lot, causing a lot of MapResposne updates. But portmapped
endpoints are worth keeping for now, as they they work right away
without requiring the firewall traversal extra RTT dance.
End result will be less control->client bandwidth. (despite negligible
increase in client->control bandwidth)
Updates tailscale/corp#1543
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
