superseded: [PATCH storage v4 2/2] nfs: use rpcinfo as a showmount fallback for nfs3

[Maximiliano Sandoval <m.sandoval@proxmox.com>]

Fiona Ebner <f.ebner@proxmox.com> writes:

> Am 07.07.26 um 9:37 AM schrieb Maximiliano Sandoval:
>> Fiona Ebner <f.ebner@proxmox.com> writes:
>> 
>>> Am 06.07.26 um 4:37 PM schrieb Maximiliano Sandoval:
>>>> diff --git a/src/PVE/Storage/NFSPlugin.pm b/src/PVE/Storage/NFSPlugin.pm
>>>> index 56dbd9ec..4fe7695a 100644
>>>> --- a/src/PVE/Storage/NFSPlugin.pm
>>>> +++ b/src/PVE/Storage/NFSPlugin.pm
>>>> @@ -198,6 +198,11 @@ sub check_connection {
>>>>          # no exports are listed
>>>>          $cmd = $nfs_get_rpcinfo_command->($server, '4');
>>>>      } else {
>>>> +        # FIXME: PVE 10: Consider switching to rpcinfo altogether.
>>>> +        # There are some potential pitfalls:
>>>> +        # - Some NFSv3 servers might only support UDP (cf. RFC 1813, Section 2.3)
>>>> +        # - The two timeouts add up to a potential worst-case scenario of 20 seconds
>>>
>>> This is not a pitfall for to the switch to rpcinfo, but related to your
>>> current implementation with showmount+rpcinfo. It's rather easy to spot,
>>> so not sure it needs a comment in the code. I'd rather have it in the
>>> commit message.
>> 
>> Running into timeouts is a very common issue in Enterprise support, I do
>> have a strong preference for keeping it as explicit as possible.
>
> But please don't mention it in the pitfalls for the switch to rpcbind,
> but clearly state that it's about the current implementation. Otherwise,
> that is just confusing.

Superseded-by: https://lore.proxmox.com/all/20260707082332.111714-1-m.sandoval@proxmox.com/T/#u
-- 
Maximiliano