summaryrefslogtreecommitdiffhomepage
path: root/tsnet
AgeCommit message (Collapse)AuthorFilesLines
2025-10-16wgengine/netlog: merge connstats into package (#17557)Joe Tsai1-2/+1
Merge the connstats package into the netlog package and unexport all of its declarations. Remove the buildfeatures.HasConnStats and use HasNetLog instead. Updates tailscale/corp#33352 Signed-off-by: Joe Tsai <joetsai@digital-static.net>
2025-10-15net/connstats: prepare to remove package (#17554)Joe Tsai1-1/+2
The connstats package was an unnecessary layer of indirection. It was seperated out of wgengine/netlog so that net/tstun and wgengine/magicsock wouldn't need a depenedency on the concrete implementation of network flow logging. Instead, we simply register a callback for counting connections. This PR does the bare minimum work to prepare tstun and magicsock to only care about that callback. A future PR will delete connstats and merge it into netlog. Updates tailscale/corp#33352 Signed-off-by: Joe Tsai <joetsai@digital-static.net>
2025-10-08go.toolchain.rev: bump Go to 1.25.2Brad Fitzpatrick1-1/+1
Updates tailscale/go#135 Change-Id: I89cfb49b998b2fd0264f8d5f4a61af839cd06626 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-07feature/linkspeed: move cosmetic tstun netlink code out to modular featureBrad Fitzpatrick1-2/+1
Part of making all netlink monitoring code optional. Updates #17311 (how I got started down this path) Updates #12614 Change-Id: Ic80d8a7a44dc261c4b8678b3c2241c3b3778370d Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-06util/checkchange: stop using deephash everywhereBrad Fitzpatrick1-1/+2
Saves 45 KB from the min build, no longer pulling in deephash or util/hashx, both with unsafe code. It can actually be more efficient to not use deephash, as you don't have to walk all bytes of all fields recursively to answer that two things are not equal. Instead, you can just return false at the first difference you see. And then with views (as we use ~everywhere nowadays), the cloning the old value isn't expensive, as it's just a pointer under the hood. Updates #12614 Change-Id: I7b08616b8a09b3ade454bb5e0ac5672086fe8aec Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-06feature/condlite/expvar: add expvar stub package when metrics not neededBrad Fitzpatrick1-0/+1
Saves ~53 KB from the min build. Updates #12614 Change-Id: I73f9544a9feea06027c6ebdd222d712ada851299 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-06feature/featuretags: make usermetrics modularBrad Fitzpatrick1-1/+1
Saves ~102 KB from the min build. Updates #12614 Change-Id: Ie1d4f439321267b9f98046593cb289ee3c4d6249 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-04wgengine: stop importing flowtrack when unusedBrad Fitzpatrick1-1/+1
Updates #12614 Change-Id: I42b5c4d623d356af4bee5bbdabaaf0f6822f2bf4 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-04feature/c2n: move answerC2N code + deps out of control/controlclientBrad Fitzpatrick2-2/+4
c2n was already a conditional feature, but it didn't have a feature/c2n directory before (rather, it was using consts + DCE). This adds it, and moves some code, which removes the httprec dependency. Also, remove some unnecessary code from our httprec fork. Updates #12614 Change-Id: I2fbe538e09794c517038e35a694a363312c426a2 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-03feature/ace: make ACE modularBrad Fitzpatrick1-1/+0
Updates #12614 Change-Id: Iaee75d8831c4ba5c9705d7877bb78044424c6da1 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-03health: make it omittableBrad Fitzpatrick1-1/+1
Saves 86 KB. And stop depending on expvar and usermetrics when disabled, in prep to removing all the expvar/metrics/tsweb stuff. Updates #12614 Change-Id: I35d2479ddd1d39b615bab32b1fa940ae8cbf9b11 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-02appc,*: publish events for route updates and storage (#17392)M. J. Fromberger1-2/+2
