Runtime Cloud — observability¶
Target
This page describes intended behaviour or structure. No implementation is claimed unless a section is explicitly marked Current implementation or Evidence.
| Metric | Service | Instrument | Dimensions |
|---|---|---|---|
connectsoft.factory.provisioning.runs |
IaCProvisioningService | counter | tenant, kind, state |
connectsoft.factory.provisioning.apply_duration |
IaCProvisioningService | histogram | kind |
connectsoft.factory.provisioning.reconciliation_diffs |
IaCProvisioningService | counter | tenant, destructive |
connectsoft.factory.runtime_cloud.host.inbox_lag |
RuntimeCloudHost | gauge | module |
connectsoft.factory.capacity.pool_utilisation |
ExecutionCapacityService | gauge | poolId, kind |
connectsoft.factory.capacity.leases_active |
ExecutionCapacityService | updown | tenant, holderKind |
connectsoft.factory.capacity.lease_wait |
ExecutionCapacityService | histogram | tenant, holderKind |
connectsoft.factory.capacity.quota_denials |
ExecutionCapacityService | counter | tenant, limit |
connectsoft.factory.runtime.environments_stale |
RuntimeEnvironmentService | updown | tenant |
connectsoft.factory.runtime.drift_open |
RuntimeEnvironmentService | updown | tenant, kind |
connectsoft.factory.runtime.reconciliation_proposals |
RuntimeEnvironmentService | updown | tenant, state, destructive |
The platform's signals¶
| Signal | Why it matters |
|---|---|
connectsoft.factory.provisioning.runs, apply_duration, reconciliation_diffs |
how often and how fast infrastructure changes, and how far reality drifts between reconciliations |
connectsoft.factory.runtime.environments_stale, drift_open, reconciliation_proposals |
whether the inventory can be trusted, and the backlog of human decisions |
connectsoft.factory.capacity.pool_utilisation, leases_active, lease_wait, quota_denials |
capacity pressure per pool and tenant; lease wait is a direct component of run latency |
Objectives and alerts¶
SloService defines platform objectives on lease wait (interactive latency component) and environment staleness; AlertRuleService fires on ProvisioningFailed, destructive RuntimeDriftDetected, pools Unavailable and sustained quota denials. Spans carry connectsoft.factory.environment_id, provisioning_run_id, pool_id, lease_id.