MarkLogic 5.0 - First Look
November 01, 2011 at 12:24 PM | categories: XQuery, MarkLogic | View Comments
In case you have missed the news, MarkLogic Server 5.0-1 is now available. The upgrade went smoothly for me, but this is a major release so it is wise to back up your databases and configuration before upgrading. The on-disk forest version appears to have changed, which will trigger reindexing of all forests. You may want to manually disable reindexing before upgrading, so that you don't have to contend with multiple forests trying to reindex at the same time.
This is also a good time to double-check your free disk space, since reindexing uses extra disk space. Some of that space won't be released when reindexing finishes, either. For example, one of my forests looked like this:
You can purge those deleted fragments by forcing a merge of the forest, or of the entire database. After doing this, my forest used less disk space.
This new release is stricter about unquoted attributes. With previous releases this would generally work, even though the XQuery 1.0 Recommendation requires quoted attribute values:
Now it throws an
I'm looking forward to learning more about the 5.0 release, but so far it looks good.
This is also a good time to double-check your free disk space, since reindexing uses extra disk space. Some of that space won't be released when reindexing finishes, either. For example, one of my forests looked like this:
You can purge those deleted fragments by forcing a merge of the forest, or of the entire database. After doing this, my forest used less disk space.
This new release is stricter about unquoted attributes. With previous releases this would generally work, even though the XQuery 1.0 Recommendation requires quoted attribute values:
<test a={xdmp:random()}/>
Now it throws an
XDMP-UNEXPECTED error. Quote the attribute value correctly, and the problem is fixed.
<test a="{xdmp:random()}"/>
I'm looking forward to learning more about the 5.0 release, but so far it looks good.