Add and wire up event publishers for these two event types in the AppConnector. Nothing currently subscribes to them, so this is harmless. Subscribers for these events will be added in a near-future commit. As part of this, move the appc.RouteInfo type to the types/appctype package. It does not contain any package-specific details from appc. Beside it, add appctype.RouteUpdate to carry route update event state, likewise not specific to appc. Update all usage of the appc.* types throughout to use appctype.* instead, and update depaware files to reflect these changes. Add a Close method to the AppConnector to make sure the client gets cleaned up when the connector is dropped (we re-create connectors). Update the unit tests in the appc package to also check the events published alongside calls to the RouteAdvertiser. For now the tests still rely on the RouteAdvertiser for correctness; this is OK for now as the two methods are always performed together. In the near future, we need to rework the tests so not require that, but that will require building some more test fixtures that we can handle separately. Updates #15160 Updates #17192 Change-Id: I184670ba2fb920e0d2cb2be7c6816259bca77afe Signed-off-by: M. J. Fromberger <fromberger@tailscale.com>
2025-10-02control/controlclient: remove x/net/http2, use net/httpBrad Fitzpatrick1-6/+3
Saves 352 KB, removing one of our two HTTP/2 implementations linked into the binary. Fixes #17305 Updates #15015 Change-Id: I53a04b1f2687dca73c8541949465038b69aa6ade Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-02feature/featuretags, all: add build features, use existing ones in more placesBrad Fitzpatrick1-1/+1
Saves 270 KB. Updates #12614 Change-Id: I4c3fe06d32c49edb3a4bb0758a8617d83f291cf5 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-01control/ts2021: rename from internal/noiseconn in prep for controlclient splitBrad Fitzpatrick1-1/+1
A following change will split out the controlclient.NoiseClient type out, away from the rest of the controlclient package which is relatively dependency heavy. A question was where to move it, and whether to make a new (a fifth!) package in the ts2021 dependency chain. @creachadair and I brainstormed and decided to merge internal/noiseconn and controlclient.NoiseClient into one package, with names ts2021.Conn and ts2021.Client. For ease of reviewing the subsequent PR, this is the first step that just renames the internal/noiseconn package to control/ts2021. Updates #17305 Change-Id: Ib5ea162dc1d336c1d805bdd9548d1702dd6e1468 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-01Makefile, cmd/*/depaware.txt: split out vendor packages explicitlyBrad Fitzpatrick1-5/+19
depaware was merging golang.org/x/foo and std's vendor/golang.org/x/foo packages (which could both be in the binary!), leading to confusing output, especially when I was working on eliminating duplicate packages imported under different names. This makes the depaware output longer and grosser, but doesn't hide reality from us. Updates #17305 Change-Id: I21cc3418014e127f6c1a81caf4e84213ce84ab57 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-01net/netmon: remove usage of direct callbacks from netmon (#17292)Claus Lensbøl1-0/+2
The callback itself is not removed as it is used in other repos, making it simpler for those to slowly transition to the eventbus. Updates #15160 Signed-off-by: Claus Lensbøl <claus@tailscale.com>
2025-10-01tsnet: remove AuthenticatedAPITransport (API-over-noise) supportBrad Fitzpatrick1-35/+0
It never launched and I've lost hope of it launching and it's in my way now, so I guess it's time to say goodbye. Updates tailscale/corp#4383 Updates #17305 Change-Id: I2eb551d49f2fb062979cc307f284df4b3dfa5956 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-10-01all: use Go 1.20's errors.Join instead of our multierr packageBrad Fitzpatrick1-1/+0
Updates #7123 Change-Id: Ie9be6814831f661ad5636afcd51d063a0d7a907d Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-30cmd/tailscaled, ipn/localapi, util/eventbus: don't link in regexp when debug ↵Brad Fitzpatrick1-4/+4
is omitted Saves 442 KB. Lock it with a new min test. Updates #12614 Change-Id: Ia7bf6f797b6cbf08ea65419ade2f359d390f8e91 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-30feature, net/tshttpproxy: pull out support for using proxies as a featureBrad Fitzpatrick2-1/+4
Saves 139 KB. Also Synology support, which I saw had its own large-ish proxy parsing support on Linux, but support for proxies without Synology proxy support is reasonable, so I pulled that out as its own thing. Updates #12614 Change-Id: I22de285a3def7be77fdcf23e2bec7c83c9655593 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-30ipn/ipnlocal: drop h2c package, use net/http's supportBrad Fitzpatrick1-2/+1
