Re: old NFS stripped out in a few days

On 12/16/2014 7:04 PM, Rick Macklem wrote:
> Bryan Drewery wrote:
>> On 12/16/2014 4:58 PM, Rick Macklem wrote:
>>> FYI, I am planning on stripping the old NFS code out of head
>>> on about Mon. Dec 22. This has been discussed before and most
>>> seemed to be in favour of it.
>>>
>>> If you see a big problem with this, please email soon with
>>> your concerns.
>>>
>>> rick
>>
>> Mind clarifying briefly the impact of this? Does the oldnfs code
>> offer a
>> single benefit over the current NFS code?
>>
> Well, I remember one person reporting that they still use the old
> NFS server and that they had problems with the new one.
> 
> Unfortunately, these servers were used in production and they didn't
> have time to update them or try and isolate what problem(s) they
> experienced with the new server. No one else has reported problems
> with the new code that they avoid with the old code. (One other
> site has a lot of local patches for the old NFS server, but I
> think they will just have to port those to the new server if/when
> they want to upgrade to FreeBSD11.)

Thanks.

> 
> Unless you use "-o" on nfsd to run the old server or do
> "mount -t oldnfs ..." to use the old client, there shouldn't
> be any impact, since you aren't using the old NFS client/server.
> 
> If you try and do "nfsd -o" after it is removed, nfsd replies that
> the server isn't available and doesn't start.
> For "mount -t oldnfs ...", it fails after the code is removed.
> 
> I, personally, don't care if it removed, but others have suggested
> it (I suspect to reduce "code bloat" and the fact keeping it means
> maintaining two NFS subsystems.)

Yes, let's not rehash that. If it's been decided then let's stick to it.
No sense having dual stacks if there's no benefits.

> 
> rick
> 
>> --
>> Regards,
>> Bryan Drewery
>>
>>


-- 
Regards,
Bryan Drewery