Commit Graph

65 Commits (6c3820e8c4434b4b547403dc37a5416ad6d6ddf8)

Author SHA1 Message Date
Brad Fitzpatrick db2436c7ff wgengine/magicsock: don't interrupt endpoint updates, merge all mutex into one
Before, endpoint updates were constantly being interrupted and resumed
on Linux due to tons of LinkChange messages from over-zealous Linux
netlink messages (from router_linux.go)

Now that endpoint updates are fast and bounded in time anyway, just
let them run to completion, but note that another needs to be
scheduled after.

Now logs went from pages of noise to just:

root@taildoc:~# grep -i -E 'stun|endpoint update' log
2020/03/13 08:51:29 magicsock.Conn: starting endpoint update (initial)
2020/03/13 08:51:30 magicsock.Conn.ReSTUN: endpoint update active, need another later ("link-change-minor")
2020/03/13 08:51:31 magicsock.Conn: starting endpoint update (link-change-minor)
2020/03/13 08:51:31 magicsock.Conn.ReSTUN: endpoint update active, need another later ("link-change-minor")
2020/03/13 08:51:33 magicsock.Conn: starting endpoint update (link-change-minor)
2020/03/13 08:51:33 magicsock.Conn.ReSTUN: endpoint update active, need another later ("link-change-minor")
2020/03/13 08:51:35 magicsock.Conn: starting endpoint update (link-change-minor)
2020/03/13 08:51:35 magicsock.Conn.ReSTUN: endpoint update active, need another later ("link-change-minor")

Or, seen in another run:

2020/03/13 08:45:41 magicsock.Conn: starting endpoint update (periodic)
2020/03/13 08:46:09 magicsock.Conn: starting endpoint update (periodic)
2020/03/13 08:46:21 magicsock.Conn: starting endpoint update (link-change-major)
2020/03/13 08:46:37 magicsock.Conn: starting endpoint update (periodic)
2020/03/13 08:47:05 magicsock.Conn: starting endpoint update (periodic)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick db31550854 wgengine: don't Reconfig on boring link changes
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick bc73dcf204 wgengine/magicsock: don't block in Send waiting for derphttp.Send
Fixes #137
Updates #109
Updates #162
Updates #163

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick 8807913be9 wgengine/magicsock: wait for previous DERP goroutines to end before new ones
Updates #109 (hopefully fixes, will wait for graphs to be happy)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick eff6dcdb4e wgengine/magicsock: log more about why we're re-STUNing 5 years ago
Brad Fitzpatrick b3ddf51a15 wgengine/magicsock: add a pointer value for logging
Updates #109

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
David Crawshaw 7ec54e0064 wgengine/magicsock: remove TODO
The TODO above derphttp.NewClient suggests it does network I/O,
but the derphttp client connects lazily and so creating one is
very cheap.

Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
5 years ago
Brad Fitzpatrick 01b4bec33f stunner: re-do how Stunner works
It used to make assumptions based on having Anycast IPs that are super
near. Now we're intentionally going to a bunch of different distant
IPs to measure latency.

Also, optimize how the hairpin detection works. No need to STUN on
that socket. Just use that separate socket for sending, once we know
the other UDP4 socket's endpoint. The trick is: make our test probe
also a STUN packet, so it fits through magicsock's existing STUN
routing.

This drops netcheck from ~5 seconds to ~250-500ms.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
David Anderson 86baf60bd4 wgengine/magicsock: synchronize epUpdate cleanup on shutdown.
Signed-off-by: David Anderson <danderson@tailscale.com>
5 years ago
Brad Fitzpatrick 023df9239e Move linkstate boring change filtering to magicsock
So we can at least re-STUN on boring updates.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick a265d7cbff wgengine/magicsock: in STUN-disabled test mode, let endpoint discovery proceed 5 years ago
Brad Fitzpatrick 5c1e443d34 wgengine/monitor: don't call LinkChange when interfaces look unchanged
Basically, don't trust the OS-level link monitor to only tell you
interesting things. Sanity check it.

