When trying to connect to Sage Intacct, it may not not respond before its own gateway timed out, and returned an HTTP 504 to Arcoro. This status is on Arcoro's automatic retry list for Sage Intacct, so Arcoro retries the same request up to five more times, with the delay between attempts roughly doubling each time plus a small random amount, before giving up and logging this message. This is a delay or outage reaching Sage Intacct, not a sign-in failure; the connector's stored credentials are not involved.
Messages in this article
| Message text | Direction | Occurs when | Outcome | Fix |
|---|---|---|---|---|
| The request timed out in Sage Intacct. | Shared connection layer. Fires on a read of any entity from Sage Intacct, and on the lookup Arcoro runs partway through a Location update write. | Sage Intacct returns HTTP 504 and Arcoro's automatic retries of that same request are exhausted. | Varies. If it happens while Arcoro is reading an entity's records from Sage Intacct, that entity's read stops for the rest of this run and is logged as a Failure; other entities in the same sync keep going. If it happens during the Location-update lookup, only that one Location is skipped and logged as an Error; other Location records continue. | Re-run the sync. If the timeout returns on re-run, escalate. |
Confirming the Error
- Open the sync run for the affected connector in Sync History.
- Find the entity row that looked incomplete: Company/Locations if the affected record was a Location, or whichever entity was being read from Sage Intacct.
- Check whether the count landed under Errors or Failures for that row. Errors means one Location record was skipped while the rest of the entity's records were still processed; Added, Updated, and Unchanged on the same row should otherwise look normal. Failures means Arcoro stopped partway through reading that entity for this run; other entities in the same sync are unaffected.
- Open the message detail and match on the exact fixed sentence The request timed out in Sage Intacct. It never has extra text appended to it in this article's scope.
- Confirm the message is this exact sentence and not a passthrough message from Sage Intacct. A passthrough message reads like Sage Intacct error: <code>-<text> - SupportId: <id>, or, near a write, Intacct error <id> - <text> encountered while attempting to save.... Either of those is a different problem, out of scope for this article.
Resolving the Error
- Re-run the sync for the affected step or entity. Arcoro already retried the original request several times, with the delay roughly doubling each attempt, before this message was logged, so a re-run sends a fresh set of attempts rather than repeating the ones that already failed.
- If the re-run completes without this message, no further action is needed. Confirm by checking that the entity's Errors and Failures counts are back to zero and that Added, Updated, and Unchanged look normal for that run.
- If the message returns on re-run, this points to a repeated problem getting a timely response from Sage Intacct rather than a one-time delay. Gather the details below and escalate rather than continuing to re-run.
When to Escalate
If any of the following occur, you may need to reach out to Arcoro support for additional assistance.
- The timeout recurs on a re-run, or recurs across more than one sync execution for the same customer.
- Before escalating, capture: the entity or record that showed the message, the full message text exactly as it appears, and whether it was counted under Errors or Failures.
- Also capture whether the message text is this exact sentence or a passthrough message containing a Sage Intacct code, and a support ID if one is present. A passthrough message is a different issue and does not belong in this escalation, but the support ID (when present) is exactly what Sage Intacct support needs to look up the original failure on their side.
- If the count was under Failures, note that the affected entity's read stopped for this run while other entities continued; that framing helps whoever picks up the escalation gauge urgency without re-deriving it from the raw log.