Hello, At remote locations, we often use Big IP F5 VPN client to connect to our VPN and then do other activities.
Whilst on VPN, we tried connecting to our Kafka Service on GCP to test pub/sub. It appears that there's always a network exception even with request.timeout.ms set as 120s for producer/consumer. And when console producer receives those exception, we hit "Return", and the consumer received the messages in reverse order. When we tried outside VPN, it appear to work without any tuning to default request.timeout.ms. We can understand the latency whilst on VPN, but the message ordering was quite confusing. Has anyone experienced the same issue? I am not quite an expert in various network technologies, but it almost seems that the batching of the packets and delivery were all mixed up. Any help/suggestion is appreciated. Regards,
