Hi Airbyte Community! Did anyone experience report...
# replication-troubleshooting
m
Hi Airbyte Community! Did anyone experience report generation slowness, or any other stream slowness in the last 2-3 weeks related to amazon seller partner?
k
There are a few recent and semi‑recent items in the docs and issues that are relevant to slowness or degraded behavior for the Amazon Seller Partner source: 1. Rate limiting / quota issues and retries - Users have reported 429 QuotaExceeded errors on report endpoints, which can cause retries and failed/very slow syncs. [[QuotaExceeded thread](https://discuss.airbyte.io/t/1922); [Rate limits docs](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#limitations--troubleshooting)] - The official docs note that report rate limits can differ from Amazon’s documentation and recommend: - Reducing
Period In Days
- Using Incremental sync (Append) - Running syncs ~every 24h to reduce the chance of rate limiting. [[Limitations & troubleshooting](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#limitations--troubleshooting)] 2. “Wait between requests” logic (known performance impact / bug) - There is a feature
Wait between requests to avoid fatal statuses in reports
that intentionally slows down certain streams to avoid report failures; it only applies to a few streams like
GET_AMAZON_FULFILLED_SHIPMENTS_DATA_GENERAL
,
GET_AFN_INVENTORY_DATA
, and
GET_FBA_ESTIMATED_FBA_FEES_TXT_DATA
. When enabled, syncs for those streams are significantly slower. [[Limitations & troubleshooting](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#limitations--troubleshooting)] - A GitHub issue describes that the
wait_to_avoid_fatal_errors
/ threshold decorator was effectively delaying each record, not just each report request, making some streams extremely slow (pages taking hours). [[wait_to_avoid_fatal_errors issue](https://github.com/airbytehq/airbyte/issues/48882)] 3. Recent connector changes that may affect behavior - v4.4.0 introduced “waiting between requests logic to avoid failed report requests”, which could slow down report streams by design. [[Changelog 4.4.0](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#changelog)] - v4.6.1 added configurable Financial Events Step Size and fixed date‑range handling; docs suggest smaller step sizes (e.g.
P1D
) to avoid timeouts/rate limits and larger sizes (e.g.
P14D
) when volumes are small. [[Changelog 4.6.1](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#changelog); [Performance considerations](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#performance-considerations)] - v4.5.0 / 4.5.x changed rate‑limiting behavior and later had follow‑up fixes; some users reported syncs failing and had to pin to 4.4.7 while a fix was being rolled out. [[v4.5.x issue](https://github.com/airbytehq/airbyte/issues/53167)] Based on the sources, there are known issues and changes that can cause slowness, especially around: - Rate limits and 429s on reports - The “wait between requests” /
wait_to_avoid_fatal_errors
logic - The new waiting/rate‑limit logic introduced around 4.4.x–4.5.x However, the knowledge sources do not explicitly mention a generalized, platform‑wide slowness affecting all users in the exact “last 2–3 weeks”; they only document the above issues and behavior changes. If you’re seeing new slowness: - Check which connector version you’re on and compare to the changelog around 4.4.x–4.6.x. [[Changelog](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#changelog)] - See if
Wait between requests to avoid fatal statuses in reports
is enabled, especially for the affected streams. - Review logs for 429 /
QuotaExceeded
or “Failed to retrieve the report … will be read during the next sync” messages. [[Limitations & troubleshooting](https://docs.airbyte.com/integrations/sources/amazon-seller-partner#limitations--troubleshooting)] The docs don’t provide more specific guidance than this, so if your case doesn’t match these patterns, there isn’t additional information in the current knowledge sources.