From: "Shan Shaji" <s.shaji@proxmox.com>
To: "Christian Ebner" <c.ebner@proxmox.com>, <pbs-devel@lists.proxmox.com>
Subject: Re: [PATCH proxmox] fix #7329: proxmox-network-api: add missing `Auto` network config method
Date: Fri, 13 Mar 2026 18:36:59 +0100 [thread overview]
Message-ID: <DH1U4VO2TY8B.VC0VV2VY0MPP@proxmox.com> (raw)
In-Reply-To: <829e80ef-1274-4ac7-9ebf-1c8fa0b55108@proxmox.com>
On Thu Mar 12, 2026 at 3:57 PM CET, Christian Ebner wrote:
> Thanks for the patch, gave this a quick spin but ran into an issue.
>
> If I do have a network config containing e.g.:
> ```
> auto nic1
> iface nic1 inet static
> address 172.16.0.18/24
>
> iface nic1 inet6 auto
> ```
> and I go about to edit this NIC via the WebUI, e.g. adding a comment,
> the suggested diff would drop this, suggesting the following diff:
>
> ``` auto nic1
> iface nic1 inet static
> address 172.16.0.18/24
> -
> -iface nic1 inet6 auto
> +#comment
> ```
Hi Chris, Thank you very much for catching that.
After digging into this a bit more, it looks like this happens in the interface
update handler on the API side. If the inet6 address family's cidr6/gateway6 values are
not set, interface.method6 value is mutated with `NetworkConfigMethod::Manual`:
```
if interface.cidr6.is_some() || interface.gateway6.is_some() {
interface.method6 = Some(NetworkConfigMethod::Static);
} else {
interface.method6 = Some(NetworkConfigMethod::Manual);
}
```
Later, when writing the config back, if the method6 value is `NetworkConfigMethod::Manual`
and there are no IPv6 specific attributes present, the inet6 stanza gets skipped:
```
fn write_iface(iface: &Interface, w: &mut dyn Write) -> Result<(), Error> {
[...]
if let Some(method6) = iface.method6 {
let mut skip_v6 = false; // avoid empty inet6 manual entry
if iface.method.is_some()
&& method6 == NetworkConfigMethod::Manual
&& iface.comments6.is_none()
&& iface.options6.is_empty()
{
skip_v6 = true;
}
if !skip_v6 {
writeln!(w, "iface {} inet6 {}", iface.name, method_to_str(method6))?;
write_iface_attributes_v6(iface, w, method6)?;
if iface.method.is_none() {
// only write common attributes once
write_iface_attributes(iface, w)?;
}
writeln!(w)?;
}
}
Ok(())
}
```
> Can this easily be avoided?
I was able to avoid this by changing the API-side update logic so that
method6 is only set to `NetworkConfigMethod::Manual` if interface.method6 is
not already set. I still want to do a bit more testing to make sure i am
not missing anything.
next prev parent reply other threads:[~2026-03-13 17:37 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-11 10:11 Shan Shaji
2026-03-12 14:57 ` Christian Ebner
2026-03-13 17:36 ` Shan Shaji [this message]
2026-03-16 15:23 ` Shan Shaji
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=DH1U4VO2TY8B.VC0VV2VY0MPP@proxmox.com \
--to=s.shaji@proxmox.com \
--cc=c.ebner@proxmox.com \
--cc=pbs-devel@lists.proxmox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.