| [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] |
All, I read all the responses to my last e-mail and I was obviously unclear. I was talking about the ogfs_jadd process of growing a ogfs cluster from N nodes to M nodes. My understanding of the process today: At mkfs.ogfs time the number of journals (N) is defined via the -j N option. For the cluster to be extended from N nodes to M nodes requires the creation of M-N additional journals via ogfs_jadd. Since journals have been tightly associated with journal sub-pools, I _assume_ that ogfs_jadd will have to be modified to understand the new external journal mechanisms. Note: I don't think ogfs_jadd works today, so this may be a non-trivial effort. Secondly, after ogfs_jadd is modified and works, I was thinking it might make sense to invoke ogfs_jadd from a EVMS FSIM. Ram, possibly you could say if that is appropriate or not. (i.e. Does EVMS know when nodes are being added to a cluster?) Greg -- Greg Freemyer ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Opengfs-devel mailing list Opengfs-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opengfs-devel