0

I just created 2 VMs instance for my web service.
I bind two ports 80 and 8080 for different web service.
Firewall policy was allow tcp:80 and 8080
First, I created a http load balancer (80 port) everything was fine.
Then I created a tcp load balancer (8080 port) the health check always fail.

load balancer fail

Since the two web service will use the different URL :
eg :
80 port a[1-8].abc.com
8080 prot m[1-8].abc.com
What is the best solution for my case ?

2
  • There is not enough information in your question to provide an answer. Edit your question and provide details on the health check and the error. Jul 28, 2019 at 4:08
  • @JohnHanley just update the picture. thanks a lot!!
    – Eric Wu
    Jul 28, 2019 at 6:22

1 Answer 1

1

Assuming your health-check for 34.80.167.244:8080 is properly configured to test port 8080 with protocol HTTP, I would suggest you double-check the following:

  • As described here, the Network Load Balancer uses legacy health check and the source IP addresses for GCP probe systems are different. Make sure you create an ingress rule in the GCP firewall allowing the traffic from those addresses to your two VMs.
  • Verify that the web servers running on your two VMs are actively listening on port 8080. You may launch a curl or telnet command on port 8080 directly to the VMs.
  • If you are running local firewalls in your VMs, confirm that you allow the incoming traffic to port 8080.

If all the above is fine, you will have to provide more details about your configurations.

1
  • The part where I had to allow healthcheck in firewall was my culprit. I wish Google did a better job communicating that I have to add this myself. Thank man!
    – Lukas
    Dec 22, 2022 at 15:46

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .