[#BATCH-539] Oracle schema could use more oracle-appropriated datatypes.

Spring Batch

Oracle schema could use more oracle-appropriated datatypes.

Created: 01/Apr/08 02:05 PM Updated: 07/Aug/08 10:07 AM

Component/s:

Core

Affects Version/s:

1.0.0

Fix Version/s:

1.0.1

Time Tracking:

Original Estimate:

0.06d

Remaining Estimate:

0.06d

Time Spent:

Not Specified

File Attachments:	1. oracle.diff (5 kB)
Environment:	oracle

Description

« Hide

While working with an Oracle DBA, it was noticed that the Oracle schema could use more oracle-appropriated datatypes. They are functionally equivalent, but are more suited to oracle.

VARCHAR->VARCHAR2 : This is recommended by oracle for all variable character data. [1]
DOUBLE PRECISION -> NUMBER : These are synonyms, but more NUMBER is clearer as to what datatype is actually being used.

[1] http://download.oracle.com/docs/cd/B19306_01/server.102/b14200/sql_elements001.htm#sthref77
[2] http://download.oracle.com/docs/cd/B19306_01/server.102/b14200/sql_elements001.htm#i54335

All

Comments

Work Log

Change History

FishEye

Builds

Sort Order:

[ Permlink | « Hide ]

Gregory Kick added a comment - 01/Apr/08 03:00 PM

A patch that includes the changes outlined in the issue.

[ Permlink | « Hide ]

Richard Kettelerij added a comment - 03/Apr/08 09:03 AM

Shouldn't there be more indexes present in the Oracle schema? Currently no indexes are created except for the ones associated with primary keys (these are auto created).

[ Permlink | « Hide ]

Gregory Kick added a comment - 03/Apr/08 09:40 AM

@Richard
AFAIK, the framework only queries based on the primary keys. Any additional indexes would depend on your particular usecase. (e.g. monitoring, reporting, etc.)

[ Permlink | « Hide ]

Dave Syer added a comment - 03/Apr/08 11:53 AM

Actually we do query on the JOB_KEY column in BATCH_JOB_EXECUTION. But if we provide indexes, someone will only want more, or to put them in a different tablespace. I would prefer that people do their own indexing. Maybe we could explicitly recommend which columns should be indexed for efficiency.

At the end of the day, even a full table scan once per job execution on our meta data tables is unlikely to be a bottleneck if the jobs are doing any significant processing. So it might not be that important.

[ Permlink | « Hide ]

Richard Kettelerij added a comment - 03/Apr/08 12:54 PM

Ok, a note in the reference documentation describing candidate indexes would suffice (for most user I think).

[ Permlink | « Hide ]

Dave Syer added a comment - 04/Apr/08 06:47 AM

Added in the section on JobRepository (Chapter 3).

[ Permlink | « Hide ]

Gregory Kick added a comment - 04/Apr/08 10:24 AM

while i appreciate extra documentation as much as the next person, the original issue had nothing to do with indexing. (see the original title, description and patch.) it seems that that should have been a separate issue linked to this one. please reopen this issue.