On portal UI, it shows sync successfully, but the new created application is not really sync to the gateway, and the gateway log shows following error message:
Unable to obtain HTTP response from https://xxxxxxxxxxxx:9443/sync/v1/applications?identifier=xxxxxxx: Received fatal alert: internal_error
Release : 9.4
Component : API GATEWAY
Dump pssg log on portal server:
docker service logs portal_pssg > pssg.log 2>&1
Search in pssg.log with the complete uri in the gateway error message, compare the timestamp in pssg.log and in the gateway log, you would see the pssg finish the sync task after the failure on gateway server. That's why it shows sync successfully on portal side, but the application data is not really sync to the gateway.
It indicates the portal is too slow to finish the sync task in time.
option 1: improve the portal performance
option 2: increase the interval for the portal sync application schedule task, and increase the read timeout of the route assertions in the scheduled task policy
Note that there are 2 route assertions in the policy, so we might evaluate the time as below,
task process time = timestamp of sync finish in pssg log - timestamp of failure in ssg log + current read timeout of the route assertion
then we could set the read timeout of each route assertion as,
read timeout = task process time + 1 minute
And set the interval of portal sync application schedule task as,
task interval = (read timeout) * 2 + 2 minutes
If the error message occurs again, repeat above