OUT_OF_RANGE — Requested sequence number is out of range

The requested sequence number is no longer in the agent's buffer, or is beyond what it has produced. Agents keep a rolling buffer, and anything older has been discarded.

Start free MTConnect → canonical

Diagnosis

When you hit this

A client that fell behind and asked for a sequence the agent has since aged out.
Resuming from a stored sequence number after the agent restarted, which resets the sequence space.
A buffer sized too small for the client's polling interval.

Resolution

How to fix it

Recover via /current
Fetch /current to get the present nextSequence and resume streaming from there. Do not retry the stale number.
Track instanceId
The agent's instanceId changes on restart. If it changed, your saved sequence is meaningless and you must resynchronize.
Size the buffer for your slowest client
Buffer size divided by data rate is your recovery window. Make it longer than any expected client outage.

Normalization

What Forge does about it

A request that fails on some points and succeeds on others still normalizes. The tags that arrived resolve to canonical fields as usual; the ones that did not come back are listed in unresolved_tags and excluded from coverage_pct, so a partial read reports as partial instead of quietly looking complete.