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
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
use std::{
sync::Arc,
time::{Duration, Instant},
};
use crate::{
ActivationStrategy, Backend, TaskDefinition, TaskSuccess, Worker, WorkerFactory,
backends::{ClaimTaskError, FailTaskError, FinishTaskError, RenewTaskError},
};
use tokio::sync::mpsc::UnboundedSender as MpscSender;
use tracing::{trace, warn};
const LEASE_DURATION: Duration = Duration::from_secs(20);
const LEASE_RENEWAL_THRESHOLD: Duration = Duration::from_secs(10);
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub(crate) struct RuntimeUpdate {
pub next_available_from_update: Option<Option<Instant>>,
pub claimed_task: bool,
}
/// An internal type that handles worker lifecycle.
#[derive(Debug)]
pub(crate) struct WorkerRuntime<B, F> {
pub backend: B,
pub factory: Arc<F>,
pub worker_id: u64,
pub update_signal: MpscSender<RuntimeUpdate>,
}
impl<B, F> WorkerRuntime<B, F>
where
B: Backend + 'static,
F: WorkerFactory + 'static,
{
pub fn run(
self,
dispatch_token: <<<F::Worker as Worker>::Task as TaskDefinition>::Trigger as ActivationStrategy>::DispatchToken,
) {
let daemon = Daemon {
backend: self.backend,
factory: self.factory,
worker_id: self.worker_id,
update_signal: self.update_signal,
status: DaemonStatus::WaitingForTask { dispatch_token },
};
tokio::spawn(daemon.run());
}
}
struct Daemon<B, F, T>
where
T: TaskDefinition,
{
backend: B,
factory: Arc<F>,
worker_id: u64,
update_signal: MpscSender<RuntimeUpdate>,
status: DaemonStatus<T>,
}
impl<B, F> Daemon<B, F, <F::Worker as Worker>::Task>
where
B: Backend,
F: WorkerFactory,
F::Worker: 'static,
{
async fn run(mut self) {
loop {
let (new_self, action, update) = self.event_loop().await;
self = new_self;
if let Some(update) = update {
let _ = self.update_signal.send(update);
}
if matches!(action, EventLoopResult::Exit) {
break;
}
}
}
async fn event_loop(mut self) -> (Self, EventLoopResult, Option<RuntimeUpdate>) {
let (status, action, update) = match self.status {
DaemonStatus::WaitingForTask { dispatch_token } => {
match <<<F::Worker as Worker>::Task as TaskDefinition>::Trigger as ActivationStrategy>::claim_task::<B, <F::Worker as Worker>::Task>(
&self.backend,
self.worker_id,
dispatch_token,
Instant::now() + LEASE_DURATION,
)
.await {
Ok(task) => (
DaemonStatus::TaskClaimed {
task_id: task.task_id,
task_payload: task.task_payload,
lease_expiration: task.lease_expiration,
},
EventLoopResult::Continue,
Some(RuntimeUpdate {
next_available_from_update: None,
claimed_task: true,
}),
),
Err(ClaimTaskError::TaskLeased { expiration }) => {
trace!("Unable to claim task with worker #{}", self.worker_id);
(
DaemonStatus::WorkerExited,
EventLoopResult::Exit,
Some(RuntimeUpdate {
next_available_from_update: Some(Some(expiration)),
claimed_task: false,
}),
)
}
Err(ClaimTaskError::TaskUnavailable { available_from }) => {
trace!("No task available for worker #{}", self.worker_id);
(
DaemonStatus::WorkerExited,
EventLoopResult::Exit,
Some(RuntimeUpdate {
next_available_from_update: Some(available_from),
claimed_task: false,
}),
)
}
Err(ClaimTaskError::TaskNotFound) => {
(DaemonStatus::WorkerExited, EventLoopResult::Exit, None)
}
Err(ClaimTaskError::Backend(err)) => {
warn!(
"Unable to claim task with worker #{} due to backend error: {}",
self.worker_id, err
);
(DaemonStatus::WorkerExited, EventLoopResult::Exit, None)
}
}
}
DaemonStatus::TaskClaimed {
task_id,
task_payload,
lease_expiration,
} => {
let worker = self.factory.build(self.worker_id);
let worker_handle =
tokio::spawn(async move { worker.process(task_id, task_payload).await });
(
DaemonStatus::WorkerStarted {
task_id,
worker_handle,
lease_expiration,
},
EventLoopResult::Continue,
None,
)
}
DaemonStatus::WorkerStarted {
task_id,
mut worker_handle,
lease_expiration,
} => {
let renewal_deadline = lease_expiration
.checked_sub(LEASE_RENEWAL_THRESHOLD)
.unwrap_or_else(Instant::now);
let renewal =
tokio::time::sleep_until(tokio::time::Instant::from_std(renewal_deadline));
tokio::select! {
_ = renewal => {
let renewal = self
.backend
.renew(
self.worker_id,
task_id,
Instant::now() + LEASE_DURATION,
)
.await;
match renewal {
Ok(new_lease) => {
trace!("Lease renewed");
(
DaemonStatus::WorkerStarted {
task_id,
worker_handle,
lease_expiration: new_lease.new_expiration,
},
EventLoopResult::Continue,
None,
)
}
Err(RenewTaskError::LeaseLost) => {
trace!(
"Lease lost for task #{}, aborting worker #{}",
task_id,
self.worker_id
);
worker_handle.abort();
(DaemonStatus::WorkerExited, EventLoopResult::Exit, None)
}
Err(RenewTaskError::Backend(err)) => {
warn!(
"Unable to renew lease for task #{} with worker #{}: {}",
task_id,
self.worker_id,
err
);
worker_handle.abort();
(DaemonStatus::WorkerExited, EventLoopResult::Exit, None)
}
}
},
join_result = &mut worker_handle => {
match join_result {
Ok(Ok(task_success)) => (
DaemonStatus::WorkerFinishedProcessing {
task_id,
task_success,
},
EventLoopResult::Continue,
None,
),
Ok(Err(task_failure)) => (
DaemonStatus::WorkerFailedProcessing {
task_id,
available_from: task_failure.available_from,
},
EventLoopResult::Continue,
None,
),
Err(err) => {
warn!(
"Worker #{} for task #{} exited unexpectedly: {}; retrying immediately",
self.worker_id,
task_id,
err
);
(
DaemonStatus::WorkerFailedProcessing {
task_id,
available_from: None,
},
EventLoopResult::Continue,
None,
)
}
}
},
}
}
DaemonStatus::WorkerFailedProcessing {
task_id,
available_from,
} => {
match self
.backend
.fail(self.worker_id, task_id, available_from)
.await
{
Ok(_) => {
trace!("Failed task #{} with worker #{}", task_id, self.worker_id);
}
Err(FailTaskError::LeaseLost) => {
warn!(
"Worker #{} failed task #{} but no longer holds its lease; assuming another worker took over",
self.worker_id, task_id
);
}
Err(FailTaskError::Backend(err)) => {
warn!(
"Unable to record failure for task #{} with worker #{} due to backend error: {}",
task_id, self.worker_id, err
);
}
}
(
DaemonStatus::WorkerExited,
EventLoopResult::Exit,
Some(RuntimeUpdate {
next_available_from_update: None,
claimed_task: false,
}),
)
}
DaemonStatus::WorkerFinishedProcessing {
task_id,
task_success,
} => {
match self
.backend
.finish::<<F::Worker as Worker>::Task>(
self.worker_id,
task_id,
task_success.callback_payload,
task_success.available_from,
)
.await
{
Ok(_) => {
trace!("Finished task #{} with worker #{}", task_id, self.worker_id);
}
Err(FinishTaskError::LeaseLost) => {
warn!(
"Worker #{} finished task #{} but no longer holds its lease; assuming another worker took over",
self.worker_id, task_id
);
}
Err(FinishTaskError::Backend(err)) => {
warn!(
"Unable to finalize task #{} with worker #{} due to backend error: {}",
task_id, self.worker_id, err
);
}
}
(
DaemonStatus::WorkerExited,
EventLoopResult::Exit,
Some(RuntimeUpdate {
next_available_from_update: None,
claimed_task: false,
}),
)
}
DaemonStatus::WorkerExited => {
unreachable!("Worker runtime event loop should have ended")
}
};
self.status = status;
(self, action, update)
}
}
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
enum EventLoopResult {
Continue,
Exit,
}
/// Worker runtime daemon lifecycle stages.
enum DaemonStatus<T>
where
T: TaskDefinition,
{
/// The runtime has just been started and have not claimed the task yet.
WaitingForTask {
dispatch_token: <T::Trigger as ActivationStrategy>::DispatchToken,
},
/// The runtime has claimed the task.
TaskClaimed {
task_id: u64,
task_payload: <T::Trigger as ActivationStrategy>::EffectivePayload,
lease_expiration: Instant,
},
/// The background worker has been spawned.
WorkerStarted {
task_id: u64,
worker_handle: tokio::task::JoinHandle<crate::TaskResult<T::Callback>>,
lease_expiration: Instant,
},
/// The worker finished unsuccessfully and the task should be released back to the backend.
WorkerFailedProcessing {
task_id: u64,
available_from: Option<Instant>,
},
/// The worker future completed successfully and the task should be finalized in the backend.
WorkerFinishedProcessing {
task_id: u64,
task_success: TaskSuccess<T::Callback>,
},
/// The worker runtime is done.
WorkerExited,
}