tailscale: control/controlclient: failure to connect on Alaska Airlines wifi
bradfitz@book1pro tailscale.com % git rev-parse HEAD
5550a17391f1a8765f7f56831c759ee862b6b434
bradfitz@book1pro tailscale.com % ping 8.8.8.8
PING 8.8.8.8 (8.8.8.8): 56 data bytes
64 bytes from 8.8.8.8: icmp_seq=0 ttl=53 time=698.840 ms
64 bytes from 8.8.8.8: icmp_seq=1 ttl=53 time=726.506 ms
64 bytes from 8.8.8.8: icmp_seq=2 ttl=53 time=948.392 ms
Request timeout for icmp_seq 3
64 bytes from 8.8.8.8: icmp_seq=3 ttl=53 time=1069.618 ms
64 bytes from 8.8.8.8: icmp_seq=4 ttl=53 time=775.248 ms
64 bytes from 8.8.8.8: icmp_seq=5 ttl=53 time=712.453 ms
64 bytes from 8.8.8.8: icmp_seq=6 ttl=53 time=734.816 ms
64 bytes from 8.8.8.8: icmp_seq=7 ttl=53 time=702.859 ms
^C
--- 8.8.8.8 ping statistics ---
9 packets transmitted, 8 packets received, 11.1% packet loss
round-trip min/avg/max/stddev = 698.840/796.092/1069.618/128.544 ms
Fails to connect:
bradfitz@book1pro tailscale.com % go run ./cmd/tailscale debug ts2021
09:44:31.408636 Fetching keys from https://controlplane.tailscale.com/key?v=106 ...
09:44:33.810578 Dial("tcp", "[2a05:d014:386:202:c5e9:4c3f:9808:f2b0]:80") ...
09:44:33.811639 Dial("tcp", "[2a05:d014:386:202:c5e9:4c3f:9808:f2b0]:80") = dial tcp [2a05:d014:386:202:c5e9:4c3f:9808:f2b0]:80: connect: no route to host
09:44:33.811718 Dial("tcp", "18.199.123.246:80") ...
09:44:33.816331 Dial("tcp", "18.199.123.246:80") = 172.19.207.90:64618 / 18.199.123.246:80
09:44:34.311099 Dial("tcp", "[2a05:d014:386:202:b1ee:6525:7c95:d0f3]:443") ...
09:44:34.312192 Dial("tcp", "[2a05:d014:386:202:b1ee:6525:7c95:d0f3]:443") = dial tcp [2a05:d014:386:202:b1ee:6525:7c95:d0f3]:443: connect: no route to host
09:44:34.312254 Dial("tcp", "18.199.123.246:443") ...
09:44:34.315810 Dial("tcp", "18.199.123.246:443") = 172.19.207.90:64626 / 18.199.123.246:443
09:44:34.704893 controlhttp.Dial = 0x140000a0558, <nil>
09:44:34.705027 did noise handshake
09:44:34.705035 final underlying conn: 172.19.207.90:64618 / 18.199.123.246:80
RoundTrip whoami request: unexpected EOF
exit status 1
About this issue
- Original URL
- State: closed
- Created 2 years ago
- Comments: 20 (9 by maintainers)
Commits related to this issue
- control/controlhttp: fix connectivity on Alaska Air wifi Updates #13597 Change-Id: Ifbf52b93fd35d64fcf80f8fddbfd610008fd8742 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com> — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: test that a control conn is healthy before using We've seen a bunch of cases where a captive portal or network with a firewall will allow a connection to the control server, succ... — committed to tailscale/tailscale by andrew-d 2 years ago
- cmd/tailscale/cli: make 'tailscale debug ts2021' try twice In prep for a future port 80 MITM fix, make the 'debug ts2021' command retry once after a failure to give it a chance to pick a new strategy... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: factor out some code in prep for future change This pulls out the clock and forceNoise443 code into methods on the Dialer as cleanup in its own commit to make a future change les... — committed to tailscale/tailscale by bradfitz 2 years ago
- cmd/tailscale/cli: make 'tailscale debug ts2021' try twice In prep for a future port 80 MITM fix, make the 'debug ts2021' command retry once after a failure to give it a chance to pick a new strategy... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: factor out some code in prep for future change This pulls out the clock and forceNoise443 code into methods on the Dialer as cleanup in its own commit to make a future change les... — committed to tailscale/tailscale by bradfitz 2 years ago
