cowboy.git/test, branch 2.8.0 Small, fast, modern HTTP server for Erlang/OTP. 204 and 304 responses must not include a body 2020-05-20T11:41:05+00:00 Loïc Hoguin essen@ninenines.eu 2020-05-20T11:41:05+00:00 39b2816255503910dc23e2fdf703ee63bbc8953e When calling cowboy_req:reply/4 with a body a crash will occur resulting in a 500 response. When calling cowboy_req:stream_reply/2,3 and then attempting to send a body a crash will occur. 
When calling cowboy_req:reply/4 with a body a crash will occur
resulting in a 500 response. When calling cowboy_req:stream_reply/2,3
and then attempting to send a body a crash will occur.
Increase the default max_keepalive HTTP option to 1000 2020-05-20T09:08:58+00:00 Loïc Hoguin essen@ninenines.eu 2020-05-20T09:08:58+00:00 8337aca4d31ffa8120704057675aacbbaa280975 100 is very low for current deployments. 1000 is more appropriate as a default value. 
100 is very low for current deployments. 1000 is more
appropriate as a default value.
Don't stop listeners that don't exist in static_handler_SUITE 2020-05-20T09:08:24+00:00 Loïc Hoguin essen@ninenines.eu 2020-05-20T09:08:24+00:00 4edc39b0030ad09098283cf9aac8182a1d5cba26 






Increase the timetrap timeout in req_SUITE on Windows
2020-04-08T08:28:27+00:00

Loïc Hoguin
essen@ninenines.eu

2020-04-08T08:28:27+00:00

0d0e7d164c72799b48275c0c79a7a9d5c08e652e

Still seeing intermittent failures on Windows due to
timetrap so let's see if doubling the timeout again helps.





Still seeing intermittent failures on Windows due to
timetrap so let's see if doubling the timeout again helps.







Get rid of timeouts in metrics_SUITE
2020-04-08T08:28:11+00:00

Loïc Hoguin
essen@ninenines.eu

2020-04-08T08:28:11+00:00

f79d60e0c77122479a1c7b711bafb8f0d2c39e24

Again to avoid intermittent issues.





Again to avoid intermittent issues.







Allow {error, enotconn} as test success in security_SUITE
2020-04-07T08:18:29+00:00

Loïc Hoguin
essen@ninenines.eu

2020-04-07T08:18:29+00:00

b33971346116640013eeacc24ed273a2a5eb1c06

This happens from time to time on Windows and is partially
due to how the test is written.





This happens from time to time on Windows and is partially
due to how the test is written.







Rely on timetrap timeouts for tracer_SUITE
2020-04-06T18:14:31+00:00

Loïc Hoguin
essen@ninenines.eu

2020-04-06T18:14:31+00:00

e51ac66728a583e0db7da0aa958959573d65763c












NO_PARALLEL=1 will disable (most) parallel testing
2020-04-06T14:39:28+00:00

Loïc Hoguin
essen@ninenines.eu

2020-04-06T14:39:28+00:00

8cba8c3c68946b297d7cad63303677ab1da028f9

This is to make up for very slow or low resource environments,
like the FreeBSD CI environment currently.





This is to make up for very slow or low resource environments,
like the FreeBSD CI environment currently.







Increase the req_SUITE timetrap timeout on Windows
2020-04-06T10:27:41+00:00

Loïc Hoguin
essen@ninenines.eu

2020-04-06T10:27:41+00:00

5203ee6a847066be0a71cef0a7795f5617434ba3

The local network is much slower so the tests take longer there.





The local network is much slower so the tests take longer there.







Reduce the pipeline test body sizes to avoid filling send buffers
2020-04-05T14:28:54+00:00

Loïc Hoguin
essen@ninenines.eu

2020-04-05T14:28:54+00:00

cc8a2e4257da2a2d2a3e3ef97d4f6e2ecfdace12

This causes the test to get stuck because both the client and
server end up waiting in Transport:send for the other side to
read data from the socket.





This causes the test to get stuck because both the client and
server end up waiting in Transport:send for the other side to
read data from the socket.