How to monitor your ImpulseVPS fleet

ImpulseVPS reports on your fleet in three places: the Dashboard for what needs attention now, the Diagnostics tab for checks and logs, and ImpulseCore's Cron page for the scheduled tasks that do the work.

The Dashboard

Open Addons → ImpulseVPS. The Dashboard shows:

• Open alerts, critical first.
• The Launch readiness checklist, until every setup step passes.
• Server counts, once you have servers: Active, Provisioning, Suspended, In grace (terminated but still restorable) and Failed (setup or restore failed).
• Provider health: each provider reads Reachable, Unreachable or Unknown, with the time of the last check. This comes from the 15-minute power_sync task, so opening the Dashboard never calls a provider.
• Recent activity (the last 10 audit events) and System status (version, registration, license and tables).

Alerts

Alerts include:

• stuck orders and actions
• provider accounts that can't be reached, or whose API key is refused
• servers that no service owns
• a suspended server still running, or an active one that is off
• catalog problems that affect what you sell
• plan, backup and storage changes that didn't reach a server
• failed teardowns

Each row links to where you fix it, such as Open service or Open Images. Three or more alerts of one kind are grouped into one row.

Most alerts close themselves once the cause is gone or the operation later succeeds. Conditions the hourly check recomputes, such as a stuck order, close after two clear runs. A few need you:

• An unowned server stays listed until you Adopt it or click Destroy instance in the orphan report on the Instances tab.
• An unknown create stays until you click Resolve unknown create on the service.
• Alerts that nothing else closes, such as a snapshot a purge couldn't delete, have a Mark resolved button. Check the provider account first, because the item may still be billing.

Alert email

Configure alert email under Settings → Alerts & Notifications:

• Send operator alert emails is on by default. Off sends no alert email, but the Dashboard still lists every alert.
• Alert recipients: left blank, alerts go to the WHMCS administrators who receive system emails. Addresses you enter need a provider under ImpulseCore → Email Center → Delivery. Until one is set, the administrators get the email.
• Re-notify interval (hours) defaults to 24. If a warning reopens within that time, its email waits until the interval is up and goes out if it's still open. A critical alert is emailed every time it opens.

A provisioning failure and a refused provider API key are emailed as soon as they open, and so are catalog alerts opened by the daily catalog refresh or by your own catalog, plan or product changes. Every other critical or warning alert goes out in one digest from the hourly check. Info alerts are never emailed. The same page also sets Provision timeout (min) (default 15), a deadline for each long-running action, and whether the orphan sweep runs.

Scheduled tasks

ImpulseVPS has no cron entry of its own. ImpulseCore runs its six tasks from the WHMCS system cron, which should run every 5 minutes.

• provision (every tick) creates queued servers and activates each service once its server starts. A server that isn't ready within the provision timeout is marked failed and raises an alert. After an upgrade, this task also updates the module's database.
• action_poll (every tick) follows power changes, reinstalls, resizes, snapshots and volume changes until they finish. It fails any action that passes its deadline.
• power_sync (every 15 minutes) refreshes power state and provider health. It raises a critical alert when a provider refuses an account's API key.
• alert_detection (hourly) finds stuck orders and actions, power-state mismatches, unowned servers and catalog problems, and sends the digest. It also retries snapshots and volumes a purge couldn't delete yet, and refreshes the update check.
• image_sync (daily) refreshes the operating-system and application catalogs.
• usage_sync (hourly) is reserved for usage collection. It records nothing in this version.

Where to see their health

• ImpulseCore → Operations → Cron: Cron Health shows each impulsevps task's schedule, last run, next run and status (for example healthy, overdue or stale). Recent runs shows each run's duration and its result or error. The button under Diagnostics → Help → Cron info opens this page filtered to ImpulseVPS.
• Diagnostics → Health Checks → Cron jobs fails if a task isn't registered, has never run or is late. The result names the task.

ImpulseCore may list provision and action_poll as every 2 minutes. With a 5-minute cron, they run every 5 minutes.

The Diagnostics tab

• Health Checks: Cron jobs, VPS providers, VPS instances, Email delivery, WHMCS product sync, Schema & runtime and Provisioning preflight. DNS provider and Cloud-Init templates report Skipped, because ImpulseVPS uses neither. Provider diagnostics checks one account, read-only. The Synthetic end-to-end test creates one real, billable server and destroys it again, even after a timeout or an error. If that destroy fails, or the provider's answer to the create is lost, you get a critical alert: find the server in the provider account and delete it.
• Activity is the audit trail, filterable by service, event and level.
• Logs shows recent failed actions and provider warnings and errors.
• License shows status, tier and features.
• Help has version and cron details and a support snapshot with secrets masked.

The update check

When the last check found a newer release, a green Update v… available badge appears in the addon's header. If the last check failed, a grey Update check failed badge appears instead. A failed check is never shown as up to date.

Diagnostics → Help → Version info shows the current state, when the last check ran and a Check now button. Opening a page never runs the check. The hourly alert_detection task refreshes a result that is more than 6 hours old.

Related

• How to find a customer's server
• Fix: an order fails at Create, or setup fails after the order was accepted
• Fix: a provider rejects an action or reports it busy
• Adding a provider account to ImpulseVPS