In Dec 2021 in d3d503d9977ed I had grand plans to make exit node DNS cheaper by using HTTP/2 over PeerAPI, at least on some platforms. I only did server-side support though and never made it to the client. In the ~4 years since, some things have happened: * Go 1.24 got support for http.Protocols (https://pkg.go.dev/net/http#Protocols) and doing UnencryptedHTTP2 ("HTTP2 with prior knowledge") * The old h2c upgrade mechanism was deprecated; see https://github.com/golang/go/issues/63565 and https://github.com/golang/go/issues/67816 * Go plans to deprecate x/net/http2 and move everything to the standard library. So this drops our use of the x/net/http2/h2c package and instead enables h2c (on all platforms now) using the standard library. This does mean we lose the deprecated h2c Upgrade support, but that's fine. If/when we do the h2c client support for ExitDNS, we'll have to probe the peer to see whether it supports it. Or have it reply with a header saying that future requests can us h2c. (It's tempting to use capver, but maybe people will disable that support anyway, so we should discover it at runtime instead.) Also do the same in the sessionrecording package. Updates #17305 Change-Id: If323f5ef32486effb18ed836888aa05c0efb701e Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-29ipn/ipnlocal, feature/posture: pull posture out into a modular featureBrad Fitzpatrick1-3/+1
Updates #12614 Change-Id: I9d08a1330b9c55e1a23e7979a707e11d8e090d79 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-29feature/clientupdate: move clientupdate to a modular feature, disabled for tsnetBrad Fitzpatrick1-35/+32
Updates #12614 Change-Id: I5f685dec84a5396b7c2b66f2788ae3d286e1ddc6 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-29ipn/store, feature/condregister: move AWS + Kube store registration to ↵Brad Fitzpatrick1-91/+11
condregister Otherwise they're uselessly imported by tsnet applications, even though they do nothing. tsnet applications wanting to use these already had to explicitly import them and use kubestore.New or awsstore.New and assign those to their tsnet.Server.Store fields. Updates #12614 Change-Id: I358e3923686ddf43a85e6923c3828ba2198991d4 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-29wgengine/router{,/osrouter}: split OS router implementations into subpackageBrad Fitzpatrick1-17/+2
So wgengine/router is just the docs + entrypoint + types, and then underscore importing wgengine/router/osrouter registers the constructors with the wgengine/router package. Then tsnet can not pull those in. Updates #17313 Change-Id: If313226f6987d709ea9193c8f16a909326ceefe7 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-29cmd/tailscale/cli,client,ipn: add appc-routes cli commandFran Bull1-1/+1
Allow the user to access information about routes an app connector has learned, such as how many routes for each domain. Fixes tailscale/corp#32624 Signed-off-by: Fran Bull <fran@tailscale.com>
2025-09-29feature/sdnotify: move util/systemd to a modular featureBrad Fitzpatrick1-2/+0
Updates #12614 Change-Id: I08e714c83b455df7f538cc99cafe940db936b480 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-28util/backoff: rename logtail/backoff package to util/backoffBrad Fitzpatrick1-1/+1
It has nothing to do with logtail and is confusing named like that. Updates #cleanup Updates #17323 Change-Id: Idd34587ba186a2416725f72ffc4c5778b0b9db4a Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-28tsconst, util/linuxfw, wgengine/router: move Linux fw consts to tsconstBrad Fitzpatrick1-1/+1
Now cmd/derper doesn't depend on iptables, nftables, and netlink code :) But this is really just a cleanup step I noticed on the way to making tsnet applications able to not link all the OS router code which they don't use. Updates #17313 Change-Id: Ic7b4e04e3a9639fd198e9dbeb0f7bae22a4a47a9 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-26doctor: add ts_omit_doctor supportBrad Fitzpatrick1-7/+2
Updates #12614 Change-Id: I84c166c4b99ca75d70abe4087e5ff3f7d90d4bcc Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-25feature/portlist: pull portlist service porting into extension, use eventbusBrad Fitzpatrick2-4/+14
And yay: tsnet (and thus k8s-operator etc) no longer depends on portlist! And LocalBackend is smaller. Removes 50 KB from the minimal binary. Updates #12614 Change-Id: Iee04057053dc39305303e8bd1d9599db8368d926 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-24derp/derpserver: split off derp.Server out of derp into its own packageBrad Fitzpatrick1-6/+5
This exports a number of things from the derp (generic + client) package to be used by the new derpserver package, as now used by cmd/derper. And then enough other misc changes to lock in that cmd/tailscaled can be configured to not bring in tailscale.com/client/local. (The webclient in particular, even when disabled, was bringing it in, so that's now fixed) Fixes #17257 Change-Id: I88b6c7958643fb54f386dd900bddf73d2d4d96d5 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-23client, cmd/tailscale/cli, feature/relayserver, net/udprelay: implement ↵Jordan Whited1-0/+1