Also, move the interfaces package into the net directory now that we
have it.
5 years ago
Brad Fitzpatrick 39c0ae1dba derp/derpmap: new DERP config package, merge netcheck into magicsock more
Fixes #153
Updates #162
Updates #163

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick 4800926006 wgengine/magicsock: add AddrSet appendDests+UpdateDst tests 5 years ago
David Anderson bb93d7aaba wgengine/magicsock: plumb logf throughout, and expose in Options.
Signed-off-by: David Anderson <danderson@tailscale.com>
5 years ago
Brad Fitzpatrick f42b9b6c9a wgengine/magicsock: don't discard UDP packet on UDP+DERP race
Fixes #155

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
David Anderson f265603110 wgengine/magicsock: fix data race in ReceiveIPv4.
The UDP reader goroutine was clobbering `n` and `err` from the
main goroutine, whose accesses are not synchronized the way `b` is.

Signed-off-by: David Anderson <danderson@tailscale.com>
5 years ago
David Anderson 77354d4617 wgengine/magicsock: unblock wireguard-go's read on magicsock shutdown.
wireguard-go closes magicsock, and expects this to unblock reads
so that its internal goroutines can wind down. We were incorrectly
blocking the read indefinitey and breaking this contract.

Signed-off-by: David Anderson <danderson@tailscale.com>
5 years ago
David Anderson fdee5fb639 wgengine/magicsock: don't mutexly reach inside Conn to tweak DERP settings.
Signed-off-by: David Anderson <danderson@tailscale.com>
5 years ago
Brad Fitzpatrick 61d83f759b wgengine/magicsock: remove redundant derpMagicIP comparison
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
David Anderson ef31dd7bb5 wgengine/magicsock: check all 3 fast paths independently.
The previous code would skip the DERP short-circuit if roamAddr
was set, which is not what we wanted. More generally, hitting
any of the fast path conditions is a direct return, so we can
just have 3 standalone branches rather than 'else if' stuff.

Signed-Off-By: David Anderson <danderson@tailscale.com>
5 years ago
David Anderson 05a52746a4 wgengine/magicsock: fix destination selection logic to work with DERP.
The effect is subtle: when we're not spraying packets, and have not yet
figured out a curAddr, and we're not spraying, we end up sending to
whatever the first IP is in the iteration order. In English, that
means "when we have no idea where to send packets, and we've given
up on sending to everyone, just send to the first addr we see in
the list."

This is, in general, what we want, because the addrs are in sorted
preference order, low to high, and DERP is the least preferred
destination. So, when we have no idea where to send, send to DERP,
right?

... Except for very historical reasons, appendDests iterated through
addresses in _reverse_ order, most preferred to least preferred.
crawshaw@ believes this was part of the earliest handshaking
algorithm magicsock had, where it slowly iterated through possible
destinations and poked handshakes to them one at a time.

Anyway, because of this historical reverse iteration, in the case
described above of "we have no idea where to send", the code would
end up sending to the _most_ preferred candidate address, rather
than the _least_ preferred. So when in doubt, we'd end up firing
packets into the blackhole of some LAN address that doesn't work,
and connectivity would not work.

This case only comes up if all your non-DERP connectivity options
have failed, so we more or less failed to detect it because we
didn't have a pathological test box deployed. Worse, codependent
bug 2839854994 made DERP accidentally
work sometimes anyway by incorrectly exploiting roamAddr behavior,
albeit at the cost of making DERP traffic symmetric. In fixing
DERP to once again be asymmetric, we effectively removed the
bandaid that was concealing this bug.

Signed-Off-By: David Anderson <danderson@tailscale.com>
5 years ago
David Anderson 97e58ad44d wgengine/magicsock: only set addrByKey once in CreateEndpoint.
Signed-Off-By: David Anderson <danderson@tailscale.com>
5 years ago
Brad Fitzpatrick fbab12c94c wgengine/magicsock: skip netcheck if external STUN aren't in use
Updates #146 (not a complete fix yet probably)
5 years ago
Brad Fitzpatrick fe0051fafd wgengine/magicsock: expand AddrSet.addrs comment 5 years ago
David Anderson 2839854994 wgengine/magicsock: never set a DERP server as a roamAddr.
DERP traffic is asymmetric by design, with nodes always sending
to their peer's home DERP server. However, if roamAddr is set,
magicsock will always push data there, rather than let DERP
server selection do its thing, so we end up accidentally
creating a symmetric flow.

