Gluster users - Dec 2008 - Unify and single namespace

Hello all,

I have reviewed the docs, instructions, etc... but I do not seem to
get the "picture" of how the single metadata namespace would fit into,
for instance a 4-brick architecture ... of Gluster.

1.  The FAQ discusses that when the Unify Translator is enabled,
metadata is stored on the namespace brick.

 Metadata Storage - When using striping (unify), how/where is the
metadata kept? The metadata is stored on the namespace brick.

What does this mean?  We would need to dedicate a separate metadata
server (in addition to my storage nodes)?  But does not seem so, as
the installation binaries seem to only come in "server" and
"client"
form.

2.  I further read the Unify Translator doc -
http://www.gluster.org/docs/index.php/Understanding_Unify_Translator

and seems like the following states that the "namespace-child" needs
to be another volume defined in the spec, configured like another
server brick, but its content is used to manage namespace. QUESTION -
is this a dedicated server?


?	LINE 5: "option namespace namespace-child"
The option specifies that namespace-child should be used as a place
where the whole filesystem's namespace (ie, directory/file tree
structure) is maintained. Here 'namespace-child' needs to be another
volume which is already defined in the spec.
'namespace-child' is configured like another server brick, except that
it's content is used to manage namespace. It should not be listed
along with the other subvolumes.
?	LINE 6: "subvolumes child1 child2 child3 child4"
Here 'subvolumes' is an identifier for parser to tell that following
volumes are the 'children' nodes for current volume.
'subvolumes'
takes one or more volume names as argument. Here child1, child2,
child3, child4 are also named as storage nodes as these are the nodes
where actual data is storaged.

=====
Could someone point me to a site that potentially diagrams a summary
of a simple Gluster with Unify and AFR installed?

I found this site:
http://ftp.zresearch.com/pub/gluster/glusterfs/talks/Z/GlusterFS.pdf -
page 5 - shows the Unify stack in the client, but I still am unable to
conceive where the metadata - single namespace information - is
stored?

I apologize if my question is quite naive, but I'm a bit confused.

thanks in advanced,

aspasia.

Hi,
please find the comments inlined.

On Wed, Dec 3, 2008 at 4:59 AM, a s p a s i a <aspasia.sf at gmail.com>
wrote:
> Hello all,
>
> I have reviewed the docs, instructions, etc... but I do not seem to
> get the "picture" of how the single metadata namespace would fit
into,
> for instance a 4-brick architecture ... of Gluster.
>
> 1.  The FAQ discusses that when the Unify Translator is enabled,
> metadata is stored on the namespace brick.
>
>  Metadata Storage - When using striping (unify), how/where is the
> metadata kept? The metadata is stored on the namespace brick.
>
> What does this mean?  We would need to dedicate a separate metadata
> server (in addition to my storage nodes)?  But does not seem so, as
> the installation binaries seem to only come in "server" and
"client"
> form.

namespace brick can have glusterfs running as server exporting a brick to
store the whole directory structure of unify with 0 byte sized files (which
is called namespace cache). As the wiki says its similar to children of
unify (nodes mentioned in subvolumes section of unify) with the difference
that instead of actual data, namespace cache is stored on it.

> 2.  I further read the Unify Translator doc -
> http://www.gluster.org/docs/index.php/Understanding_Unify_Translator
>
> and seems like the following states that the "namespace-child"
needs
> to be another volume defined in the spec, configured like another
> server brick, but its content is used to manage namespace. QUESTION -
> is this a dedicated server?

Its not clear what do you mean by "dedicated server". All it needs is
to
have the filesystem which has enough space to hold namespace cache (which is
small compared to actual data) and which can support large number of inodes
(the whole directory structure of unify is replicated on this node with 0
byte sized files).

>
>
> ?       LINE 5: "option namespace namespace-child"
> The option specifies that namespace-child should be used as a place
> where the whole filesystem's namespace (ie, directory/file tree
> structure) is maintained. Here 'namespace-child' needs to be
another
> volume which is already defined in the spec.
> 'namespace-child' is configured like another server brick, except
that
> it's content is used to manage namespace. It should not be listed
> along with the other subvolumes.
> ?       LINE 6: "subvolumes child1 child2 child3 child4"
> Here 'subvolumes' is an identifier for parser to tell that
following
> volumes are the 'children' nodes for current volume.
'subvolumes'
> takes one or more volume names as argument. Here child1, child2,
> child3, child4 are also named as storage nodes as these are the nodes
> where actual data is storaged.
>
> =====>
> Could someone point me to a site that potentially diagrams a summary
> of a simple Gluster with Unify and AFR installed?
>
> I found this site:
> http://ftp.zresearch.com/pub/gluster/glusterfs/talks/Z/GlusterFS.pdf -
> page 5 - shows the Unify stack in the client, but I still am unable to
> conceive where the metadata - single namespace information - is
> stored?
>
> I apologize if my question is quite naive, but I'm a bit confused.
>
> thanks in advanced,
>
> aspasia.
>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users
>

regards,
-- 
Raghavendra G
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20081203/8aa13ec6/attachment.html>