Tangentially related to #987, #177, #594, #925, #505
Motivated by rebooting a launchd-controlled tailscaled and it going
into SetNetworkUp(false) mode immediately because there really is no
network up at system boot, but then it got stuck in that paused state
forever, without a monitor implementation.
When a handshake race occurs, a queued data packet can get lost.
TestTwoDevicePing expected that the very first data packet would arrive.
This caused occasional flakes.
Change TestTwoDevicePing to repeatedly re-send packets
and succeed when one of them makes it through.
This is acceptable (vs making WireGuard not drop the packets)
because this only affects communication with extremely old clients.
And those extremely old clients will eventually connect,
because the kernel will retry sends on timeout.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
We modified the standard net package to not allocate a *net.UDPAddr
during a call to (*net.UDPConn).ReadFromUDP if the caller's use
of the *net.UDPAddr does not cause it to escape.
That is https://golang.org/cl/291390.
This is the companion change to magicsock.
There are two changes required.
First, call ReadFromUDP instead of ReadFrom, if possible.
ReadFrom returns a net.Addr, which is an interface, which always allocates.
Second, reduce the lifetime of the returned *net.UDPAddr.
We do this by immediately converting it into a netaddr.IPPort.
We left the existing RebindingUDPConn.ReadFrom method in place,
as it is required to satisfy the net.PacketConn interface.
With the upstream change and both of these fixes in place,
we have removed one large allocation per packet received.
name old time/op new time/op delta
ReceiveFrom-8 16.7µs ± 5% 16.4µs ± 8% ~ (p=0.310 n=5+5)
name old alloc/op new alloc/op delta
ReceiveFrom-8 112B ± 0% 64B ± 0% -42.86% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
ReceiveFrom-8 3.00 ± 0% 2.00 ± 0% -33.33% (p=0.008 n=5+5)
Co-authored-by: Sonia Appasamy <sonia@tailscale.com>
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
addrSet maintained duplicate lists of netaddr.IPPorts and net.UDPAddrs.
Unify to use the netaddr type only.
This makes (*Conn).ReceiveIPvN a bit uglier,
but that'll be cleaned up in a subsequent commit.
This is preparatory work to remove an allocation from ReceiveIPv4.
Co-authored-by: Sonia Appasamy <sonia@tailscale.com>
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
I based my estimation of the required timeout based on locally
observed behavior. But CI machines are worse than my local machine.
16s was enough to reduce flakiness but not eliminate it. Bump it up again.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
It only affects 'go install ./...', etc, and only on darwin/arm64 (M1 Macs) where
the go-ole package doesn't compile.
No need to build it.
Updates #943
This was in place because retrieved allowed_ips was very expensive.
Upstream changed the data structure to make them cheaper to compute.
This commit is an experiment to find out whether they're now cheap enough.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
We removed the "fast retry" code from our wireguard-go fork.
As a result, pings can take longer to transit when retries are required.
Allow that.
Fixes#1277
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
The fix can make this test run unconditionally.
This moves code from 5c619882bc for
testability but doesn't fix it yet. The #1282 problem remains (when I
wrote its wake-up mechanism, I forgot there were N DERP readers
funneling into 1 UDP reader, and the code just isn't correct at all
for that case).
Also factor out some test helper code from BenchmarkReceiveFrom.
The refactoring in magicsock.go for testability should have no
behavior change.
Magicsock started dropping all traffic internally when Tailscale is
shut down, to avoid spurious wireguard logspam. This made the benchmark
not receive anything. Setting a dummy private key is sufficient to get
magicsock to pass traffic for benchmarking purposes.
Fixes#1270.
Signed-off-by: David Anderson <danderson@tailscale.com>
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>
Unused for now, but I want to backport this commit to 1.4 so 1.6 can
start sending these and then at least 1.4 logs will stringify nicely.
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Use tb.Cleanup to simplify both the API and the implementation.
One behavior change: When the number of goroutines shrinks, don't log.
I've never found these logs to be useful, and they frequently add noise.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
Previously we disabled v6 support if the disable_policy knob was
missing in /proc, but some kernels support policy routing without
exposing the toggle. So instead, treat disable_policy absence as a
"maybe", and make the direct `ip -6 rule` probing a bit more
elaborate to compensate.
Fixes#1241.
Signed-off-by: David Anderson <danderson@tailscale.com>
This is mostly code movement from the wireguard-go repo.
