Steffen, Please find answers in-line. On 6/27/07, Steffen Grunewald <steffen.grunewald@xxxxxxxxxx > wrote:
I'm looking for a Howto document (or two):
http://gluster.org/docs/index.php/GlusterFS_User_Guide How to convert a simple pool of storage servers into a glusterfs pool?
(Suppose you've got a collection of data, which someone has stored on a bunch of storage servers, each with its own RAID. Now we want to unify the namespace, and access all of the files through glusterfs.) How to do that with "poor man's RAID-1" (suppose the storage servers come in pairs, storing the same set of data on each side of the mirror, while the access is read-only[!])? (I *know* that the ideal way would be to start with an empty filesystem, but with more than a dozen TB of data I cannot afford that right now.)
Glusterfs now has self-heal. It does not force users to start with empty filesystems. BTW, what's the state of 1.3.0? I know that there's a glusterfs-1.3.0-pre4.2
in qa-releases/, how does this relate to the current SVN? Would I be better off building from SVN? Is the debian/ tree up-to-date enough to dpkg-buildpackage?
-- Gowda (benki)
