system2 0.1.0

An (unofficial) idiomatic Rust SDK for TypeSafe's System One model, Jev.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224

//! Tests of the retries of a request.

mod support;

use std::sync::atomic::Ordering;
use std::time::{Duration, Instant, SystemTime};
use support::{
    INSTRUCTIONS, STATE, client_for, mock_api, mock_api_then, noul, noul_response, partial_api,
    request_count,
};
use system2::AskError;
use tokio::time::{sleep, timeout};
use wiremock::ResponseTemplate;

// the first wait of the backoff, less the most that jitter subtracts
const SHORTEST_BACKOFF: Duration = Duration::from_millis(375);

fn retry_now(status: u16) -> ResponseTemplate {
    ResponseTemplate::new(status).insert_header("retry-after-ms", "0")
}

#[tokio::test]
async fn ask_retries_408_429_and_5xx() {
    for status in [408, 429, 500, 503, 599] {
        let server = mock_api_then(retry_now(status), noul_response(0.92)).await;

        let answer = client_for(&server)
            .ask(noul(INSTRUCTIONS), STATE)
            .await
            .unwrap();

        assert_eq!(answer.probability(), 0.92, "{status}");
        assert_eq!(request_count(&server).await, 2, "{status}");
    }
}

#[tokio::test]
async fn ask_returns_other_statuses_with_no_retry() {
    for status in [400, 401, 402, 404, 413, 422] {
        let server = mock_api(retry_now(status)).await;

        let error = client_for(&server)
            .ask(noul(INSTRUCTIONS), STATE)
            .await
            .unwrap_err();

        assert!(
            matches!(error, AskError::Http { status: s, .. } if s == status),
            "{status}: {error:?}"
        );
        assert_eq!(request_count(&server).await, 1, "{status}");
    }
}

#[tokio::test]
async fn ask_retries_as_many_times_as_max_retries() {
    for (max_retries, attempts) in [(None, 3), (Some(0), 1), (Some(1), 2), (Some(4), 5)] {
        let server = mock_api(retry_now(503)).await;
        let client = client_for(&server);
        let client = match max_retries {
            Some(max_retries) => client.max_retries(max_retries),
            None => client,
        };

        let error = client.ask(noul(INSTRUCTIONS), STATE).await.unwrap_err();

        assert!(
            matches!(error, AskError::Http { status: 503, .. }),
            "{max_retries:?}: {error:?}"
        );
        assert_eq!(request_count(&server).await, attempts, "{max_retries:?}");
    }
}

#[tokio::test]
async fn ask_waits_as_a_header_tells() {
    // each wait is longer than the longest first backoff
    let cases = [
        ("retry-after-ms", "600", Duration::from_millis(600)),
        ("retry-after", "1", Duration::from_secs(1)),
    ];
    for (header, value, wait) in cases {
        let first = ResponseTemplate::new(429).insert_header(header, value);
        let server = mock_api_then(first, noul_response(0.92)).await;

        let start = Instant::now();
        client_for(&server)
            .ask(noul(INSTRUCTIONS), STATE)
            .await
            .unwrap();

        assert!(start.elapsed() >= wait, "{header}: {:?}", start.elapsed());
    }
}

#[tokio::test]
async fn ask_waits_for_the_backoff_with_no_header() {
    let server = mock_api_then(ResponseTemplate::new(503), noul_response(0.92)).await;

    let start = Instant::now();
    client_for(&server)
        .ask(noul(INSTRUCTIONS), STATE)
        .await
        .unwrap();

    assert!(start.elapsed() >= SHORTEST_BACKOFF, "{:?}", start.elapsed());
}

#[tokio::test]
async fn ask_returns_the_error_when_a_header_asks_for_more_than_30_seconds() {
    let in_an_hour = httpdate::fmt_http_date(SystemTime::now() + Duration::from_secs(3600));
    let cases = [
        ("retry-after-ms", "30001".to_owned()),
        ("retry-after", "31".to_owned()),
        ("retry-after", in_an_hour),
    ];
    for (header, value) in cases {
        let first = ResponseTemplate::new(429).insert_header(header, value.as_str());
        let server = mock_api_then(first, noul_response(0.92)).await;

        let error = client_for(&server)
            .ask(noul(INSTRUCTIONS), STATE)
            .await
            .unwrap_err();

        assert!(
            matches!(error, AskError::Http { status: 429, .. }),
            "{value}: {error:?}"
        );
        assert_eq!(request_count(&server).await, 1, "{value}");
    }
}

#[tokio::test]
async fn ask_follows_no_redirect() {
    for status in [301, 307] {
        let response = ResponseTemplate::new(status).insert_header("location", "/v1/systemone");
        let server = mock_api(response).await;

        let error = client_for(&server)
            .ask(noul(INSTRUCTIONS), STATE)
            .await
            .unwrap_err();

        assert!(
            matches!(error, AskError::Http { status: s, .. } if s == status),
            "{status}: {error:?}"
        );
        assert_eq!(request_count(&server).await, 1, "{status}");
    }
}

#[tokio::test]
async fn ask_retries_a_timeout() {
    let slow = noul_response(0.1).set_delay(Duration::from_secs(5));
    let server = mock_api_then(slow, noul_response(0.92)).await;

    let answer = client_for(&server)
        .timeout(Duration::from_millis(500))
        .ask(noul(INSTRUCTIONS), STATE)
        .await
        .unwrap();

    assert_eq!(answer.probability(), 0.92);
    assert_eq!(request_count(&server).await, 2);
}

#[tokio::test]
async fn ask_retries_a_failed_connection() {
    // the connection closes before the response, or in its body
    let cases: [&[u8]; 2] = [b"", b"HTTP/1.1 200 OK\r\ncontent-length: 100\r\n\r\n{"];
    for partial_response in cases {
        let (client, connections) = partial_api(partial_response).await;

        let error = client
            .max_retries(1)
            .ask(noul(INSTRUCTIONS), STATE)
            .await
            .unwrap_err();

        assert!(matches!(error, AskError::Transport(_)), "{error:?}");
        assert_eq!(connections.load(Ordering::SeqCst), 2);
    }
}

#[tokio::test]
async fn ask_keeps_the_status_of_a_response_that_fails_in_its_body() {
    let cases: [(&[u8], u16); 2] = [
        (
            b"HTTP/1.1 401 Unauthorized\r\ncontent-length: 100\r\n\r\n{",
            401,
        ),
        (
            b"HTTP/1.1 429 Too Many Requests\r\nretry-after: 31\r\ncontent-length: 100\r\n\r\n{",
            429,
        ),
    ];
    for (partial_response, status) in cases {
        let (client, connections) = partial_api(partial_response).await;

        let error = client.ask(noul(INSTRUCTIONS), STATE).await.unwrap_err();

        assert!(
            matches!(error, AskError::Http { status: s, .. } if s == status),
            "{status}: {error:?}"
        );
        assert_eq!(connections.load(Ordering::SeqCst), 1, "{status}");
    }
}

#[tokio::test]
async fn ask_sends_no_more_attempts_when_dropped_in_a_wait() {
    let first = ResponseTemplate::new(503).insert_header("retry-after-ms", "1000");
    let server = mock_api_then(first, noul_response(0.92)).await;

    // wide margins, so a busy machine cannot run the retry before the drop
    let asking = client_for(&server).ask(noul(INSTRUCTIONS), STATE);
    timeout(Duration::from_millis(100), asking)
        .await
        .expect_err("the request must still wait for its retry");
    sleep(Duration::from_millis(1200)).await;

    assert_eq!(request_count(&server).await, 1);
}