Signed-Off-By: David Anderson <danderson@tailscale.com>
5 years ago
David Anderson 4f5c0da1ae wgengine/magicsock: log when home DERP server changes. 5 years ago
Brad Fitzpatrick 6978b93bdd derp, magicsock: track home (preferred) vs visiting connections for stats 5 years ago
Brad Fitzpatrick 12b77f30ad wgengine/magicsock: close stale DERP connections 5 years ago
Brad Fitzpatrick 2cff9016e4 net/dnscache: add overly simplistic DNS cache package for selective use
I started to write a full DNS caching resolver and I realized it was
overkill and wouldn't work on Windows even in Go 1.14 yet, so I'm
doing this tiny one instead for now, just for all our netcheck STUN
derp lookups, and connections to DERP servers. (This will be caching a
exactly 8 DNS entries, all ours.)

Fixes #145 (can be better later, of course)
5 years ago
Brad Fitzpatrick a36ccb8525 wgengine/magicsock: actually add to the activeDerp map
Fixes bug just introduced in 8f9849c140; not tested enough :(
5 years ago
Brad Fitzpatrick 8f9849c140 wgengine/magicsock: collapse three DERP maps down into one 5 years ago
Brad Fitzpatrick 40ebba1373 magicsock: use [unexpected] convention more
Fixes #136 (not entirely, but we have a convention now)
5 years ago
Brad Fitzpatrick 848a2bddf0 wgengine/magicsock: update set of DERP nodes 5 years ago
David Crawshaw 7932481b95 magicsock: lookup AddrSet by key from DERP
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
5 years ago
Brad Fitzpatrick eac62ec5ff ipn, wgengine/magicsock: add ipn.Prefs.DisableDERP bool
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick bf704a5218 derp: protocol negotiation, add v2: send src pub keys to clients in packets
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
David Crawshaw a33419167b magicsock: plumb through derpTLSConfig variable (for testing)
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
5 years ago
David Crawshaw 34859f8e7d wgengine, magicsock: add a CreateBind method
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
5 years ago
Brad Fitzpatrick b27d4c017a magicsock, wgengine, ipn, controlclient: plumb regular netchecks to map poll
Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick 724c37fb41 wgengine/magicsock: start tracking nearest DERP node 5 years ago
Brad Fitzpatrick 4675c70464 wgengine/magicsock: check STUN regularly 5 years ago
Brad Fitzpatrick bc7bc43fb8 magicsock, interfaces: move some code from magicsock to interfaces 5 years ago
Brad Fitzpatrick af7a01d6f0 wgengine/magicsock: drop donec channel, rename epUpdateCtx to serve its purpose 5 years ago
David Crawshaw cc4afa775f magicsock: rate limit send error log messages
The x/time/rate dependency adds 24kb to tailscaled binary size.

Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
5 years ago
David Crawshaw 0752c77dc2 magicsock: keep DERP magic IPs out of the address map
Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
5 years ago
David Crawshaw c6550135d5 magicsock: remove the index from indexedAddrs
The value predates the introduction of AddrSet which replaces
the index by tracking curAddr directly.

Signed-off-by: David Crawshaw <crawshaw@tailscale.com>
5 years ago
Brad Fitzpatrick 1abf2da392 wgengine/magicsock: reset favorite address on handshakes
Updates #92 (not a complete fix; could be better/faster?)

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick 67ede8d6d2 wgengine, magicsock: fix SetPrivateKey data race
Updates #112

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago
Brad Fitzpatrick c185e6b4b0 stunner: support IPv6, add latency info to callbacks, use unique TxIDs per retry
And some more docs.

Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
5 years ago