tailscale debug peer-relay-sessions (#17239) Fixes tailscale/corp#30035 Signed-off-by: Dylan Bargatze <dylan@tailscale.com> Signed-off-by: Jordan Whited <jordan@tailscale.com> Co-authored-by: Dylan Bargatze <dylan@tailscale.com>
2025-09-22cmd/k8s-operator: send operator logs to tailscale (#17110)David Bond1-4/+4
This commit modifies the k8s operator to wrap its logger using the logtail logger provided via the tsnet server. This causes any logs written by the operator to make their way to Tailscale in the same fashion as wireguard logs to be used by support. This functionality can also be opted-out of entirely using the "TS_NO_LOGS_NO_SUPPORT" environment variable. Updates https://github.com/tailscale/corp/issues/32037 Signed-off-by: David Bond <davidsbond93@gmail.com>
2025-09-19feature/linuxdnsfight: move inotify watching of /etc/resolv.conf out to a ↵Brad Fitzpatrick1-2/+0
feature tsnet apps in particular never use the Linux DNS OSManagers, so they don't need DBus, etc. I started to pull that all out into separate features so tsnet doesn't need to bring in DBus, but hit this first. Here you can see that tsnet (and the k8s-operator) no longer pulls in inotify. Updates #17206 Change-Id: I7af0f391f60c5e7dbeed7a080346f83262346591 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-19net/ace, control/controlhttp: start adding ACE dialing supportBrad Fitzpatrick1-0/+1
Updates tailscale/corp#32227 Change-Id: I38afc668f99eb1d6f7632e82554b82922f3ebb9f Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-19tsnet,internal/client/tailscale: resolve OAuth into authkeys in tsnet (#17191)Naman Sood2-0/+17
* tsnet,internal/client/tailscale: resolve OAuth into authkeys in tsnet Updates #8403. * internal/client/tailscale: omit OAuth library via build tag Updates #12614. Signed-off-by: Naman Sood <mail@nsood.in>
2025-09-17net/dns/recursive: remove recursive DNS resolverBrad Fitzpatrick1-4/+2
It doesn't really pull its weight: it adds 577 KB to the binary and is rarely useful. Also, we now have static IPs and other connectivity paths coming soon enough. Updates #5853 Updates #1278 Updates tailscale/corp#32168 Change-Id: If336fed00a9c9ae9745419e6d81f7de6da6f7275 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-16feature/portmapper: make the portmapper & its debugging tools modularBrad Fitzpatrick2-2/+6
Starting at a minimal binary and adding one feature back... tailscaled tailscale combined (linux/amd64) 30073135 17451704 31543692 omitting everything + 480302 + 10258 + 493896 .. add debugportmapper + 475317 + 151943 + 467660 .. add portmapper + 500086 + 162873 + 510511 .. add portmapper+debugportmapper Fixes #17148 Change-Id: I90bd0e9d1bd8cbe64fa2e885e9afef8fb5ee74b1 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-16health,ipn/ipnlocal: introduce eventbus in heath.Tracker (#17085)Claus Lensbøl1-3/+3
The Tracker was using direct callbacks to ipnlocal. This PR moves those to be triggered via the eventbus. Additionally, the eventbus is now closed on exit from tailscaled explicitly, and health is now a SubSystem in tsd. Updates #15160 Signed-off-by: Claus Lensbøl <claus@tailscale.com>
2025-09-16all: add ts_omit_serve, start making tailscale serve/funnel be modularBrad Fitzpatrick1-0/+1
tailscaled tailscale combined (linux/amd64) 29853147 17384418 31412596 omitting everything + 621570 + 219277 + 554256 .. add serve Updates #17128 Change-Id: I87c2c6c3d3fc2dc026c3de8ef7000a813b41d31c Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-12tsnet: expose logtail's Logf method (#17057)David Bond1-0/+10
This commit adds a new method to the tsnet.Server type named `Logger` that returns the underlying logtail instance's Logf method. This is intended to be used within the Kubernetes operator to wrap its existing logger in a way such that operator specific logs can also be sent to control for support & debugging purposes. Updates https://github.com/tailscale/corp/issues/32037 Signed-off-by: David Bond <davidsbond93@gmail.com>
2025-09-09tsnet: remove APIClient() which is deprecated and now unused (#17073)Alex Chan2-24/+3
Updates tailscale/corp#22748 Signed-off-by: Alex Chan <alexc@tailscale.com>
2025-09-08cmd/tailscale/cli, derp: use client/local instead of deprecated ↵Alex Chan1-1/+1
client/tailscale (#17061) * cmd/tailscale/cli: use client/local instead of deprecated client/tailscale Updates tailscale/corp#22748 Signed-off-by: Alex Chan <alexc@tailscale.com> * derp: use client/local instead of deprecated client/tailscale Updates tailscale/corp#22748 Signed-off-by: Alex Chan <alexc@tailscale.com> --------- Signed-off-by: Alex Chan <alexc@tailscale.com>
2025-09-02util/syspolicy: finish plumbing policyclient, add feature/syspolicy, move ↵Brad Fitzpatrick1-1/+2
global impl This is step 4 of making syspolicy a build-time feature. This adds a policyclient.Get() accessor to return the correct implementation to use: either the real one, or the no-op one. (A third type, a static one for testing, also exists, so in general a policyclient.Client should be plumbed around and not always fetched via policyclient.Get whenever possible, especially if tests need to use alternate syspolicy) Updates #16998 Updates #12614 Change-Id: Iaf19670744a596d5918acfa744f5db4564272978 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-02util/syspolicy/{setting,ptype}: move PreferenceOption and Visibility to new ↵Brad Fitzpatrick1-0/+1
leaf package Step 3 in the series. See earlier cc532efc2000 and d05e6dc09e. This step moves some types into a new leaf "ptype" package out of the big "settings" package. The policyclient.Client will later get new methods to return those things (as well as Duration and Uint64, which weren't done at the time of the earlier prototype). Updates #16998 Updates #12614 Change-Id: I4d72d8079de3b5351ed602eaa72863372bd474a2 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-09-01util/syspolicy/policyclient: add policyclient.Client interface, start plumbingBrad Fitzpatrick1-2/+3
This is step 2 of ~4, breaking up #14720 into reviewable chunks, with the aim to make syspolicy be a build-time configurable feature. Step 1 was #16984. In this second step, the util/syspolicy/policyclient package is added with the policyclient.Client interface. This is the interface that's always present (regardless of build tags), and is what code around the tree uses to ask syspolicy/MDM questions. There are two implementations of policyclient.Client for now: 1) NoPolicyClient, which only returns default values. 2) the unexported, temporary 'globalSyspolicy', which is implemented in terms of the global functions we wish to later eliminate. This then starts to plumb around the policyclient.Client to most callers. Future changes will plumb it more. When the last of the global func callers are gone, then we can unexport the global functions and make a proper policyclient.Client type and constructor in the syspolicy package, removing the globalSyspolicy impl out of tsd. The final change will sprinkle build tags in a few more places and lock it in with dependency tests to make sure the dependencies don't later creep back in. Updates #16998 Updates #12614 Change-Id: Ib2c93d15c15c1f2b981464099177cd492d50391c Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-08-31util/syspolicy/*: move syspolicy keys to new const leaf "pkey" packageBrad Fitzpatrick1-0/+1
This is step 1 of ~3, breaking up #14720 into reviewable chunks, with the aim to make syspolicy be a build-time configurable feature. In this first (very noisy) step, all the syspolicy string key constants move to a new constant-only (code-free) package. This will make future steps more reviewable, without this movement noise. There are no code or behavior changes here. The future steps of this series can be seen in #14720: removing global funcs from syspolicy resolution and using an interface that's plumbed around instead. Then adding build tags. Updates #12614 Change-Id: If73bf2c28b9c9b1a408fe868b0b6a25b03eeabd1 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>
2025-08-28wgengine/magicsock: shorten process internal DERP queueJames Tucker1-1/+0
DERP writes go via TCP and the host OS will have plenty of buffer space. We've observed in the wild with a backed up TCP socket kernel side buffers of >2.4MB. The DERP internal queue being larger causes an increase in the probability that the contents of the backbuffer are "dead letters" - packets that were assumed to be lost. A first step to improvement is to size this queue only large enough to avoid some of the initial connect stall problem, but not large enough that it is contributing in a substantial way to buffer bloat / dead-letter retention. Updates tailscale/corp#31762 Signed-off-by: James Tucker <james@tailscale.com>
Copyright © 2025 - 2026 Wayne Cole. WTFPL. Attribution appreciated. No warranty. Not liable for bodily damages.