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
use ;
use Notify;
/// Scheduling intent after a successful worker tick.
///
/// A tick has exactly one intent. Setting a schedule replaces the previous intent.
/// Both [`super::WorkerTick`] and [`super::BatchRunSummary`] use this type.
/// Worker-level errors always use failure backoff instead. Shutdown takes
/// precedence over every schedule, and each new tick still checks its lease.
/// Coalescing wakeup signal for one worker runner.
///
/// Clone the handle for producers and install it with
/// [`super::WorkerRunOptions::with_wakeup`]. Notify **after** the enqueue
/// transaction commits. External listeners can forward events through the same
/// handle; the worker has no dependency on their transport or database.
///
/// Notifications carry no tasks and are not counted: at most one pending signal
/// is retained, including a signal sent before the worker starts waiting. A
/// notification during a tick can cause one extra queue check. Use a separate
/// handle for each runner; clones share a signal, they do not broadcast it.
/// Dropping producer handles leaves polling operational.
///
/// Only [`WorkerSchedule::WaitForWork`] consumes signals. Notifications never
/// interrupt startup delays, explicit delays, failure backoff, or lease waits.