[ClusterLabs] Feedback wanted: changing "master/slave" terminology

[Ken Gaillot]

On Wed, 2018-01-17 at 11:19 +0100, Kristoffer Grönlund wrote:
> Ken Gaillot <kgaillot at redhat.com> writes:
> 
> > 
> > For Pacemaker 2, I'd like to replace the <master> resource type
> > with
> > <clone stateful="true">. (The old syntax would be transparently
> > upgraded to the new one.) The role names themselves are not likely
> > to
> > be changed in that time frame, as they are used in more external
> > pieces
> > such as notification variables. But it would be the first step.
> > 
> > I hope that this will be an uncontroversial change in the
> > ClusterLabs
> > community, but because such changes have been heated elsewhere,
> > here is
> > why this change is desirable:
> > 
> 
> I agree 100% about this change. In Hawk, we've already tried to hide
> the
> Master/Slave terms as much as possible and replace them with
> primary/secondary and "Multi-state", but I'm happy to converge on
> common
> terms.
> 
> I'm partial to "Promoted" and "Started" since it makes it clearer
> that
> the secondary state is a base state and that it's the promoted state
> which is different / special.
> 
> However, can I throw a wrench in the machinery? When replacing the
> <master> resource type with <clone stateful="true">, why not go a
> step
> further and merge both <master> and <clone> with the basic
> <primitive>?
> 
> <primitive replicas="2"> => clone
> <primitive replicas="2" stateful="true"> => master
> 
> or for groups,
> 
> <group replicas="2" stateful="true">
> 
> I have never understood the usefulness of separate meta-attribute
> sets
> for the <clone> and <primitive> nodes.

I can see the point, but I do like having <clone> separate.

A clone with a single instance is not identical to a primitive. Think
of building a cluster, starting with one node, and configuring a clone
-- it has only one instance, but you wouldn't expect it to show up as a
primitive in status displays.

Also, there are a large number of clone meta-attributes that aren't
applicable to simple primitives. By contrast, master adds only two
attributes to clones.

>From the XML perspective, I think the current approach is logically
structured, a <clone> wrapped around a <primitive> or <group>, each
with its own meta-attributes.
-- 
Ken Gaillot <kgaillot at redhat.com>