[OCF]Standards for review?

[Lars Marowsky-Bree]

On 2003-10-20T15:49:01,
   Alan Robertson <alanr@unix.sh> said:

> What remains for us to do before releasing the resource agent for public 
> comments?

I've done the only remaining modification to the draft on my side now.
Sorry it took so long.

As per tytso's recommendation:
	
> It is true that this means that there isn't any way to return an
> error code for the "status" init action.  The current assumption
> is that the status action can never fail (since normally it's
> simply a matter of finding the pid file and checking to see
> whether or not the process exists).
...
> As for why things were done this way, the answer is to preserve
> compatibility with distribution behaviour at the time when the
> specification was drafted.  Either Red Hat, SuSE, or both (I
> don't remember, we could do digging historically if we *really*
> cared) was already using the existing exit return codes, and we
> decided it was better to adopt the already-existing scheme than
> to do something else that would have made that distribution
> instantly LSB non-compliant.
...
> Yes, redefining exit status codes probably doesn't make any
> sense.
> 
> For the current init scripts, it's not at all clear to me that
> we   really need to be able to return error status codes.
> However, in a cluster environment, I can easily see how being
> able to return errors would be useful.  Using a new action such
> as "monitor" as a cluster extension might be the easist way to
> solve this particular problem.

I've changed the draft to use "monitor" instead of "status" again and
committed it to CVS.

Did I miss anything else?

I think that should be about it.


Sincerely,
    Lars Marowsky-Brée <lmb@suse.de>

-- 
High Availability & Clustering	      \ ever tried. ever failed. no matter.
SUSE Labs			      | try again. fail again. fail better.
Research & Development, SUSE LINUX AG \ 	-- Samuel Beckett