redhat: adjust IP bonding to bond0 only #58
Conversation
It's possible that adding IPs to bond1 can create a scenario where networking isn't happy. Lets remove those accordingly
|
we need to add the tests for this too to confirm the fix |
It gets replaced by a prior formatting pass
packetnetworking/distros/redhat/templates/bonded/etc_sysconfig_network-scripts_ifcfg-bondX.j2
Outdated
Show resolved
Hide resolved
packetnetworking/distros/redhat/templates/bonded/etc_sysconfig_network-scripts_ifcfg-bondX.j2
Outdated
Show resolved
Hide resolved
…_network-scripts_ifcfg-bondX.j2 Co-authored-by: nuclearbob <6314562+nuclearbob@users.noreply.github.com>
…_network-scripts_ifcfg-bondX.j2 Co-authored-by: nuclearbob <6314562+nuclearbob@users.noreply.github.com>
…t-networking into rh-bond1-fixup-and-more
| @@ -1,5 +1,6 @@ | |||
| DEVICE={bond} | |||
| NAME={bond} | |||
| {{% if "{bond}" == "bond0" %}} | |||
There was a problem hiding this comment.
| {{% if "{bond}" == "bond0" %}} | |
| {{% if bond == "bond0" %}} |
There was a problem hiding this comment.
🤦 this is python formatted before it's handled with jinja2, so this is proper
|
Testing on actual machines went well except for n3.xlarge.x86, which shows problems on alma_8, rocky_8, and centos_8. alma and rocky install correctly with current osie, so I'm looking into what goes wrong when we use the new one. |
|
alma_8 is now broken with or without this, so let's go ahead and land it, and I'll keep working on alma. |
|
Oh, let's get a review from another team? Maybe @nshalman can recommend someone? |
|
Do we still want a review from another team on this, or is it okay to go in since it's a fix requested by another team and it's pretty small? Either way, I'd like to keep moving on this since it fixes issues in some of the plans we're enabling. |
It's possible that adding IPs to bond1 can create a scenario where
networking isn't happy. Lets remove those accordingly