Default example is not working

[aleasims]

I'm running a default example in an empty directory:

sam init && sam run

And I get this:

 TEST  Testing Example Test Suite ...
 TEST    It should respond to HTTP requests...😭: Runtime error: Failed to get URL: error sending request for url (http://127.0.0.1:8080/hello.txt) (line 8, position 5) in call to function 'fetch' @ 'example' (line 6, position 33) in closure call (line 4, position 5) (514us 898ns)
 TEST  Testing Example Test Suite failed! 😭 (1 tests failed out of 1) (545us 8ns)

It looks like caddy container is destroyed immediately after creation:

$ podman ps -a
CONTAINER ID  IMAGE       COMMAND     CREATED     STATUS      PORTS       NAMES

I tried debugging by adding this line right after podman command execution:

let o = Command::new("podman").args(["logs", "caddy"]).output().await.unwrap();
println!("{}", String::from_utf8_lossy(&o.stdout));
println!("{}", String::from_utf8_lossy(&o.stderr));

and it became behaving completely unpredictable: sometimes the test succeeds, sometimes it fails with no obvious difference:

$ sam run

{"level":"warn","ts":1737733486.5284944,"logger":"admin","msg":"admin endpoint disabled"}
{"level":"warn","ts":1737733486.5285542,"logger":"http.auto_https","msg":"server is listening only on the HTTP port, so no automatic HTTPS will be applied to this server","server_name":"static","http_port":80}
{"level":"info","ts":1737733486.5286343,"logger":"tls.cache.maintenance","msg":"started background certificate maintenance","cache":"0xc0003ebb00"}
{"level":"warn","ts":1737733486.528772,"logger":"http","msg":"HTTP/2 skipped because it requires TLS","network":"tcp","addr":":80"}
{"level":"warn","ts":1737733486.528779,"logger":"http","msg":"HTTP/3 skipped because it requires TLS","network":"tcp","addr":":80"}
{"level":"info","ts":1737733486.5287817,"logger":"http.log","msg":"server running","name":"static","protocols":["h1","h2","h3"]}
{"level":"info","ts":1737733486.528789,"msg":"Caddy serving static files on :80"}
{"level":"info","ts":1737733486.5349355,"logger":"tls","msg":"cleaning storage unit","storage":"FileStorage:/data/caddy"}
{"level":"info","ts":1737733486.5373535,"logger":"tls","msg":"finished cleaning storage units"}

 TEST  Testing Example Test Suite ...
 TEST    It should respond to HTTP requests...😭 (4ms 96us 704ns)
 TEST      Expected valid response from server (failed)
 TEST  Testing Example Test Suite failed! 😭 (1 tests failed out of 1) (4ms 115us 914ns)
Error: Test("Test run failed with 1 failed assertions")

So it actually behaves completely randomly.

My setup:

$ rustc --version
rustc 1.84.0 (9fc6b4312 2025-01-07)
$ podman --version
podman version 5.3.2
$ uname -a
Linux alea-pc-arch 6.12.10-arch1-1 #1 SMP PREEMPT_DYNAMIC Sat, 18 Jan 2025 02:26:57 +0000 x86_64 GNU/Linux

[aleasims]

UPD.

If I add sleep right after container run

tokio::time::sleep(std::time::Duration::from_secs(10)).await;

the container will appear and I will be able to manually curl it:

$ curl http://127.0.0.1:8080/hello.txt
hello world

however the test still fails:

$ sam run
 TEST  Testing Example Test Suite ...
 TEST    It should respond to HTTP requests...😭 (971us 346ns)
 TEST      Expected valid response from server (failed)
 TEST  Testing Example Test Suite failed! 😭 (1 tests failed out of 1) (989us 566ns)
Error: Test("Test run failed with 1 failed assertions")

So looks like container is getting killed somewhere in the middle.

Something weird is going on with that processes here (maybe related to this?).

[trusch]

It's a race condition. Caddy starts async but right after that we already try to do the get request. This is even faster than doing an exec(curl) because the http request is directly supported by the engine.

The fix should be to showcase the wait_until function or to add a health check to the yaml that would cause that start_component() or the default start would block until it's reported healthy or run into an error