Schedule cadence

Every server has one update interval, and it comes from its tier:

Tier	Default interval
B2B · Tier 1 · Tier 2	3 months
Tier 3 · MGMT	6 months

The split is between what is reachable from outside — the internet or a partner — and what sits inside. The two error directions are not symmetric: patching an internal machine that did not need it costs a maintenance window, and failing to patch an externally reachable one costs exposure.

Both numbers are editable

Settings → General settings → Schedule cadence holds two fields, B2B / T1 / T2 and T3 / MGM, both in months. Changing them affects the whole inventory on the next read, not only newly added servers.

The next due date

Next update = last update + the interval. The two columns in the server list — Frequency and Next update — come from the same calculation, so they cannot contradict each other.

The addition is calendar-month addition, not day arithmetic: 31/01 plus one month is 28/02 (or 29/02 in a leap year), not 03/03.

When the last-update date cannot be read, the next due date stays as it arrived from the inventory source — it is not blanked. Erasing a date you cannot improve on is worse than leaving it.

How it reads on screen

The server list carries three adjacent columns: Last update, Next update and Frequency. A server that is past due and a server that will be due within 30 days are marked separately, and both can be filtered on directly.

The same split feeds the Maintenance debt card in the calendar's context panel: how many servers are past due with no window scheduled, and how many will be due in the next 30 days.

The cadence is a planning default, not a lock: a window can be scheduled at any date. It answers "why have we not touched this server yet", not "am I allowed to touch it now".