Re: nfs-utils: support/nfs/rpcmisc.c

[Jeff Layton <jlayton@redhat.com>]

On Wed, 28 Jan 2009 13:05:39 -0500
Chuck Lever <chuck.lever@oracle.com> wrote:

> On Jan 28, 2009, at 11:37 AM, Jeff Layton wrote:
> > On Tue, 27 Jan 2009 15:20:02 -0500
> > Chuck Lever <chuck.lever@oracle.com> wrote:
> >> I'm considering creating a version of rpc_init() under utils/statd
> >> that strips out all of the unnecessary features, uses note() instead
> >> of xlot(), and then adds support for AF_INET6.
> >>
> >
> > Is there some reason to not just convert statd to use xlog()? It
> > seems like that would be a better long term solution.
> 
> I'm not sure what the advantage of note() vs. xlog() is.
> 
> Be it also noted that sm-notify uses yet another logging mechanism,  
> and I've already coded up patches to move it over to using note().   
> This is because I'm adding some IPv6-enabled RPC handlers that are  
> shared with statd...
> 
> And, the RPC libraries also write directly to stderr.  All of those  
> error messages are lost in the daemon case, but when using -Fd they  
> are useful, and appear on the controlling terminal.
> 
> > For a proof-of-concept sort of thing, that might be reasonable though.
> 
> 
> It could be an additional clean-up once I'm done with the IPv6  
> support.  But I need to look at both implementations to see if one is  
> more appropriate for all of nfs-utils.
> 

FWIW, I don't have any particular attachment to either logging scheme.
It just seems like we could benefit from some standardization in this
area.

Not a high-priority item, but something to keep in mind if you poking
around in the code anyway.

-- 
Jeff Layton <jlayton@redhat.com>