Eclipse Community Forums: EclipseLink » Eclipse Data Partition Question

Help

Home

Home » Eclipse Projects » EclipseLink » Eclipse Data Partition Question

Show: Today's Messages :: Show Polls :: Message Navigator

Eclipse Data Partition Question [message #1061781]

Tue, 04 June 2013 03:21

Oliver Zhou

Messages: 41
Registered: February 2012

Member

Hi,

I download the eclipselink data partition example and I can run it successfully. But I have a few questions about it.

1. In the Customer class, I see two partition policy annotation applied:

@ReplicationPartitioning(name="Replicate", connectionPools={"default","node2"})
@HashPartitioning(name="HashPartitionById", 
	partitionColumn=@Column(name="ID"), connectionPools={"default","node2"})

I think these two policies looks like incompatible, and in practise, only HashPartitoning policy takes effect.

2. Suppose I would like to issue a query like this:

SELECT c FROM Customer c WHERE c.id < 10 ORDER BY c.id

The result are not globally sorted and only partial sorted by each partition.
Is there a way to use a merge sort algorithm to sort the merged result?

Best wishes,
Oliver

[Updated on: Tue, 04 June 2013 03:22]

Report message to a moderator

Re: Eclipse Data Partition Question [message #1061886 is a reply to message #1061781]

Tue, 04 June 2013 14:10

James Sutherland

Messages: 1939
Registered: July 2009
Location: Ottawa, Canada

Senior Member

1 - You are correct, the ReplicationPartitioning is not used by Customer. It is defined there so it can be used as the session default policy in the persistence.xml, this allows for DDL to be replicated for schema generation.

2 - Two sort across partitions you need to sort in Java. You can just use the Java Collections sort() function.

James : Wiki : Book : Blog : Twitter

Report message to a moderator