error proxying a service deployed behind the firewall

Hello,

I am trying to proxy a backend URL which is deployed in a customer network and it is giving me the below error

{"fault":{"faultstring":"The Service is temporarily unavailable","detail":{"errorcode":"messaging.adaptors.http.flow.ServiceUnavailable"}}}

Is it due to any network settings or firewall? if so how do i solve this?

Thanks in advance,

Prashanthi

Solved Solved
0 3 306
1 ACCEPTED SOLUTION

adas
Participant V

Yes, its mostly because of network setting or firewall on the target backend side. To confirm this, can you replace the target with any public url like http://httpbin.org or even http://google.com.

Its mostly because of firewall settings on the target side. If so, you need to understand the restrictions from your network admin and if you are a paid Apigee Edge customer (either cloud or on-premise) you can ask your network admin to whitelist the message processor IPs so that you can connect to the backend target. If you are a free customer, it would not be possible since we do not guarantee fixed IPs for our free customers.

View solution in original post

3 REPLIES 3

adas
Participant V

Yes, its mostly because of network setting or firewall on the target backend side. To confirm this, can you replace the target with any public url like http://httpbin.org or even http://google.com.

Its mostly because of firewall settings on the target side. If so, you need to understand the restrictions from your network admin and if you are a paid Apigee Edge customer (either cloud or on-premise) you can ask your network admin to whitelist the message processor IPs so that you can connect to the backend target. If you are a free customer, it would not be possible since we do not guarantee fixed IPs for our free customers.

@arghya das I understand that this can't be guaranteed, but could you give us some idea of whether IPs for a free customer would be likely to change very frequently (say, every few hours) or whether we could reasonably expect it to stay the same for a few days?

@David Warwick McIntyre In the current model, the IPs do not change that often, maybe every few weeks but going forward that can change. If we were to scale automatically, the Edge IPs that actually connect to your backend target endpoint can change frequently, so that's why I don't think its a good idea to build a solution around that assumption. But in today's model the IPs do not change as frequently. But when they do, there won't be any prior notification to the trial pod customers and org admins. I hope that answers your question.