Most of the new wgcfg package corresponds to the wireguard-go wgcfg package.
wgengine/wgcfg/device{_test}.go was device/config{_test}.go.
There were substantive but simple changes to device_test.go to remove
internal package device references.
The API of device.Config (now wgcfg.DeviceConfig) grew an error return;
we previously logged the error and threw it away.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
On Windows, configureInterface starts a goroutine reconfiguring the
Windows firewall.
But if configureInterface fails later, that goroutine kept running and
likely failing forever, spamming logs. Make it stop quietly if its
launching goroutine filed.
Rewrite log lines on the fly, based on the set of known peers.
This enables us to use upstream wireguard-go logging,
but maintain the Tailscale-style peer public key identifiers
that the rest of our systems (and people) expect.
Fixes#1183
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
Also, don't try to use IPv6 LinkLocalUnicast addresses for now. Like endpoints
exchanged with control, we share them but don't yet use them.
Updates #1172
c8c493f3d9 made it always say
`created=false` which scared me when I saw it, as that would've implied
things were broken much worse. Fortunately the logging was just wrong.
DstToString is used in two places in wireguard-go: Logging and uapi.
We are switching to use uapi for wireguard-go config.
To preserve existing behavior, we need the full set of addrs.
And for logging, having the full set of addrs seems useful.
(The Addrs method itself is slated for removal. When that happens,
the implementation will move to DstToString.)
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
To save CPU and wakeups, don't run the DERP cleanup timer regularly
unless there is a non-home DERP connection open.
Also eliminates the goroutine, moving to a time.AfterFunc.
Updates #1034
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This reverts commit 08baa17d9a.
It caused deadlocks due to lock ordering violations.
It was not the right fix, and thus should simply be reverted
while we look for the right fix (if we haven't already found it
in the interim; we've fixed other logging-after-test issues).
Fixes#1161
context.cancelCtx.Done involves a mutex and isn't as cheap as I
previously assumed. Convert the donec method into a struct field and
store the channel value once. Our one magicsock.Conn gets one pointer
larger, but it cuts ~1% of the CPU time of the ReceiveFrom benchmark
and removes a bubble from the --svg output :)
This test serves two purposes:
+ check that Write() returns an error if the tstun has been
closed.
+ ensure that the close-related code in tstun is exercised in
a test case. We were getting spurious code coverage adds/drops
based on timing of when the test case finished.
Signed-off-by: Denton Gentry <dgentry@tailscale.com>
22507adf54 stopped relying on
our fork of wireguard-go's UpdateDst callback.
As a result, we can unwind that code,
and the extra return value of ReceiveIPv{4,6}.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
TwoDevicePing is explicitly testing the behavior of the legacy codepath, everything
else is happy to assume that code no longer exists.
Signed-off-by: David Anderson <danderson@tailscale.com>
Previously, this benchmark relied on behavior of the legacy
receive codepath, which I changed in 22507adf. With this
change, the benchmark instead relies on the new active discovery
path.
Signed-off-by: David Anderson <danderson@tailscale.com>
This prevents us from continuing to do unnecessary work
(including logging) after the connection has closed.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
This adds a new IP Protocol type, TSMP on protocol number 99 for
sending inter-tailscale messages over WireGuard, currently just for
why a peer rejects TCP SYNs (ACL rejection, shields up, and in the
future: nothing listening, something listening on that port but wrong
interface, etc)
Updates #1094
Updates tailscale/corp#1185
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Commit 68ddf1 removed code that reads
`SOFTWARE\Tailscale IPN\SearchList` registry value. But the commit
left code that writes that value.
So now this package writes and never reads the value.
Remove the code to stop pointless work.
Updates #853
Signed-off-by: Alex Brainman <alex.brainman@gmail.com>
This eliminates a dependency on wgcfg.Endpoint,
as part of the effort to eliminate our wireguard-go fork.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
This makes connectivity between ancient and new tailscale nodes slightly
worse in some cases, but only in cases where the ancient version would
likely have failed to get connectivity anyway.
Signed-off-by: David Anderson <danderson@tailscale.com>
This is what every other DNS resolver I could find does, so tsdns
should do it to. This also helps avoid weird error messages about
non-existent records being unimplemented, and thus fixes#848.
Signed-off-by: Smitty <me@smitop.com>
In sendDiscoMessage there is a check of whether the connection is
closed, which is not being reliably exercised by other tests.
