Files
archie/internal/infrastructure/cronstore/schedule.go
T
sam 84450ac96c fix(cronstore): write a schedule interval as a duration, not nanoseconds
The schedules document stored interval as a bare time.Duration, so its JSON was
the nanosecond count -- 1800000000000 for thirty minutes -- and the Web UI, which
now derives field metadata from the document type, correctly rendered a number
input. Same defect 02rm fixed for rate_limit.window, in the second stored time
this tree has. Nobody should be asked for a count of nanoseconds to configure a
schedule.

cronstore.Duration writes the string time.ParseDuration reads and reads that, a
nanosecond count, and nothing else: every schedule stored before this shape
existed carries a count, in both the control plane's schedules resource and this
package's own jobs table, which persists the spec as JSON. A value that parses as
neither is refused rather than decoding as zero, because a zero interval is a job
that never fires.

The derived schema needed no change and gets it for free -- which exposed a fix
the derivation itself needed: it classified durations by NAME, so cronstore.Duration
derived as an integer. durationLike now asks the type instead -- does it expose the
standard duration, and does it marshal as a quoted string -- so the next duration
type is classified correctly rather than silently rendering as a number box. That
failure mode was invisible until a live daemon showed it.

normalizeSchedules already re-marshals, so a stored document converges on its next
write while reads keep tolerating the count.

Verified: cronstore, crondelivery, controlplane, scheduling, webui and archied
suites, go vet ./..., go build ./..., golangci-lint run ./... (0 issues),
docs:check. Against a live daemon (8484): the catalog serves schedule.interval as
{type: string, format: duration}.
2026-09-22 04:22:46 +10:00

