Platform

Quotas & limits

Q: What are foro.sh's plan limits?

Free includes 1 real server (plus the auto-provisioned example project, which never counts against the limit) and 1 seat. Starter includes 3 servers and 1 seat. Team includes 10 servers and 10 seats. Enterprise starts from the Team base with custom limits by agreement.

Q: How long are logs and metrics kept on each plan?

Two separate windows. Persisted logs and per-call metric rows - the high-fidelity tier that answers exact latency percentiles and carries verbatim error strings - are kept for 1 day on Free, 7 days on Starter, 30 days on Team and 90 days on Enterprise. Daily metric rollups are kept far longer: 7 days on Free, 30 days on Starter, 365 days on Team and 730 days on Enterprise. So a plan buys both more detail and more range, and call counts stay visible long after the individual calls behind them have aged out.

Q: What happens when a free workspace runs out of tool calls?

Free workspaces are metered: 10,000 tool calls per UTC calendar month, counted per workspace rather than per server. Crossing it suspends every non-example server the workspace has running, not just the one whose traffic tripped it. The auto-provisioned example project is exempt - its calls are never counted and it is never suspended. A manual restart is refused until the cap clears, which happens by upgrading the workspace's plan or by next month's counter starting at zero, whichever comes first.

What your plan entitles your workspace to, and exactly what happens when you reach any of it.

What are foro.sh's plan limits?

Plan	Servers	Seats	Tool calls	Logs	Metrics
Free	1	1	10,000 / month	1 day	7 days
Starter	3	1	Unlimited	7 days	30 days
Team	10	10	Unlimited	30 days	365 days
Enterprise	10 base, custom	10 base, custom	Unlimited	90 days	730 days

The auto-provisioned example project every workspace starts with doesn't count against the server limit or the tool-call cap - a free workspace can explore it fully before deploying a real project requires an upgrade.

How long are logs and metrics kept on each plan?

The two columns above are separate windows, and they answer different questions. The Logs window covers persisted logs and the per-call metric rows behind them - the high-fidelity tier that can still show you an exact latency percentile, an hourly breakdown, and the verbatim text of an error. The Metrics window covers the daily rollups: call counts, error counts and a latency histogram per day.

Rollups always outlive the calls they summarise, so your traffic history stays readable long after the individual calls behind it have been pruned - on Team that is 365 days of counts against 30 days of detail. Both windows are enforced by the worker's hourly retention pass, not just hidden from the dashboard: once a row is past its window it is deleted.

What happens when a free workspace runs out of tool calls?

Free workspaces are metered at 10,000 tool calls per UTC calendar month, counted per workspace rather than per server. Crossing it suspends every non-example server the workspace has running, not only the one whose traffic tripped it - the cap is on the workspace's total usage, not any one server's.

The auto-provisioned example project is exempt: its calls are never counted and it's never suspended. A manual restart of a suspended server is refused until the cap clears, which happens either by upgrading the workspace's plan or by next month's counter starting at zero - no plan above Free has a ceiling.

What happens to a workspace after a downgrade?

Limits are enforced at the action, never retroactively - but a downgrade can still stop a live server directly. If the workspace ends up running more servers than its new plan allows, the most recently deployed servers are kept and the rest are suspended immediately, not just blocked from restarting. Nothing is deleted: the suspended projects keep their state, and a later upgrade restores them automatically, oldest suspension first, as room allows.