Jobs: recurrence settings
Activate or pause a recurring-job sequence
/a/JobRecurrence_setActive
Replace scheduler registration and optionally remove future incomplete children when pausing recurrence.
When to use this endpoint
Use the Active control in the recurring-jobs table. Pausing is not a harmless visibility toggle: it attempts cleanup of generated future jobs.
Authentication and permissions
Authenticate with signed API-key fields or a logged-in CRM session. Tenant context is required. The Job edit-restriction bit 128 applies, with the existing client-administrator exception.
IDs resolve inside the authenticated tenant. There is no additional department-visibility or recurrence-ownership filter.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
id | Integer string | Yes | Positive recurrence ID in canonical integer form; not a job ID. The record must exist. |
val | Integer string | Yes | Exactly 0 pauses or 1 enables; other values are rejected. |
skip-run | Legacy truthy flag | No | Default false. A truthy value suppresses only the scheduler-priority bump. It does not suppress the settings save, explicit cleanup, expiry registration updates or reminder reconciliation. |
Example request
$result = FM_api('JobRecurrence_setActive', [
'id' => '1042',
'val' => '1',
]);Example response
{"ok":1}A post-save follow-up failure adds saved=1 and warning/warnings without undoing the settings. Read the warning and reload affected objects rather than blindly replaying the operation.
Errors and unusual responses
| Response | Condition |
|---|---|
{"error":"not logged in"} | No authenticated tenant context is available. Invalid signed credentials can be rejected by the dispatcher before the handler runs. |
{"error":"missing or invalid parameter: id"} | ID missing, invalid or not positive. |
{"error":"job recurrence 1042 does not exist"} | No matching recurrence; numeric ID varies. |
{"error":"No permission to edit"} | The Job edit restriction applies and the caller is not a client administrator. |
{"error":"Unable to save job delivery changes"} | The checked local transaction failed; settings/history are rolled back. |
{"error":"Invalid saved recurrence metadata"} | The saved metadata cannot be decoded as an array/object; it is not silently erased. |
{"error":"missing or invalid parameter: val"} | The value is missing when required, has an unsupported type, or fails the operation-specific range/shape/date rules. |
Legacy application errors may be returned with HTTP 200. Check the decoded response for error, not only the HTTP status.
Behaviour and edge cases
- Pausing explicitly resets from now: it preserves the first valid cached anchor and completed/deleted jobs, and attempts deletion only of incomplete active children strictly after the current instant.
- Primary active state is saved before cleanup. A cleanup error or generator-busy result is returned as a saved warning, with cleanup details; settings are not silently rolled back.
- Global registration and type-7 expiry reminders are reconciled together in a checked global transaction after the local save. Only active recurrences with an end date receive an expiry reminder.
- Enabling uses recurrence-days and the stored expiry; notification_types.job_recurrence_expires_soon.num supplies reminder lead days. skip-run does not suppress registration/reminder reconciliation.
- A priority bump does not synchronously generate jobs: it backdates global last_run by seven days. Missing active registration is created using recurrence-days. Inactive sequences have their registration removed without being regenerated; an active sequence with blank on_arr yields a saved warning instead of a misleading rolled-back error.
- The surrounding Job_save recurrence editor invokes this setter internally. Its wider generation-setting save separately requests reset-from-now and now propagates failures/saved warnings. Direct endpoint effects above do not describe the whole editor save.
Side effects
- The existing recurrence is locked and changed settings, edit timestamp/operation marker and history are committed together in a checked local transaction. Failed local persistence rolls back; an unchanged value normally returns ok without a new local write.
- Cleanup, child-target changes and global scheduler/reminder reconciliation run after the local save. A failed follow-up preserves committed settings and returns ok=1, saved=1, warning/warnings. Inspect current state and any partial progress before retrying; tenant and global effects are not distributed-atomic.
- A priority bump does not synchronously generate jobs: it backdates global last_run by seven days. Missing active registration is created using recurrence-days. Inactive sequences have their registration removed without being regenerated; an active sequence with blank on_arr yields a saved warning instead of a misleading rolled-back error.
See it in Fieldmotion
CRM1
Open Jobs → Recurring Jobs and toggle the Active checkbox; filter for JobRecurrence_setActive. The customer recurring-jobs tab uses it too.
CRM2
No direct production CRM2 call to JobRecurrence_setActive was found. Its recurring-jobs module writes active through mapped recurrence updates and frequency changes through JobRecurrence_set. The shared job recurrence editor also reaches this setter indirectly through Job_save; observe Job_save in the Network panel rather than expecting a separate request to this focused route.
Related endpoints
JobRecurrence_getreloads the stored settings after an edit.