[Gluster-devel] Multiple NFS Servers (Gluster NFS in 3.x, unfsd, knfsd, etc.)

[Shehjar Tikoo]

Gordan Bobic wrote:
> Shehjar Tikoo wrote:
>> ----- "Gordan Bobic" <gordan at bobich.net> wrote:
>>> I'm not sure if this is the right place to ask, but Google has failed
>>> me and it is rather related to the upcoming Gluster NFS export 
>>> feature. What I'm interested in knowing is whether it will be 
>>> possible to run Gluster NFS export for glfs mounts while using knfsd 
>>> as per standard
>>> for exporting non glfs paths?
>>
>> With native NFS there'll be no need to first mount a glusterFS
>> FUSE based volume and then export it as NFS. The way it has been 
>> developed is that
>> any glusterfs volume in the volfile can be exported using NFS by adding
>> an NFS volume over it in the volfile. This is something that will become
>> clearer from the sample vol files when 3.0.1 comes out.
> 
> It may be worth checking the performance of that solution vs the 
> performance of the standalone unfsd unbound to portmap/mountd over 
> mounted glfs volumes, as I discovered today that the performance feels 
> very similar to native knfsd and server-side AFR, but without the 
> fuse.ko complications of the former and the buggyness of the latter 
> (e.g. see bug 186: 
> http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=186 - that bug 
> has been driving me nuts since before 2.0.0 was released)
> 
> I'd hate to see this be another wasted effort like booster when there is 
> a solution that already works.
> 
>> The answer to your question is, yes, it will be possible to export your
>> local file system with knfsd and glusterfs distributed-replicated volumes
>> with Gluster NFS translator BUT not in the first release.
> 
> See comment above. Isn't that all the more reason to double check 
> performance figures before even bothering?
> 
> In fact, I may have just convinced myself to acquire some iozone 
> performance figures. Will report later.
> 
>>> e.g. if /home is a glfs mounted volume and /usr/src is on a raw block
>>>
>>> device, will it be possible to have /home handled by the glfs NFS
>>> export while having /usr/src handled by the native knfsd?
>>>
>>
>> Also, you'll be able to export both using the Gluster NFS server and
>> not need the knfsd at all but that is irrelevant for you I suppose.
> 
> You mean use glusterfsd to export both glfs volumes and raw disk 
> volumes? Why on earth re-invent unfsd?
> 

It is not re-invention but a consequence of the translator stack
design in glusterfs. Stack the nfs xlator over storage/posix and what
you get is an architecture similar to unfsd's way of exporting local
directories.

-Shehjar

>> The NFS translator right now does not make it possible for
>> the knfsd and Gluster NFS to co-exist on the same box(even if the above
>> options are used) but it is not hard to do and I will include this option
>> in one of the subsequent releases.
> 
> Let me get some iozone figures on unfsd in this new performance enhanced 
> way vs. server-side AFR glfs client. It may end up saving you some effort.
> 
>>> One possible work-around I can think of is to have one daemon listen
>>> for NFS connections on TCP and the other on UDP, but this is a bit lame.
>> BTW, we're only supporting NFS over TCP in the upcoming NFS translator.
> 
> Hmm... I'm not sure of that's a good thing or a bad thing...
> 
> Gordan
> 
> 
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel