What is the relation between GCP load balancer and Apigee X

[anonymous]

Hi experts,

I’m facing an issue with Apigee X and GCP load balancer. I have updated the timeout to 30 seconds for the GCP load balancer of Apigee project. Below the screenshot for load balancer settings

After that I have invoked an API having 45 seconds of timeout. Then I’m getting a response as below

On the other hand, I can see some success response in Apigee trace. Is it because by Apigee default timeout is 55 seconds?

So, My query is how the GCP load balancer and Apigee are dependent? Can someone share an explanation for this

Please share your valuable comments @dchiesa1

[debjitd18]

With my limited knowledge, the Load Balancer handles the traffic to the Apigee. If you can expand the timeout on the Load Balancer to greater than 30 seconds, you should not be seeing the time out error.

@dchiesa1 please correct if am wrong in providing my explanation for this issue. Or if you provide more of your insights will be helpful.

Thanks,

Debjit

[anonymous]

@devfied

So, My query is how the GCP load balancer and Apigee are dependent? Can someone share an explanation for this

they aren’t related. I suspect if you create a simple API not running on Apigee, and connect it to a GCP LB, and that API takes 45 seconds to complete, you will experience the same behavior.

If the timeout on the LB is set lower than the backend, you should be seeing the LB timeout standard message (whatever that it).

[dknezic]

The load balancer doesn’t handle traffic directly to Apigee.

Depending on the approach, the load balancer will

(as per this case) forward traffic to a proxy/mig, and the proxy forwards the traffic to Apigee.

or if you’re using PSC between Apigee and the LB, the proxying is taken care of under the hood to Apigee

[dknezic]

Your LB timeout looks to be 30 seconds while your API Proxy is taking more than 30 seconds to return a response.