This shows up in code coverage reports, the lines of code in
sendDiscoMessage are alternately added and subtracted from
code coverage.
Add a test to specifically exercise and verify this code path.
Signed-off-by: Denton Gentry <dgentry@tailscale.com>
In derpWriteChanOfAddr when we call derphttp.NewRegionClient(),
there is a check of whether the connection is already errored and
if so it returns before grabbing the lock. The lock might already
be held and would be a deadlock.
This corner case is not being reliably exercised by other tests.
This shows up in code coverage reports, the lines of code in
derpWriteChanOfAddr are alternately added and subtracted from
code coverage.
Add a test to specifically exercise this code path, and verify that
it doesn't deadlock.
This is the best tradeoff I could come up with:
+ the moment code calls Err() to check if there is an error, we
grab the lock to make sure it would deadlock if it tries to grab
the lock itself.
+ if a new call to Err() is added in this code path, only the
first one will be covered and the rest will not be tested.
+ this test doesn't verify whether code is checking for Err() in
the right place, which ideally I guess it would.
Signed-off-by: Denton Gentry <dgentry@tailscale.com>
netaddr.IP no longer allocates, so don't need a cache or all its associated
code/complexity.
This totally removes groupcache/lru from the deps.
Also go mod tidy.
* wengine/netstack: bump gvisor to latest version
Signed-off-by: Naman Sood <naman@tailscale.com>
* update dependencies
Signed-off-by: Naman Sood <naman@tailscale.com>
* Don't change hardcoded IP
Signed-off-by: Naman Sood <naman@tailscale.com>
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
* show DNS name over hostname, removing domain's common MagicDNS suffix.
only show hostname if there's no DNS name.
but still show shared devices' MagicDNS FQDN.
* remove nerdy low-level details by default: endpoints, DERP relay,
public key. They're available in JSON mode still for those who need
them.
* only show endpoint or DERP relay when it's active with the goal of
making debugging easier. (so it's easier for users to understand
what's happening) The asterisks are gone.
* remove Tx/Rx numbers by default for idle peers; only show them when
there's traffic.
* include peers' owner login names
* add CLI option to not show peers (matching --self=true, --peers= also
defaults to true)
* sort by DNS/host name, not public key
* reorder columns
The log lines that wireguard-go prints as it starts
and stops its worker routines are mostly noise.
They also happen after other work is completed,
which causes failures in some of the log testing packages.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
This appears to have been the intent of the previous code,
but in practice, it only returned A records.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
To be honest I'm not fond of Golden Bytes tests like this, but
not so much as to want to rewrite the whole test. The DNS byte
format is essentially immutable at this point, the encoded bytes
aren't going to change. The rest of the test assumptions about
hostnames might, but we can fix that when it comes.
Signed-off-by: Denton Gentry <dgentry@tailscale.com>
eccc167 introduced closeHandle which opened the handle,
but never closed it.
Windows handles should be closed.
Updates #921
Signed-off-by: Alex Brainman <alex.brainman@gmail.com>
Previously the client had heuristics to calculate which DNS search domains
to set, based on the peers' names. Unfortunately that prevented us from
doing some things we wanted to do server-side related to node sharing.
So, bump MapRequest.Version to 9 to signal that the client only uses the
explicitly configured DNS search domains and doesn't augment it with its own
list.
Updates tailscale/corp#1026
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
This is a replacement for the key-related parts
of the wireguard-go wgcfg package.
This is almost a straight copy/paste from the wgcfg package.
I have slightly changed some of the exported functions and types
to avoid stutter, added and tweaked some comments,
and removed some now-unused code.
To avoid having wireguard-go depend on this new package,
wgcfg will keep its key types.
We translate into and out of those types at the last minute.
These few remaining uses will be eliminated alongside
the rest of the wgcfg package.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
The windows key timeout is longer than the wgengine watchdog timeout,
which means we never reach the timeout, instead the process exits.
Reduce the timeout so if we do hit it, at least the process continues.
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
On Win10, there's a hardcoded GUID and this works.
On Win7, this GUID changes and we need to ask the tun for its
LUID and convert that from the GUID.
This commit uses the computed GUID that is placed in InterfaceName.
Diagnosed by Jason Donnenfeld. (Thanks!)
Lazy wg configuration now triggers if a peer has only endpoint
addresses (/32 for IPv4, /128 for IPv6). Subnet routers still
trigger eager configuration to avoid the need for a CIDR match
in the hot packet path.
