Gluster users - Jan 2009 - GlusterFS-1.4 to become GlusterFS-2.0

Dear Community,
Since the 1.3 release, we have come a long way to reach this level of code
maturity.
We are preparing to call the next release as the 2.0.0 instead of 1.4.0. There
are
real technical reasons to do so. We want to sincerely apologize for this
confusion.

REASONS FOR THE VERSION CHANGE:
* Significant architectural enhancements and code-rewrites since v1.3.X.
   - AFR rewrite, non-blocking IO, binary protocol and single address mode
* Current code base will remain as the foundation for upcoming major releases,
   featuring hot-management, active-heal, defrag, snapshot, NDMP and cloning.
* New features:
   - Distribute: elastic hash based clustering
   - Apache / Lighttpd embeddable
   - Distributed BerkeleyDB
   - Atomic writes in AFR
   - High-Availability translator
   - NUFA translator
* Improved performance and scalability, particularly for small files.
* Changes towards simplification
   - "distribute" supersedes complicated "unify"
   - "AFR" renamed as "replicate"
   - Simpler documentation using "distribute", "replicate"
and "stripe"
   - Standardized volume templates for most common use cases

2.0.0rc1 will be available in a week. Before we call it 2.0.0, we want to
to achieve the following:
* Seamless migration for 1.3 users
* Clean and complete documentation
* Binary packages for GNU/Linux (rpm/deb), FreeBSD, Mac OS X and Solaris
(server)
* And most importantly the community blessing for 2.0 stable release.

Happy Hacking!
--
GlusterFS Team

I'm about to switch from 1.3.11 to the latest rc of 1.4... how much  
different will 2.0 be from 1.4? Or will 2.0 effectively be the latest  
version of 1.4 rc7 + bug fixes etc? Or will it be a totally different  
branch?

I suppose the best course of action would be to wait until 2.0 but I  
really need that io-cache leak fix now.


Dan Parsons


On Jan 5, 2009, at 3:59 AM, Anand Babu Periasamy wrote:
> Dear Community,
> Since the 1.3 release, we have come a long way to reach this level  
> of code maturity.
> We are preparing to call the next release as the 2.0.0 instead of  
> 1.4.0. There are
> real technical reasons to do so. We want to sincerely apologize for  
> this confusion.
>
> REASONS FOR THE VERSION CHANGE:
> * Significant architectural enhancements and code-rewrites since  
> v1.3.X.
>  - AFR rewrite, non-blocking IO, binary protocol and single address  
> mode
> * Current code base will remain as the foundation for upcoming major  
> releases,
>  featuring hot-management, active-heal, defrag, snapshot, NDMP and  
> cloning.
> * New features:
>  - Distribute: elastic hash based clustering
>  - Apache / Lighttpd embeddable
>  - Distributed BerkeleyDB
>  - Atomic writes in AFR
>  - High-Availability translator
>  - NUFA translator
> * Improved performance and scalability, particularly for small files.
> * Changes towards simplification
>  - "distribute" supersedes complicated "unify"
>  - "AFR" renamed as "replicate"
>  - Simpler documentation using "distribute",
"replicate" and "stripe"
>  - Standardized volume templates for most common use cases
>
> 2.0.0rc1 will be available in a week. Before we call it 2.0.0, we  
> want to
> to achieve the following:
> * Seamless migration for 1.3 users
> * Clean and complete documentation
> * Binary packages for GNU/Linux (rpm/deb), FreeBSD, Mac OS X and  
> Solaris (server)
> * And most importantly the community blessing for 2.0 stable release.
>
> Happy Hacking!
> --
> GlusterFS Team
>
>
>
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>

>
> I'm about to switch from 1.3.11 to the latest rc of 1.4... how much
> different will 2.0 be from 1.4? Or will 2.0 effectively be the latest
> version of 1.4 rc7 + bug fixes etc? Or will it be a totally different
> branch?
>
It will be 1.4 rc7 + bug fixes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20090105/024eec37/attachment.html>

Anand Babu Periasamy wrote:
> 2.0.0rc1 will be available in a week. Before we call it 2.0.0, we want to
> to achieve the following:
> * Binary packages for GNU/Linux (rpm/deb), FreeBSD, Mac OS X and Solaris 
> (server)
As for the RPM spec file, are you amenable to some alterations?  IMO it 
would be preferable to build three packages instead of one, 
glusterfs-common, glusterfs-client and glusterfs-server, following the 
common way client/server packages are built in RHEL/CentOS.

I remember Matt's spec file used to be this way, but it looks like you 
guys specifically changed it:

* Sat Apr 19 2008 Amar Tumballi <amar at zresearch.com> - 1.3.8pre6
- Merged common, client and server packages into one package.

Is there a reason for that?

Note:  I would rather any changes to the spec file NOT hold up 2.0, so 
if you would rather hold off on this until after 2.0 (if you think it's 
valid), please do so.
> * And most importantly the community blessing for 2.0 stable release.
You have mine.  :)

-- 

-Kevan Benson
-A-1 Networks