Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [eclipselink-users] Hibernate ID Migration

Both sequence objects and table sequencing are generated on the database. 
Sequence objects (supported in Oracle and a few other databases) allow
preallocation so have very good performance and do not have any
synchronization issues, as they are allocated outside of the transaction

Table sequencing also supports a preallocation size, and TopLink can
allocate sequence numbers outside the transaction context (sequence
connection pool) so does not have any synchronization issues (especially if
a large preallocation size is used).

Not using a sequence preallocation size, or using identity sequencing will
lead to poor performance, and can lead to synchronization issues with table
sequencing.  But also long as you use a large sequence preallocation size
there are no issues.

Frans Thamura-2 wrote:
>> EclipseLink (and JPA) do not have a UUID generator, you could implement
>> your
>> own through subclassing the Sequence class
>> (org.eclipse.persistence.sequencing) (and perhaps contribute it to
>> EclipseLink), but I would recommend you use one of the standard id
>> generators.
> James,
> i dont get it.
> we use the UUID because we have bad experience with PrimaryKey
> Implementation esp for sync mechanism.
> is the table sequence directlyfrom database or like hibernate's UUID
> F

--- James Sutherland
 EclipseLink ,
Wiki: EclipseLink , TopLink 
Forums: TopLink , EclipseLink 
Book: Java Persistence 
View this message in context:
Sent from the EclipseLink - Users mailing list archive at

Back to the top