Midori is happy to announce that the Archiving Plugin for Confluence 6.1.0 is generally available at Atlassian Marketplace!
New in this release
Performance, performance and more performance
In the previous add-on version (6.0.0), we introduced the so-called data limits to save your server from wasting computing resources (CPU, memory) on irrelevant data. In this version, we rebuilt the end of the data processing pipeline for even more performance and lower memory footprint. This enabled to significantly increase the defaults of the data limits, which are configurable now!
|Data limit parameter||Default value|
Using these data limits, several thousand notification emails (for hundreds of spaces and several hundred thousand pages) were sent without problems in our internal load tests.
- Fixed: Hibernate "owning Session was closed" error while collecting the addressees for notification emails.
- Fixed: Empty page lists in archiving events cause NPEs with Oracle.
- Fixed: Blacklist checking crashes on the spaces that have no description object associated with (i.e. their data is corrupt).
- Fixed: When all spaces are blacklisted, the Content Quality Stats screen misleadingly suggests that the system is in an unindexed state.
Breadcrumb style page lists removed
Breadcrumbs-style page lists in the notification emails are not supported anymore. You can't displays the ancestors of an expired page, like "grandparent > parent > actual page", only display the actual page. We did for two reasons. First: that's not the Confluence standard. Second: carrying all ancestors with the pages until the mail rendering logic generated extra complication (ancestors are lazy loaded Hibernate objects) and extra use of memory (ancestors are heavy weight), so we're keeping things simple, robust and scalable.