[pve-devel] [PATCH qemu-server] fix #3588: helper: consider NIC count for config-specific timeout

[Fiona Ebner]

Am 15.11.24 um 11:00 schrieb Hannes Laimer:
> There have been some reports about `qm start` timeouts on VMs that have a
> lot of NICs assigned.
> This patch considers the number of NICs when calculating the config-specific
> timeout. Since the increase in start time is linearly related to the number
> of NICs, a constant timeout increment per NIC was chosen.
> 
> Signed-off-by: Hannes Laimer <h.laimer at proxmox.com>

Acked-by: Fiona Ebner <f.ebner at proxmox.com>

> ---
>  PVE/QemuServer/Helpers.pm | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/PVE/QemuServer/Helpers.pm b/PVE/QemuServer/Helpers.pm
> index 0afb6317..99cb3ab8 100644
> --- a/PVE/QemuServer/Helpers.pm
> +++ b/PVE/QemuServer/Helpers.pm
> @@ -167,6 +167,14 @@ sub config_aware_timeout {
>  	$timeout = 150;
>      }
>  
> +    # Some testing showed that adding a NIC increased the start time by ~450ms
> +    # consistantly across different NIC models, options and already existsing

s/consistantly/consistently/
s/existsing/existing/

> +    # number of NICs.
> +    # So, 10x that to account for any potential system differences seemed
> +    # reasonable. 

trailing whitespace here

Nit: Maybe also (or alternatively) have a sentence condensing the user
reports with some "real-life" values?
User in bug report: 20+ -> 40-50 seconds
Forum user 1: 25 -> 45 seconds
Forum user 2: 17 -> 42 seconds

The 5 seconds per NIC seems good in light of those too :)

> +    my $nic_count = grep { /^net/ } keys %{$config};

I'd also match for the number(s), so we don't match some other future
"netXYZ" option. E.g. "net-defaults" is on the radar (i.e. in a branch
of mine :P) already for the future guest profiles feature.

Nit: I'd prefer having an explicit scalar() in such cases.

> +    $timeout += $nic_count * 5;
> +
>      return $timeout;
>  }
>