Issues fixed in MarkLogic Server 9.0-8.2

DUPURI for property fragments

With some settings, DUPURI result in directory property fragments when automatic directory creation is on and directories are created concurrently

SQL queries on TDE views return duplicate rows for non-admin users

When a view is defined in several TDE templates that are inserted with different permissions, SQL queries might return duplicated rows of data unless run with the admin role.

S3 incremental restore is very slow

Restoring an incremental backup is excessively I/O intensive. This is most noticeable restoring from S3.

XCC loaded document with AdhocQuery & variable bindings not findable using cts:element-value-query

Under some circumstances, using XCC and an AdhocQuery to load a document into MarkLogic Server where the document is bound as an external variable (using ValueFactory.newElement). A subsequent cts:element-value-query will not find the document unless unstemmed searching is specifed. This has been seen with a MarkLogic 8 instance that has been upgraded to MarkLogic 9.

Azure VM fails to start with missing custom data file

MarkLogic Server will not start on Azure if there is no /var/lib/waagent/CustomData file for the VM.

[9.0-8 Regression] SPARQL Insert and Delete query produced two bnodes for a predicate.

Starting in MarkLogic Server 9.0-8, SPARQL Insert and Delete query produced two bnodes for a predicate.

XCC exception with chunk transfer encoded.

XCC throws exception when the response is chunk transfer encoded. This has been seen when the response is large and sent through an AWS ELB.

JavaScript Performance Improvement

Module Cache efficiencies for Javascript modules