GitHub user deajan added a comment to the discussion: Bridge interface not
recognized for newly added host if they don't have a physical interface attached
Sorry, as I said, at one moment, I think I have mixed both while doing tests
creating/deleting bridges to find what the issue is.
For instance, br_npf has eno2 as slave interface, br_bgp0 has a vxlan as slave
interface.
So basically the test to see whether there is `eth*, bond*, team*, vlan*, em*,
p*p*, ens*, eno*, enp*, or enx*` in `/sys/devices/virtual/net/${bridge}/brif`
is not sufficient.
I would change that check to report successful if at least one slave different
than "vnet*" "or "vmbr*" is present (so we outrule running VM interfaces from
uplinks).
This would allow to have bridge slaves interfaces being VXLan, GRETAP, Geneve
or whatever fancy interfaces someone would like to use. It would also allow
Cloudstack to be future proof if someday a new ethernet driver naming comes out.
GitHub link:
https://github.com/apache/cloudstack/discussions/10804#discussioncomment-13004300
----
This is an automatically sent email for users@cloudstack.apache.org.
To unsubscribe, please send an email to: users-unsubscr...@cloudstack.apache.org
