Stream: git-wasmtime

Topic: wasmtime / issue #12895 Spurious networking-related failures

Wasmtime GitHub notifications bot (Mar 30 2026 at 18:08):

alexcrichton opened issue #12895:

Apparently github actions is having... uh... a day. Either that our CI all of a sudden got real flaky. A whole bunch of flaky networking failures are happening:

1. p2_udp_connect.rs address-in-use failure on Windows
2. p2_ip_name_lookup name resolution failure on macOS
3. p3_sockets_tcp_stream on macOS didn't get connection broken

cc @badeend

Wasmtime GitHub notifications bot (Mar 30 2026 at 18:08):

alexcrichton added the ci label to Issue #12895.

Wasmtime GitHub notifications bot (Mar 30 2026 at 18:33):

alexcrichton edited issue #12895:

Apparently github actions is having... uh... a day. Either that our CI all of a sudden got real flaky. A whole bunch of flaky networking failures are happening:

1. p2_udp_connect.rs address-in-use failure on Windows
2. p2_ip_name_lookup name resolution failure on macOS

3. p3_sockets_tcp_stream on macOS didn't get connection broken

   • another

cc @badeend

Wasmtime GitHub notifications bot (Mar 30 2026 at 18:33):

alexcrichton edited issue #12895:

Apparently github actions is having... uh... a day. Either that our CI all of a sudden got real flaky. A whole bunch of flaky networking failures are happening:

1. p2_udp_connect.rs address-in-use failure on Windows
2. p2_ip_name_lookup name resolution failure on macOS

3. p3_sockets_tcp_stream on macOS didn't get connection broken

   • another

cc @badeend

Wasmtime GitHub notifications bot (Mar 30 2026 at 18:34):

alexcrichton edited issue #12895:

Apparently github actions is having... uh... a day. Either that our CI all of a sudden got real flaky. A whole bunch of flaky networking failures are happening:

1. p2_udp_connect.rs address-in-use failure on Windows
2. p2_ip_name_lookup name resolution failure on macOS

3. p3_sockets_tcp_stream on macOS didn't get connection broken

   • another
   • another

cc @badeend

Wasmtime GitHub notifications bot (Mar 30 2026 at 19:59):

badeend commented on issue #12895:

• For the p3_sockets_tcp_stream failure, let's see if #12890 can shine some more light on this.
• The p2_ip_name_lookup test performs external requests, so _some_ degree of flaky'ness is expected. I've opened #12898 to reduce the likelihood.
• The p2_udp_connect test fails when binding a socket on localhost on port 0. This should never fail, unless the host is truly out of ports.

We could:

• chalk it up to github being github again and hope things will resolve themselves soon.
• continue the game of whack-a-mole and update the p2_udp_connect test retry the bind if it returns EADDRINUSE
• mark all networking tests with some kind of "flaky_in_ci" attribute and demote errors to warnings when running with the CI environment variable. That way, nothing would be hard blocked by these failures. But unless someone is actively monitoring the logs, it also allows code to be silently become permanently broken.

not super excited by any of these.

Wasmtime GitHub notifications bot (Mar 31 2026 at 18:53):

alexcrichton closed issue #12895:

Apparently github actions is having... uh... a day. Either that our CI all of a sudden got real flaky. A whole bunch of flaky networking failures are happening:

1. p2_udp_connect.rs address-in-use failure on Windows
2. p2_ip_name_lookup name resolution failure on macOS

3. p3_sockets_tcp_stream on macOS didn't get connection broken

   • another
   • another

cc @badeend

Wasmtime GitHub notifications bot (Mar 31 2026 at 18:53):

alexcrichton commented on issue #12895:

Yeah I think we've probably done as best as we can here. Everything ceased after the morning so I suppose let's just hope that github actions behaves a bit better over time...

Wasmtime GitHub notifications bot (Mar 31 2026 at 19:02):

badeend commented on issue #12895:

¯\\_\(ツ)\_/¯

Last updated: Apr 12 2026 at 23:10 UTC