System health
Last updated on
Clynto's background jobs run in one process at a time — either the web server or a separate background worker — chosen by a lease in the database, so a job never runs twice. Settings → System info → System health (admins) shows whether they're alive.
- Workers — how many are running, and a row per worker with what it does and how often it checks.
- Running in — which process holds the jobs (web server or background worker) and when it last checked in.
- Database — whether it's reachable and how fast it answered.
| Worker | What it does | Checks every |
|---|---|---|
| Integration sync scheduler | Runs each workspace's scheduled syncs | 1 min |
| Health recompute scheduler | Recomputes health and journey | 30 min |
| Workflow scheduler | Runs every live workflow's trigger check | 15 min |
| Workflow fast-poll | Re-checks workflows in testing mode | 15 s |
| Portfolio digest | Builds each CSM's daily digest | 1 h |
| Inbound email poller | Pulls newly received customer email | 1 min |
| Campaign scheduler / sender | Starts scheduled campaigns; sends queued emails | 30 s / 5 s |
| Larry Time questions | Finds missing data to ask about | 15 min |
| Document memory | Reads new account documents for Larry | 20 s |
| Status auto-map | Maps new status values to Active or Lost | 5 min |
| Retention sweep | Erases what has passed its deletion window | 1 h |
| Market Intelligence scanner | Reads public news about accounts (when turned on) | 15 min check; per workspace every 6 h by default, or the cadence you choose (6 hours to 90 days) |
| Quiet-hours email delivery | Sends proactive emails held overnight, once it's morning for the recipient | 1 min |
| Sent-folder sync | Checks CSM mailboxes for the drafts Clynto placed: sent, discarded or not sent; asks to close the task they were for | 5 min |
| Workspace context summaries | Condenses long workspace-context uploads | 30 s |
| Reminders | Fires people's personal reminders when they're due, and repeats them until done | 1 min |
That's 17 workers in all. The daily credits-pace update runs inside the Health recompute scheduler, not as a worker of its own.
Recheck reads the state again. Wake workers restarts any that stopped inside a process that is already running; when the jobs run in the background worker, the restart is handed to it and happens within about 30 seconds. It can't stop the app from going to sleep — see the tip.
Tip
If the host idles the app when nobody is using it, every job stops. Point an uptime monitor at the /health address shown on the page every few minutes (no login needed), or turn on your host's Always On setting.