Closed tkam8 closed 7 months ago
Created [CONTCNTR-4067] for internal tracking.
@tkam8 dev build with fix quay.io/f5networks/k8s-bigip-ctlr-devel:41639a9b3164a240a136daac56b2a8109802d4c8.Could you please provide your feedback with this image.
Closed as completed.
Setup Details
CIS Version : 2.13.1
Build: f5networks/k8s-bigip-ctlr:latest
BIGIP Version: Big IP 15.1.8
AS3 Version: 3.45
Agent Mode: AS3/CCCL
Orchestration: K8S/OSCP
Orchestration Version: v1.21.5
Pool Mode: Nodeport
Additional Setup details: calico
Description
Unable to get LB to both pools in the separate namespaces.
Steps To Reproduce
1) Deploy ab vs with differing weights 2) Test with client hitting the vip 3) Notice that only the first rule is hit, only the alternatebackend is displayed
Expected Result
I get LB to both pools
Actual Result
I only get LB to the alternate backend pool
Diagnostic Information
vs
Observations (if any)
If I create virtualServer just for each of the namespaces separately, I can access the backend pods correctly.
In my LTM policy I see 3 identical rules, is this normal?:
vs_bgcafe_examplecom vs_bgcafe_example_com_nginxsvcbg_80_blue_bgcafe_example_com vs_bgcafe_example_com_nginxsvcbg_80_green_bgcafe_example_com
each with conditions:
HTTP Header full string named 'host' is 'bgcafe.example.com:443', or 'bgcafe.example.com' at request time.
and logs:Log message 'a/b pool' at request time.
bigip ltm logs show below that only the first rule gets hit I believe: