smithp01 at mcrcr6.med.nyu.edu wrote:
: We currently keep local databases and plan to continue to do so. Howver,
: at this point we are about to run out of disk space for GenBank (maybe
: sometime in the next 1-2 releases). What we will need to do is to split GB
: over several disks. Is there a way to do this?
I'm not sure I understand either this question or the original post, but
it is possible to split genbank (or any database) over as many disks as
you like. Move the individual datafiles under the genbank subdirectory
to any other directory on any disk you have, then make symbolic links
from inside the genbank directory to point to the new location, like:
/usr/local/gcgdata/genbank/gb_est5.ref -> /home/gcg/gcgdata/gb_est5.ref
In UNIX, the command line would be:
ln -s /home/gcg/gcgdata/gb_est5.ref
which should be executed from within the genbank subdirectory. After you
set these up for as many of the large files as necessary for your disk
space requirements, subsequent updates of genbank automatically follow
the links during installation. In my example, /usr/local/ and /home/
are separate disks.
A better solution would be to install a disk array and stripe the data
across multiple disks (which behave as a single large file system).
That would also significantly improve performance, since the search
would be spread over multiple disks simultaneously.
As to the original post, I must be missing something, but how could
you have GCG without the database local?
David D. Womble Phone: 313-577-2374
Center for Molecular Medicine and Fax: 313-577-6200
Genetics, Wayne State University E-mail: dwomble at cmb.biosci.wayne.edu
5047 Gullen Mall, Detroit, MI 48202 http://cmmg.biosci.wayne.edu/