Cleaning Up Mismatched Entries in the LB Controller Pod Database for VMware/Openstack
(Doc ID 3053644.1)
Last updated on JANUARY 27, 2025
Applies to:
Oracle Communications Cloud Native Core - 5G - Version 2.22.4 and laterInformation in this document applies to any platform.
This is applicable for OCCNE version 22.3.x onwards
It is important to understand the basics of the working of the OCCNE-LBVM solution. Two important things to note are:
The loadbalancer services are accessible if the entry of the service is present inside the file /etc/haproxy/haproxy.cfg inside the Active LBVM. Also, the interface corresponding to the service should be attached to the same Active LBVM.
The occne-lb-controller-server pod's database is a mere record of what loadbalancer services are attached to the LBVM. If any entry is missing from the database but attached to the LBVM and present inside /etc/haproxy/haproxy.cfg file of the LBVM, then also the loadbalancer service should work fine. This scenario might cause other issues including but not limited to the messing up of the occne-lb-controller-pod's database.
Symptoms
When the database inside the occne-lb-controller-server pod does not match with the actual LoadBalancer services list, this procedure is used to clean up the mismatched entries in the occne-lb-controller pod's database.
This can be seen in the logs
{'time':'2024-10-10 11:38:52,832', 'name':'BGP', 'level':'INFO', 'message':'Service with ip xx.xxx.xxx.xxx still exists', 'pathname':'/create_lb/lbCtrl/get_bgp_external_ip.py:137'}
{'time':'2024-10-10 11:38:52,833', 'name':'BGP', 'level':'INFO', 'message':'Service IP xx.xxx.xxx.xxx still in use, IP will not be removed.',
Changes
No changes were done.
Cause
To view full details, sign in with your My Oracle Support account. |
|
Don't have a My Oracle Support account? Click to get started! |
In this Document
| Symptoms |
| Changes |
| Cause |
| Solution |