Signed-off-by: David Anderson <danderson@tailscale.com>
The previous code used a lot of whole-function variables and shared
behavior that only triggered based on prior action from a single codepath.
Instead of that, move the small amounts of "shared" code into each switch
case.
Signed-off-by: David Anderson <danderson@tailscale.com>
Before, tailscaled would log every 10 seconds when the periodic noteRecvActivity
call happens. This is noisy, but worse it's misleading, because the message
suggests that the disco code is starting a lazy config run for a missing peer,
whereas in fact it's just an internal piece of keepalive logic.
With this change, we still log when going from 0->1 tunnel for the peer, but
not every 10s thereafter.
Signed-off-by: David Anderson <danderson@tailscale.com>
While the code was correct, I broke it during a refactoring and
tests didn't detect it. This fixes that glitch.
Signed-off-by: David Anderson <danderson@tailscale.com>
Doesn't materially affect benchmarks, but shrinks match6 by 30 instructions
and halves memory loads.
Part of #19.
Signed-off-by: David Anderson <danderson@tailscale.com>
Part of #19.
name old time/op new time/op delta
Filter/icmp4-8 32.2ns ± 3% 32.5ns ± 2% ~ (p=0.524 n=10+8)
Filter/icmp6-8 49.7ns ± 6% 43.1ns ± 4% -13.12% (p=0.000 n=9+10)
Signed-off-by: David Anderson <danderson@tailscale.com>
The packet filter still rejects all IPv6, but decodes enough from v6
packets to do something smarter in a followup.
name time/op
Decode/tcp4-8 28.8ns ± 2%
Decode/tcp6-8 20.6ns ± 1%
Decode/udp4-8 28.2ns ± 1%
Decode/udp6-8 20.0ns ± 6%
Decode/icmp4-8 21.7ns ± 2%
Decode/icmp6-8 14.1ns ± 2%
Decode/unknown-8 9.43ns ± 2%
Signed-off-by: David Anderson <danderson@tailscale.com>
In tests, we force binding to localhost to avoid OS firewall warning
dialogs.
But for IPv6, we were trying (and failing) to bind to 127.0.0.1.
You'd think we'd just say "localhost", but that's apparently ill
defined. See
https://tools.ietf.org/html/draft-ietf-dnsop-let-localhost-be-localhost
and golang/go#22826. (It's bitten me in the past, but I can't
remember specific bugs.)
So use "::1" explicitly for "udp6", which makes the test quieter.
The goal is to move some of the shenanigans we have elsewhere into the filter
package, so that all the weird things to do with poking at the filter is in
a single place, behind clean APIs.
Signed-off-by: David Anderson <danderson@tailscale.com>
We still use the packet.* alloc-free types in the data path, but
the compilation from netaddr to packet happens within the filter
package.
Signed-off-by: David Anderson <danderson@tailscale.com>
os.IsNotExist doesn't unwrap errors. errors.Is does.
The ioutil.ReadFile ones happened to be fine but I changed them so
we're consistent with the rule: if the error comes from os, you can
use os.IsNotExist, but from any other package, use errors.Is.
(errors.Is always would also work, but not worth updating all the code)
The motivation here was that we were logging about failure to migrate
legacy relay node prefs file on startup, even though the code tried
to avoid that.
See golang/go#41122
Amazingly, there doesn't seem to be a documented way of updating network
configuration programmatically in a way that Windows takes notice of.
The naturopathic remedy for this is to invoke ipconfig /registerdns, which
does a variety of harmless things and also invokes the private API that
tells windows to notice new adapter settings. This makes our DNS config
changes stick within a few seconds of us setting them.
If we're invoking a shell command anyway, why futz with the registry at
all? Because netsh has no command for changing the DNS suffix list, and
its commands for setting resolvers requires parsing its output and
keeping track of which server is in what index. Amazingly, twiddling
the registry directly is the less painful option.
Fixes#853.
Signed-off-by: David Anderson <danderson@tailscale.com>
Updating the Windows firewall is usually reasonably fast, but
sometimes blocks for 20 seconds, 4 minutes, etc. Not sure why.
Until we understand that's happening, configure it in the background
without blocking the normal control flow.
Updates #785
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
The RusagePrefixLog is rarely useful, hasn't been useful in a long
time, is rarely the measurement we need, and is pretty spammy (and
syscall-heavy). Disable it by default. We can enable it when we're
debugging memory.
At startup the client doesn't yet have the DERP map so can't do STUN
queries against DERP servers, so it only knows it local interface
addresses, not its STUN-mapped addresses.
We were reporting the interface-local addresses to control, getting
the DERP map, and then immediately reporting the full set of
updates. That was an extra HTTP request to control, but worse: it was
an extra broadcast from control out to all the peers in the network.
Now, skip the initial update if there are no stun results and we don't
have a DERP map.
More work remains optimizing start-up requests/map updates, but this
is a start.
Updates tailscale/corp#557
There was a bug with the lazy wireguard config code where, if the
minimum set of peers to tell wireguard didn't change, we skipped
calling userspaceEngine.updateActivityMapsLocked which updated
the various data structures that matched incoming traffic to later
reconfigure the minimum config.
That meant if an idle peer restarted and changed discovery keys, we
skipped updating our maps of disco keys/IPs that would caused us to
lazily inflate the config for that peer later if/when it did send
traffic.
Use golang.zx2c4.com/wireguard/windows/tunnel/winipcfg
instead of github.com/tailscale/winipcfg-go package.
Updates #760
Signed-off-by: Alex Brainman <alex.brainman@gmail.com>
Due to a copy/paste-o, we were monitoring address changes twice, and
not monitoring route changes at all.
Verified with 'tailscale debug --monitor' that this actually works now (while
running 'route add 10.3.0.0 mask 255.255.0.0 10.0.0.1' and 'route delete (same)'
back and forth in cmd.exe)
In practice route changes are accompanied by address changes and this
doesn't fix any known issues. I just noticed this while reading this
code again. But at least the code does what it was trying to do now.
This function is only called in fake mode, which won't do anything more
with the packet after we respond to it anyway, so dropping it in the
prefilter is not necessary. And it's kinda semantically wrong: we did
not reject it, so telling the upper layer that it was rejected produces
an ugly error message.
Signed-off-by: Avery Pennarun <apenwarr@tailscale.com>
At some point faketun got implemented as a loopback (put a packet in
from wireguard, the same packet goes back to wireguard) which is not
useful. It's supposed to be an interface that just sinks all packets,
and then wgengine adds *only* and ICMP Echo responder as a layer on
top.
This caused extremely odd bugs on darwin, where the special case that
reinjects packets from local->local was filling the loopback channel
and creating an infinite loop (which became jammed since the reader and
writer were in the same goroutine).
Signed-off-by: Avery Pennarun <apenwarr@tailscale.com>
If no interfaces are up, calm down and stop spamming so much. It was
noticed as especially bad on Windows, but probably was bad
everywhere. I just have the best network conditions testing on a
Windows VM.
Updates #604
Gracefully skips touching the v6 NAT table on systems that don't have
it, and doesn't configure IPv6 at all if IPv6 is globally disabled.
Signed-off-by: David Anderson <danderson@tailscale.com>
When the network link changes, existing UDP sockets fail immediately
and permanently on macOS.
The forwarder set up a single UDP conn and never changed it.
As a result, any time there was a network link change,
all forwarded DNS queries failed.
To fix this, create a new connection when send requests
fail because of network unreachability.
This change is darwin-only, although extended it to other platforms
should be straightforward.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
While we're here, parseQuery into a plain function.
This is helpful for fuzzing. (Which I did a bit of. Didn't find anything.)
And clean up a few minor things.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
Part of unforking our winipcfg-go and using upstream (#760), move our
additions into our repo. (We might upstream them later if upstream has
interest)
Originally these were:
@apenwarr: "Add ifc.SyncAddresses() and SyncRoutes()."
609dcf2df5
@bradfitz: "winipcfg: make Interface.AddRoutes do as much as possible, return combined error"
e9f93d53f3
@bradfitz: "prevent unnecessary Interface.SyncAddresses work; normalize IPNets in deltaNets"
decb9ee8e1
Might fix it. I've spent too much time failing to reproduce the issue. This doesn't
seem to make it worse, though (it still runs for me), so I'll include this and
see if it helps others while I still work on a reliable way to reproduce it.
Updates tailscale/corp#474
Otherwise when PAC server is down, we log, and each log entry is a new
HTTP request (from logtail) and a new GetProxyForURL call, which again
logs, non-stop. This is also nicer to the WinHTTP service.
Then also hook up link change notifications to the cache to reset it
if there's a chance the network might work sooner.
This change is to restore /etc/resolv.conf after tailscale down is called. This is done by setting the dns.Manager before errors occur. Error collection is also added.
Fixes#723
* wgengine/router/router_linux.go: Switched `cidrDiff("addr")` and `cidrDiff("route")` order
Signed-off-by: Christina Wen <christina@tailscale.com>
Co-authored-by: Christina Wen <christina@tailscale.com>
Also remove rebinding logic from the windows router. Magicsock will
instead rebind based on link change signals.
Signed-off-by: David Anderson <danderson@tailscale.com>
Updates #654. See that issue for a discussion of why
this timeout reduces flakiness, and what next steps are.
Signed-off-by: Josh Bleecher Snyder <josh@tailscale.com>
For now. Get it working again so it's not stuck on 0.98.
Subnet relay can come later.
Updates #451
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Start of making the IPN state machine react to link changes and down
its DNS & routes if necessary to unblock proxy resolution (e.g. for
transitioning from public to corp networks where the corp network has
mandatory proxies and WPAD PAC files that can't be resolved while
using the DNS/routes configured previously)
This change should be a no-op. Just some callback plumbing.
It turns out that otherwise we don't know what exactly was set.
Also remove the now unused RootDomain config option.
Signed-off-by: Dmytro Shynkevych <dmytro@tailscale.com>
Previously, a type AAAA query would be answered with an A record
if only an IPv4 address was available. This is irrelevant for us
while we only use IPv4, but it will be a bug one day,
so it's worth being precise about semantics.
Signed-off-by: Dmytro Shynkevych <dmytro@tailscale.com>
Rather than consider bigs jumps in last-received-from activity as a
signal to possibly reconfigure the set of wireguard peers to have
configured, instead just track the set of peers that are currently
excluded from the configuration. Easier to reason about.
Also adds a bit more logging.
This might fix an error we saw on a machine running a recent unstable
build:
2020-08-26 17:54:11.528033751 +0000 UTC: 8.6M/92.6M magicsock: [unexpected] lazy endpoint not created for [UcppE], d:42a770f678357249
2020-08-26 17:54:13.691305296 +0000 UTC: 8.7M/92.6M magicsock: DERP packet received from idle peer [UcppE]; created=false
2020-08-26 17:54:13.691383687 +0000 UTC: 8.7M/92.6M magicsock: DERP packet from unknown key: [UcppE]
If it does happen again, though, we'll have more logs.
Seems to break linux CI builder. Cannot reproduce locally,
so attempting a rollback.
This reverts commit cd7bc02ab1.
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
Without this, a freshly started ipn client will be stuck in the
"Starting" state until something triggers a call to RequestStatus.
Usually a UI does this, but until then we can sit in this state
until poked by an external event, as is evidenced by our e2e tests
locking up when DERP is attached.
(This only recently became a problem when we enabled lazy handshaking
everywhere, otherwise the wireugard tunnel creation would also
trigger a RequestStatus.)
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
The previous approach modifies name in-place in the request slice to avoid an allocation.
This is incorrect: the question section of a DNS request
must be copied verbatim, without any such modification.
Software may rely on it (we rely on other resolvers doing it it in tsdns/forwarder).
Signed-off-by: Dmytro Shynkevych <dmytro@tailscale.com>
control/controlclient, wgengine/filter: extract parsePacketFilter to new constructor in wgengine/filter
Signed-off-by: chungdaniel <daniel@tailscale.com>
Consider:
Hard NAT (A) <---> Hard NAT w/ mapped port (B)
If A sends a packet to B's mapped port, A can disco ping B directly,
with low latency, without DERP.
But B couldn't establish a path back to A and needed to use DERP,
despite already logging about A's endpoint and adding a mapping to it
for other purposes (the wireguard conn.Endpoint lookup also needed
it).
This adds the tracking to discoEndpoint too so it'll be used for
finding a path back.
Fixestailscale/corp#556
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
For example:
$ tailscale ping -h
USAGE
ping <hostname-or-IP>
FLAGS
-c 10 max number of pings to send
-stop-once-direct true stop once a direct path is established
-verbose false verbose output
$ tailscale ping mon.ts.tailscale.com
pong from monitoring (100.88.178.64) via DERP(sfo) in 65ms
pong from monitoring (100.88.178.64) via DERP(sfo) in 252ms
pong from monitoring (100.88.178.64) via [2604:a880:2:d1::36:d001]:41641 in 33ms
Fixes#661
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
1) we weren't waking up a discoEndpoint that once existed and
went idle for 5 minutes and then got a disco message again.