183 lines
7.0 KiB
Go

package cronstore
import (
"encoding/json"
"fmt"
"time"
)
// Duration is a schedule interval as the document writes it: the human string
// time.ParseDuration reads, never a nanosecond count. It reads the count too,
// because every schedule stored before this shape existed carries one -- both in
// the control plane's schedules resource and in this package's own jobs table,
// which persists the spec as JSON.
//
// The type is deliberately local, as config.Duration is for the file document
// and channelDuration is in internal/app/controlplane: each document owns its own
// wire form, and what is actually worth unifying later is the legacy-count
// tolerance rather than the tag shape.
type Duration time.Duration
func (d Duration) MarshalJSON() ([]byte, error) {
return json.Marshal(time.Duration(d).String())
}
func (d *Duration) UnmarshalJSON(data []byte) error {
var text string
if err := json.Unmarshal(data, &text); err == nil {
value, err := time.ParseDuration(text)
if err != nil {
return fmt.Errorf("%w: interval must be a Go duration string such as %q: %w", ErrInvalidSpec, "30m", err)
}
*d = Duration(value)
return nil
}
var nanos int64
if err := json.Unmarshal(data, &nanos); err != nil {
return fmt.Errorf("%w: interval must be a Go duration string such as %q or a nanosecond count: %w", ErrInvalidSpec, "30m", err)
}
*d = Duration(nanos)
return nil
}
// Std is the standard-library duration, for arithmetic and comparison.
func (d Duration) Std() time.Duration { return time.Duration(d) }
// String renders the interval the way the document writes it, so an error
// message or a log line reads "30m0s" rather than 1800000000000.
func (d Duration) String() string { return time.Duration(d).String() }
// Schedule decides when a job is due. The Kind field selects which of
// Interval, Cron, At the store uses to compute nextRun. Empty Kind
// resolves to "interval" so the simplest recurring job needs no
// explicit kind in its JSON.
type Schedule struct {
// Kind is one of ScheduleInterval, ScheduleCron, ScheduleOnce.
// Empty resolves to ScheduleInterval.
Kind string `json:"kind,omitempty"`
// Interval is the period for ScheduleInterval. Required when Kind
// is ScheduleInterval; ignored otherwise. It is a Duration rather
// than a time.Duration so the document carries "30m0s" and not the
// nanosecond count that asked an operator to type 1800000000000.
Interval Duration `json:"interval,omitempty"`
// Cron is the 5-field cron expression for ScheduleCron. Required
// when Kind is ScheduleCron; ignored otherwise.
Cron string `json:"cron,omitempty"`
// At is the one-shot firing time for ScheduleOnce. Required when
// Kind is ScheduleOnce; ignored otherwise.
At *time.Time `json:"at,omitempty"`
// Start anchors the first run for ScheduleInterval / ScheduleCron.
// Empty means "use creation time" — a freshly-created job with no
// Start still has a finite next_run. Omitted from the wire when
// unset so operator-readable files are not noisy.
Start time.Time `json:"start"`
}
// Schedule kinds the store recognises. Slice 1 fully implements
// Interval; Once is accepted at Create / fires at Due exactly once, but
// has no next-run arithmetic for MarkRun (one-shot by definition). Cron
// is parsed at Create but has no firing arithmetic yet — Slice 2 will
// add the parser. MarkRun on a kind this build does not implement
// returns ErrScheduleUnsupported rather than silently advancing next_run.
const (
// ScheduleInterval fires every Interval after the last successful run
// (or after Start, if never run). The simplest recurring schedule.
ScheduleInterval = "interval"
// ScheduleCron fires at the next moment matching a standard 5-field
// cron expression after the last successful run. Slice 2 will add
// the parser; until then, MarkRun on a cron job returns
// ErrScheduleUnsupported.
ScheduleCron = "cron"
// ScheduleOnce fires once at the configured At time. Subsequent
// MarkRun calls return ErrScheduleUnsupported because there is no
// recurring next-run to compute.
ScheduleOnce = "once"
)
// Validate rejects empty or unknown schedule kinds up front so Create does
// not have to defer the check until MarkRun. Interval additionally requires
// a strictly positive duration; the others defer their full validation.
func (s Schedule) Validate() error {
switch s.Kind {
case "", ScheduleInterval:
if s.Interval <= 0 {
return fmt.Errorf("%w: interval must be positive, got %v", ErrInvalidSpec, s.Interval)
}
case ScheduleCron:
if s.Cron == "" {
return fmt.Errorf("%w: cron expression must be non-empty", ErrInvalidSpec)
}
case ScheduleOnce:
if s.At == nil {
return fmt.Errorf("%w: once schedule requires an `at` time", ErrInvalidSpec)
}
default:
return fmt.Errorf("%w: unknown schedule kind %q", ErrInvalidSpec, s.Kind)
}
return nil
}
// resolve maps the empty kind onto its documented default — "interval" is
// what an operator gets when they leave the field unset on the simplest
// recurring job.
func (s Schedule) resolve() Schedule {
if s.Kind == "" {
s.Kind = ScheduleInterval
}
return s
}
// firstRun returns when the job should fire the very first time it has
// never run before. For Interval, it is Start + Interval (or now + Interval
// when Start is zero, so a freshly-created job with no Start still has a
// finite next-run). For Once, it is At. For Cron, it is left as a future
// responsibility — Slice 1 will not synthesise a value it cannot honour.
func (s Schedule) firstRun(now time.Time) (time.Time, error) {
switch s.resolve().Kind {
case ScheduleInterval:
if !s.Start.IsZero() {
return s.Start.Add(s.Interval.Std()), nil
}
return now.Add(s.Interval.Std()), nil
case ScheduleOnce:
if s.At == nil {
return time.Time{}, fmt.Errorf("%w: once schedule has no at time", ErrScheduleUnsupported)
}
return *s.At, nil
case ScheduleCron:
return time.Time{}, fmt.Errorf("%w: cron parsing not yet implemented (slice 2)", ErrScheduleUnsupported)
default:
return time.Time{}, fmt.Errorf("%w: unknown kind %q", ErrScheduleUnsupported, s.Kind)
}
}
func (s Schedule) FirstRun(now time.Time) (time.Time, error) { return s.firstRun(now) }
func (s Schedule) NextRun(lastRun time.Time) (time.Time, error) { return s.nextRun(lastRun) }
func (s Schedule) Resolved() Schedule { return s.resolve() }
// nextRun returns when the job should fire after a successful run at
// lastRun. It is the engine-side counterpart to firstRun: every MarkRun
// advances next_run through this function. Interval is fully implemented;
// Once and Cron deliberately fail so a silently-wrong value cannot reach
// the ticker.
func (s Schedule) nextRun(lastRun time.Time) (time.Time, error) {
switch s.resolve().Kind {
case ScheduleInterval:
return lastRun.Add(s.Interval.Std()), nil
case ScheduleOnce:
return time.Time{}, fmt.Errorf("%w: once schedule has no next run", ErrScheduleUnsupported)
case ScheduleCron:
return time.Time{}, fmt.Errorf("%w: cron arithmetic not yet implemented (slice 2)", ErrScheduleUnsupported)
default:
return time.Time{}, fmt.Errorf("%w: unknown kind %q", ErrScheduleUnsupported, s.Kind)
}
}