net/http: TestTransportCloseResponseBody failures

[gopherbot]

#!watchflakes
post <- pkg == "net/http" && test == "TestTransportCloseResponseBody" && `too long waiting`

Issue created automatically to collect these failures.

Example (log):

1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
2023/01/26 01:38:34 httptest.Server blocked in Close after 5 seconds, waiting for connections:
  *tls.Conn 0xc000666000 127.0.0.1:50229 in state active
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
--- FAIL: TestTransportCloseResponseBody (0.00s)
    --- FAIL: TestTransportCloseResponseBody/h1 (19.43s)
        transport_test.go:2752: too long waiting for write error

— watchflakes

[gopherbot]

Found new dashboard test flakes for:

#!watchflakes
post <- pkg == "net/http" && test == "TestTransportCloseResponseBody"

2023-01-25 19:25 darwin-amd64-12_0 go@2bf0f54b net/http.TestTransportCloseResponseBody (log)

1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
2023/01/26 01:38:34 httptest.Server blocked in Close after 5 seconds, waiting for connections:
  *tls.Conn 0xc000666000 127.0.0.1:50229 in state active
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
1 second passes in backend, proxygone= false
--- FAIL: TestTransportCloseResponseBody (0.00s)
    --- FAIL: TestTransportCloseResponseBody/h1 (19.43s)
        transport_test.go:2752: too long waiting for write error

2023-01-30 17:48 darwin-amd64-12_0 go@4373c6f5 net/http.TestTransportCloseResponseBody (log)

--- FAIL: TestTransportCloseResponseBody (0.00s)
    --- FAIL: TestTransportCloseResponseBody/h2 (15.96s)
        transport_test.go:2744: too long waiting for close

2023-02-01 19:47 linux-amd64-race go@ab0f0459 net/http.TestTransportCloseResponseBody (log)

--- FAIL: TestTransportCloseResponseBody (0.00s)

— watchflakes

[bcmills]

linux-amd64-race is #58168; the others are probably in the nebulous darwin cluster (attn @neild).

[gopherbot]

Found new dashboard test flakes for:

#!watchflakes
post <- pkg == "net/http" && test == "TestTransportCloseResponseBody" && `too long waiting`

2023-02-10 17:29 darwin-amd64-13 go@1fa2deb1 net/http.TestTransportCloseResponseBody (log)

--- FAIL: TestTransportCloseResponseBody (0.00s)
    --- FAIL: TestTransportCloseResponseBody/h2 (12.59s)
        transport_test.go:2752: too long waiting for write error

2023-02-14 18:12 darwin-amd64-10_15 go@202a1a57 net/http.TestTransportCloseResponseBody (log)

--- FAIL: TestTransportCloseResponseBody (0.00s)
    --- FAIL: TestTransportCloseResponseBody/h1 (11.53s)
        transport_test.go:2752: too long waiting for write error
2023/02/14 15:12:41 httptest.Server blocked in Close after 5 seconds, waiting for connections:
  *net.TCPConn 0xc001164170 127.0.0.1:50922 in state active

— watchflakes

[gopherbot]

Found new dashboard test flakes for:

#!watchflakes
post <- pkg == "net/http" && test == "TestTransportCloseResponseBody" && `too long waiting`

2023-02-27 20:25 darwin-amd64-10_15 go@b611b3a8 net/http.TestTransportCloseResponseBody (log)

2023/03/01 06:57:33 httptest.Server blocked in Close after 5 seconds, waiting for connections:
  *tls.Conn 0xc000024700 127.0.0.1:50577 in state active
--- FAIL: TestTransportCloseResponseBody (0.00s)
    --- FAIL: TestTransportCloseResponseBody/h1 (16.22s)
        transport_test.go:2752: too long waiting for write error

— watchflakes

[gopherbot]

Found new dashboard test flakes for:

#!watchflakes
post <- pkg == "net/http" && test == "TestTransportCloseResponseBody" && `too long waiting`

2023-03-17 19:45 darwin-amd64-12_0 go@3ff6dbdf net/http.TestTransportCloseResponseBody (log)

2023/03/17 18:56:00 httptest.Server blocked in Close after 5 seconds, waiting for connections:
  *tls.Conn 0xc00031fc00 127.0.0.1:50398 in state active
--- FAIL: TestTransportCloseResponseBody (0.00s)
    --- FAIL: TestTransportCloseResponseBody/h1 (15.05s)
        transport_test.go:2752: too long waiting for write error

— watchflakes

[bcmills]

Fixed by CL 477196 on the main branch. (Note that the most recent linked failure is on release-branch.go1.20.)