2) userspaceEngine.noteReceiveActivity had a buggy check; fixed
and added a test
This removes the atomic bool that tried to track whether we needed to acquire
the lock on a future recursive call back into magicsock. Unfortunately that
hack doesn't work because we also had a lock ordering issue between magicsock
and userspaceEngine (see issue). This documents that too.
Fixes#644
If a node is behind a hard NAT and is using an explicit local port
number, assume they might've mapped a port and add their public IPv4
address with the local tailscaled's port number as a candidate endpoint.
A comparison operator was backwards.
The bad case went:
* device A send packet to B at t=1s
* B gets added to A's wireguard config
* B gets packet
(5 minutes pass)
* some other activity happens, causing B to expire
to be removed from A's network map, since it's
been over 5 minutes since sent or received activity
* device A sends packet to B at t=5m1s
* normally, B would get added back, but the old send
time was not zero (we sent earlier!) and the time
comparison was backwards, so we never regenerated
the wireguard config.
This also refactors the code for legibility and moves constants up
top, with comments.
The OS (tries) to send these but we drop them. No need to worry the
user with spam that we're dropping it.
Fixes#402
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
Starting with fe68841dc7, some e2e tests
got flaky. Rather than debug them (they're gnarly), just revert to the old
behavior as far as those tests are concerned. The tests were somehow
using magicsock without a private key and expecting it to do ... something.
My goal with fe68841dc7 was to stop log spam
and unnecessary work I saw on the iOS app when when stopping the app.
Instead, only stop doing that work on any transition from
once-had-a-private-key to no-longer-have-a-private-key. That fixes
what I wanted to fix while still making the mysterious e2e tests
happy.
There is a race in natlab where we might start shutdown while natlab is still running
a goroutine or two to deliver packets. This adds a small grace period to try and receive
it before continuing shutdown.
Signed-off-by: David Anderson <danderson@tailscale.com>
The first packet to transit may take several seconds to do so, because
setup rates in wgengine may result in the initial WireGuard handshake
init to get dropped. So, we have to wait at least long enough for a
retransmit to correct the fault.
Signed-off-by: David Anderson <danderson@tailscale.com>
Active discovery lets us introspect the state of the network stack precisely
enough that it's unnecessary, and dropping the initial DERP packets greatly
slows down tests. Additionally, it's unrealistic since our production network
will never deliver _only_ discovery packets, it'll be all or nothing.
Signed-off-by: David Anderson <danderson@tailscale.com>
Uses natlab only, because the point of this active discovery test is going to be
that it should get through a lot of obstacles.
Signed-off-by: David Anderson <danderson@tailscale.com>
The deadlock was:
* Conn.Close was called, which acquired c.mu
* Then this goroutine scheduled:
if firstDerp {
startGate = c.derpStarted
go func() {
dc.Connect(ctx)
close(c.derpStarted)
}()
}
* The getRegion hook for that derphttp.Client then ran, which also
tries to acquire c.mu.
This change makes that hook first see if we're already in a closing
state and then it can pretend that region doesn't exist.
wireguard-go uses 3 goroutines per peer (with reasonably large stacks
& buffers).
Rather than tell wireguard-go about all our peers, only tell it about
peers we're actively communicating with. That means we need hooks into
magicsock's packet receiving path and tstun's packet sending path to
lazily create a wireguard peer on demand from the network map.
This frees up lots of memory for iOS (where we have almost nothing
left for larger domains with many users).
We should ideally do this in wireguard-go itself one day, but that'd
be a pretty big change.
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
sync.Pools should almost always be packate globals, even though in this
case we only have exactly 1 TUN device anyway, so it matters less.
Still, it's unusual to see a Pool that's not a package global, so move it.
We originally picked those numbers somewhat at random, but with the idea
that 8 is a traditionally lucky number in Chinese culture. Unfortunately,
"88" is also neo-nazi shorthand language.
Use 52 instead, because those are the digits above the letters
"TS" (tailscale) on a qwerty keyboard, so we're unlikely to collide with
other users. 5, 2 and 52 are also pleasantly culturally meaningless.
Signed-off-by: David Anderson <danderson@tailscale.com>
Brad Fitzpatrick