On Thu, 08 Nov 2012, Giovanni Di Milia wrote:
> which is the quickest and easiest way to extract the collections
> configuration of WebSearch and apply it on another instance of
> Invenio?
If you think about replicating collection definitions and the tree, then
something like the following is enough:
On source node:
$ mysqldump --skip-opt --add-drop-table -u root -p invenio collection > c.sql
$ mysqldump --skip-opt --add-drop-table -u root -p invenio
collection_collection > cc.sql
On target node:
$ cat c.sql | /opt/invenio/bin/dbexec
$ cat cc.sql | /opt/invenio/bin/dbexec
Notes:
* Adding `--skip-opt' permits to easily grep away some wanted/unwanted
values from the dump.
* If the site name on the target node is different, then you may want to
amend name of your home collection (ID=1) to be the same as the target
site name, e.g. by running:
$ sudo -u www-data /opt/invenio/bin/inveniocfg --reset-sitename
* Obviously you could do such a dump/load for other collection-related
tables as well, such as `collection_portalbox' and the joined table
`portalbox', not only for the above two tables. See:
$ echo "SHOW TABLES" | /opt/invenio/bin/dbexec | grep -i colle
Depends on what exactly you are transferring: everything or only some
parts.
* See also `inveniocfg_dumperloader.py', but this is experimental. With
the recent move to the SQLAlchemy, the dumper/loader tool will become
more functional. We thought of using Python Fixture format for the
exported configuration file, so that it would also become more
comfortably editable. We can talk about the dumper/loader tool next
week, if that interests you.
Best regards
--
Tibor Simko