- cmd/tailscale/cli: make 'tailscale debug ts2021' try twice In prep for a future port 80 MITM fix, make the 'debug ts2021' command retry once after a failure to give it a chance to pick a new strategy... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: factor out some code in prep for future change This pulls out the clock and forceNoise443 code into methods on the Dialer as cleanup in its own commit to make a future change les... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: fix connectivity on Alaska Air wifi Updates #13597 Change-Id: Ifbf52b93fd35d64fcf80f8fddbfd610008fd8742 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com> — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: fix connectivity on Alaska Air wifi Updates #13597 Change-Id: Ifbf52b93fd35d64fcf80f8fddbfd610008fd8742 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com> — committed to tailscale/tailscale by bradfitz 2 years ago
- net/dnscache: don't fall back to an IPv6 dial if we don't have IPv6 I noticed while debugging a test failure elsewhere that our failure logs (when verbosity is cranked up) were uselessly attributing ... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: rename a param from addr to optAddr for clarity And update docs. Updates #cleanup Updates #13597 (tangentially; noted this cleanup while debugging) Change-Id: I62440294c78b0bb3... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/control{client,http}: don't noise dial localhost:443 in http-only tests 1eaad7d3deb regressed some tests in another repo that were starting up a control server on `http://127.0.0.1:nnn`. Beca... — committed to tailscale/tailscale by bradfitz 2 years ago
- net/dnscache: don't fall back to an IPv6 dial if we don't have IPv6 I noticed while debugging a test failure elsewhere that our failure logs (when verbosity is cranked up) were uselessly attributing ... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: rename a param from addr to optAddr for clarity And update docs. Updates #cleanup Updates #13597 (tangentially; noted this cleanup while debugging) Change-Id: I62440294c78b0bb3... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/control{client,http}: don't noise dial localhost:443 in http-only tests 1eaad7d3deb regressed some tests in another repo that were starting up a control server on `http://127.0.0.1:nnn`. Beca... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/control{client,http}: don't noise dial localhost:443 in http-only tests 1eaad7d3deb regressed some tests in another repo that were starting up a control server on `http://127.0.0.1:nnn`. Beca... — committed to tailscale/tailscale by bradfitz 2 years ago
- net/dnscache: don't fall back to an IPv6 dial if we don't have IPv6 I noticed while debugging a test failure elsewhere that our failure logs (when verbosity is cranked up) were uselessly attributing ... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/controlhttp: rename a param from addr to optAddr for clarity And update docs. Updates #cleanup Updates #13597 (tangentially; noted this cleanup while debugging) Change-Id: I62440294c78b0bb3... — committed to tailscale/tailscale by bradfitz 2 years ago
- control/control{client,http}: don't noise dial localhost:443 in http-only tests 1eaad7d3deb regressed some tests in another repo that were starting up a control server on `http://127.0.0.1:nnn`. Beca... — committed to tailscale/tailscale by bradfitz 2 years ago
I’m currently on an Alaska flight. Here’s a screenshot showing the macOS client (dev build, not yet released) working:
@gw, @jeffmo, see the PR references above. We have two possible fixes outstanding for this now. And I’m considering a third. But we also want to add a regression test for this too before landing.
I have another Alaska flight home in about 15 minutes and I’ll continue working on this on the flight. It should be in the next release.
Just ran into this on an Alaska flight, wasn’t working on Mac App Store version 1.72.2.
Upgraded to 1.80.2 and everything is working great!