Bug #161
closed
need replication mechanism for metacat
Added by Matt Jones about 24 years ago.
Updated almost 23 years ago.
Description
Metacat was intended to be a centralized index of metadata that was replicated
across several sites. This replication feature has not been designed or
implemented. It needs to accomplish the following:
1) mirror XML content among multiple sites
2) mirroring controlled by both source and destination metacats
3) locking issues resolved without chance of data loss
The presentation that Matt Jones gave at the SDSC annual meeting has an overview
of some potential architectures. See the PBI web site for the slides.
Work is in progress- Models have been submitted, refined and resubmitted for
comment. Work is currently in progress on some test implementations.
most of replication is done. The following is a list of what still needs to be
done:
1) add a flag in xml_replication to tell the server whether to copy it's files
out or not. This will allow for "one-way" replication.
2) create a better web interface to control replication
3) increase the error checking and error resolution capabilities.
Numbers 1 and 2 from the 12/19/2000 comments have been finished. Error checking
and resolution is now the primary issue remaining with replication.
Replication, including the replication of package information and the
xml_catalog table now works. Replication can be controlled through an html
interface called "replControl.html".
Original Bugzilla ID was 161
Also available in: Atom
PDF