This is a quick 'n' dirty Ruby / Rails script that helps to mirror data from one Java Content Repository (JCR) to another.

It is particularly helpful when the JCR XML import and exports fail, and no other options exist (e.g. underlying database is corrupt or bloated).

At this time, it supports most fields, including date and binary fields, but not JCR references.

It has been tested against Jackrabbit 2.10.1 and works with the WebDAV (HTTP) interface.

NOTE: not yet tested against Jackrabbit Oak

Run something like this:

WEBDAV_SRC=http://source-jcr.com:8080 WEBDAV_DST=http://target-jcr.com:8080 USERNAME=admin PASSWORD=pw ROOT=/jcrpath ruby test.rb

The Ruby script stores a running tally of synced data state in a Sqlite3 database. The script can be aborted and it will automatically start where it left off.

There is a status column in the nodes database table that determines whether the node has finished syncing. This can be set to :incomplete (value = 1) to restart the sync for the node.

There is also a last_synced_at database column but it is not currently used. It is an exercise for the reader to extend this script to re-start the sync to scan for node changes after a particular sync time (or alternatively, the start time of the script).

I hope you find this useful.