<@U04S8PLJWMT> Pods are fine: source-gcs-check-0fa...
# replication-ask-ai
v
@kapa.ai Pods are fine: source-gcs-check-0fa76e48-6559-4bff-91ad-4eb066c14a60-0-giwfv 0/2 Completed 0 4m31s source-gcs-check-ca3cfcab-8682-429a-9731-602912315022-0-ujppi 0/2 Completed 0 9m lovelocal@Lovelocals-MacBook-Pro airflow-data-pipelines % But in UI I get error - An unknown error occurred. (HTTP 504)
k
Thinking completed An HTTP 504 (Gateway Timeout) error in the UI—especially when the check or discovery pod itself runs successfully and marks as `Completed`—is typically caused by an intermediate ingress, reverse proxy, or load balancer terminating the connection before the Airbyte server finishes returning the response [Ingress; Multi-property syncs only process first property in version 2.9.20]. Timeout values that are too short on the Airbyte server ingress commonly lead to 504 errors in the UI during connector checks, source/destination creation, or schema discovery [Ingress]. By default, many ingress controllers (such as NGINX) have a default proxy timeout of 60 seconds [Multi-property syncs only process first property in version 2.9.20]. How to resolve this Depending on how your Airbyte ingress or proxy is deployed: 1. If using Kubernetes NGINX Ingress Increase the proxy timeout annotations on your Airbyte
Ingress
resource [Multi-property syncs only process first property in version 2.9.20]:
Copy code
metadata:
  annotations:
    nginx.ingress.kubernetes.io/proxy-connect-timeout: "600"
    nginx.ingress.kubernetes.io/proxy-send-timeout: "600"
    nginx.ingress.kubernetes.io/proxy-read-timeout: "600"
2. If using AWS ALB Increase the idle timeout on the load balancer [Manual ingress examples]:
Copy code
metadata:
  annotations:
    alb.ingress.kubernetes.io/load-balancer-attributes: idle_timeout.timeout_seconds=600
3. If using Docker Compose with Airbyte's basic auth proxy If you are running via Docker rather than Kubernetes, you may need to increase the timeout environment variable in your
.env
file [Forum Post: Airbyte Proxy - Nginx 504 Gateway Time-out EC2 Docker deployment]:
Copy code
BASIC_AUTH_PROXY_TIMEOUT=600
👍 0 button 👎 0 button ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
v
@kapa.ai Changes were done, it's working fine for other connection but a connection which has 15 streams in it, it is failing with - An unknown error occurred. (HTTP 524)
k
Thinking completed I am sorry, but the provided knowledge sources do not contain enough specific information about an
HTTP 524
error occurring during connection setup or stream checks in Airbyte. The sources mention an
Unexpected 524 Error
solely in the context of an API failure within the Recurly connector (
source-recurly
), where it was flagged as an external API issue requiring connector handling [Forum Post: Recurly Unexpected 524 Error]. However, in standard networking, an HTTP 524 typically represents a "A Timeout Occurred" error from Cloudflare, which triggers when an HTTP connection to the origin web server succeeds, but the origin server does not send an HTTP response before the default Cloudflare timeout (typically 100 seconds). If your Airbyte deployment or domain sits behind Cloudflare proxying (the orange-cloud icon), discovering or verifying a connection with 15 streams may be taking longer than 100 seconds, causing Cloudflare to terminate the client request with a 524 error while the backend is still running. 👍 0 button 👎 0 button ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.