Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1765780

Summary: [4.2] SDN cannot add egress IP to team interface
Product: OpenShift Container Platform Reporter: Eric Paris <eparis>
Component: NetworkingAssignee: Ricardo Carrillo Cruz <ricarril>
Networking sub component: openshift-sdn QA Contact: huirwang
Status: CLOSED NEXTRELEASE Docs Contact:
Severity: urgent    
Priority: urgent CC: aos-bugs, bbennett, huirwang, jolee, palonsor, ricarril, zzhao
Version: 4.2.0   
Target Milestone: ---   
Target Release: 4.2.z   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: 1746871 Environment:
Last Closed: 2020-05-08 19:11:58 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On: 1746871    
Bug Blocks: 1746870    
Attachments:
Description Flags
sdn log none

Comment 1 Ricardo Carrillo Cruz 2020-01-15 08:56:04 UTC
Apparently this made it on 4.2 during the 4.2/4.3 split:

https://github.com/openshift/sdn/pull/27

Comment 4 huirwang 2020-01-16 08:42:22 UTC
Created attachment 1652669 [details]
sdn log

Comment 8 Ricardo Carrillo Cruz 2020-01-17 09:20:08 UTC
Yes, can you please recreate this and let me know the env details?

Thanks

Comment 9 huirwang 2020-01-18 02:29:15 UTC
I did't remove the env, there is something wrong with the vsphere lab. Once it is back, I will check the cluster works or not, if not then will create a new one.

Comment 11 Ricardo Carrillo Cruz 2020-02-03 12:31:00 UTC
Hi there

I'm sorry but was out of the office when you posted here the environment.
Can you please reproduce again and paste kubeconfig link ?

Thanks

Comment 13 Ricardo Carrillo Cruz 2020-02-04 09:54:59 UTC
Can you please let me know how to login to the nodes?

[ricky@ricky-laptop ~]$ ssh core.76.13 -i ~/.ssh/openshift-dev.pem 
Received disconnect from 139.178.76.13 port 22:2: Too many authentication failures
Disconnected from 139.178.76.13 port 22
[ricky@ricky-laptop ~]$ ssh root.76.13 -i ~/.ssh/openshift-dev.pem 
Received disconnect from 139.178.76.13 port 22:2: Too many authentication failures
Disconnected from 139.178.76.13 port 22
[ricky@ricky-laptop ~]$

Comment 16 Ricardo Carrillo Cruz 2020-02-04 12:06:42 UTC
Was the node really rebooted?

I killed openshift-sdn 'sdn' pods and after get them redeployed, the IP was added back to the interface:

[test@huirwang-vshpere421-72s68-rhel-0 ~]$ ip a show ServerB
3: ServerB: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
    link/ether 00:50:56:8b:a2:3b brd ff:ff:ff:ff:ff:ff
    inet 139.178.76.55/26 brd 139.178.76.63 scope global noprefixroute dynamic ServerB
       valid_lft 1396sec preferred_lft 1396sec
    inet 139.178.76.61/26 brd 139.178.76.63 scope global secondary ServerB:eip
       valid_lft forever preferred_lft forever
    inet6 fe80::be0:589c:a260:5ad2/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever
[test@huirwang-vshpere421-72s68-rhel-0 ~]$ 

------------

[ricky@ricky-laptop ~]$ oc get hostsubnet
NAME                               HOST                               HOST IP         SUBNET          EGRESS CIDRS        EGRESS IPS
compute-0                          compute-0                          139.178.76.13   10.129.0.0/23                       
control-plane-0                    control-plane-0                    139.178.76.11   10.128.0.0/23                       
huirwang-vshpere421-72s68-rhel-0   huirwang-vshpere421-72s68-rhel-0   139.178.76.55   10.130.0.0/23   [139.178.76.0/26]   [139.178.76.61]
huirwang-vshpere421-72s68-rhel-1   huirwang-vshpere421-72s68-rhel-1   139.178.76.54   10.131.0.0/23

Comment 18 Ricardo Carrillo Cruz 2020-02-05 10:33:06 UTC
Yeah, this is odd.
The only thing I can think of is that team interfaces lifecycle differs from 4.3 to 4.2 due to maybe different kernel or packages.

I'll try to compare on both versions.

Comment 20 Ricardo Carrillo Cruz 2020-02-06 12:06:09 UTC
You can tear it down for now, I need to compare kernel and network manager versions between 4.2 and 4.3.

Thanks

Comment 21 Ben Bennett 2020-05-08 19:11:58 UTC
This is fixed in 4.3.