Re: [Gluster-devel] Proposal for 3.5: Make adding custom translators easy

[Anand Babu Periasamy]

We have had this discussion during our management framework and GUI design sessions. My suggestion is to separate the management framework in to a glusterd like daemon running on each node. They expose RESTful APIs. Client can be implemented in python like higher level scripting languages. RESTful APIs should also implement storage management (LVM, mkfs, RAID...), network management and other related node level commands. Next question is, does it make sense to leverage any other existing project for management framework or keep it limited to gluster operations alone. Unlike oVirt based management framework currently in development, I would like to see something simpler and light weight. No Java or JBoss dependency please :).

On Fri, Aug 16, 2013 at 12:15 PM, Justin Clift <address@hidden> wrote:

On 16/08/2013, at 8:11 PM, Harshavardhana wrote:
> On Fri, Aug 16, 2013 at 11:48 AM, Justin Clift <address@hidden> wrote:

<snip>

> As a separate project, any thoughts on us replacing the current CLI with
> a new one, based upon having a more complete RESTful API + a new CLI that
> does nothing but be an interface to it?
>
> In that case CLI can be completely re-written in python or any other language. This would eventually bring in more contributors to the project.

That's a very good point. :)

I'd only been thinking of it from the point of view of giving
us a "full" RESTful API that's equivalent to the CLI.  So
people could do some very in depth automation and integration
with it.

But having the CLI as Python does sound like it would be much
easier to develop + get contributors for too.

Proposal for 3.6 maybe? :)

+ Justin

--
Open Source and Standards @ Red Hat

twitter.com/realjustinclift

_______________________________________________
Gluster-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/gluster-devel

--
-ab

Imagination is more important than knowledge --Albert Einstein