<p dir="ltr">Il 01 giu 2016 22:06, "Gmail" <<a href="mailto:b.s.mikhael@gmail.com">b.s.mikhael@gmail.com</a>> ha scritto:<br>
> stat() on NFS, is just a single stat() from the client to the storage node, then all the storage nodes in the same replica group talk to each other using libgfapi (no FUSE overhead)<br>
><br>
> conclusion, I’d prefer NFS over FUSE with small files.<br>
> drawback, NFS HA is more complicated to setup and maintain than FUSE.<br></p>
<p dir="ltr">NFS HA with ganesha should be easier than kernel NFS</p>
<p dir="ltr">Skipping the whole fuse stack should be good also for big files<br>
with nfs replication is made directly by gluster servers with no client involved?<br>
In this case would be possibile to split the gluster networks with 10gb used for replication and multiple 1gb bonded for clients.<br>
I can see only advantage for nfs over native gluster</p>
<p dir="ltr">One question: with no gluster client that always know on which node a single file is located, who is telling nfs where to find the required file? Is nfs totally distributed with no "gateway"/"proxy" or any centralized server? </p>