<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Sat, Aug 19, 2017 at 4:27 PM, Csaba Henk <span dir="ltr"><<a href="mailto:chenk@redhat.com" target="_blank">chenk@redhat.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Niels,<br>
<span class=""><br>
On Fri, Aug 11, 2017 at 2:33 PM, Niels de Vos <<a href="mailto:ndevos@redhat.com">ndevos@redhat.com</a>> wrote:<br>
> On Fri, Aug 11, 2017 at 05:50:47PM +0530, Ravishankar N wrote:<br>
</span>[...]<br>
<span class="">>> To me it looks like fadvise (mm/fadvise.c) affects only the linux page cache<br>
>> behavior and is decoupled from the filesystem itself. What this means for<br>
>> fuse is that the 'advise' is only to the content that the fuse kernel<br>
>> module has stored in that machine's page cache. Exposing it as a FOP would<br>
>> likely involve adding a new fop to struct file_operations that is common<br>
>> across the entire VFS and likely won't fly with the kernel folks. I could<br>
>> be wrong in understanding all of this. :-)<br>
><br>
> Thanks for checking! If that is the case, we need a good use-case to add<br>
> a fadvise function pointer to the file_operations. It is not impossible<br>
> to convince the Linux VFS developers, but it would not be as trivial as<br>
> adding it to FUSE only (but that requires the VFS infrastructure to be<br>
> there).<br>
<br>
</span>Well, question is: are strategies of the caching xlators' mapping well to<br>
the POSIX_FADV_* hint set? Would an application that might run on<br>
a GlusterFS storage backend use fadvise(2) anyway or would fadvise calls<br>
be added particularly to optimize the GlusterFS backed scenario?<br></blockquote><div><br></div><div>+Pranith, +ravi.</div><div><br></div><div>If I am not wrong, afr too has strategies like eager-locking if writes are sequential. Wondering whether afr can benefit from a feature like this.</div><div> <br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Because if usage of fadvise were specifically to address the GlusterFS<br>
backend -- either because of specifc semantic or specific behavior --, then<br>
I don't see much point in force-fitting this kind of tuning into the fadvise<br>
syscall. We can just as well operate then via xattrs.<br>
<span class="HOEnZb"><font color="#888888"><br>
Csaba<br>
</font></span><div class="HOEnZb"><div class="h5">______________________________<wbr>_________________<br>
Gluster-devel mailing list<br>
<a href="mailto:Gluster-devel@gluster.org">Gluster-devel@gluster.org</a><br>
<a href="http://lists.gluster.org/mailman/listinfo/gluster-devel" rel="noreferrer" target="_blank">http://lists.gluster.org/<wbr>mailman/listinfo/gluster-devel</a><br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br><div class="gmail_signature" data-smartmail="gmail_signature">Raghavendra G<br></div>
</div></div>