Thanks for the new thread ( I will do not more hickjack in future, I'll promise
:-)
Dan Langille wrote:
> Dan Langille wrote:
>> This was originally posted under PostgreSQL createdb changes
>>
>> Bruno Friedmann wrote:
>>
>>> Hi Dan
>>>
>>> Can you check for constency with the other databases (Mysql/Sqlite) how are
>>> spell the table,columns and their order in declaration.
>>> I know we don't care about sensitive case under postgresql, but it would
>>> make easier to migrate from one db to another
>>> (dump stupid sql) or csv file ( but actually you need to know the order to
>>> re-import them)
>>>
>>> I've done many transformation for a job like this one month ago. (was on my
>>> todo list ... )
>>> Perharps we need to have some sort of reflexion, as I've not checked if
>>> bacula save catalog job use full insert or not.
>> I think you are asking me to make the PostgreSQL names case-sensitive.
>> I would rather not do that. Here is why:
>>
>> test=# create table "Pool" (col1 text);
>> CREATE TABLE
>> test=# select * from pool;
>> ERROR: relation "pool" does not exist
>> test=# select * from Pool;
>> ERROR: relation "pool" does not exist
>> test=# select * from "Pool";
>> col1
>> ------
>> (0 rows)
>>
>> test=#
>>
>> That is, you MUST use double quotes.
>>
>> As for the ordering of the columns, I agree, the column orders should be
>> the same.
>>
>> Do you have notes from the transformations you did?
>
> In my short look, I found only one instance of columns out of order.
>
> The PostgreSQL Job table has priorjobid defined at the end of the CREATE
> TABLE. It should be right after filesetid.
>
> Recommendation: move it.
It's also in my notes.
>
> Prognosis: This would only affect users if they were doing something
> very specifi. Otherwise, it should not.
Just to be sure, there's one case, if people have made a backup catalog are
they able to reinject it's content
with the new create_table in case of rescue or some disaster.
(I've not checked carefully what use the make_catalog_backup for postgresql,
full insert or copy with or without cols name)
One thing that can be used (just need to verify encoding trouble between
host/db/engine) is to make a little script exporting
table data into csv, and try to reimport them inside other engine.
Mysql export -> import postgresql / Sqlite
Pg export -> import Sqlite / mysql
Sqlite export -> import pg / mysql
Perharps this could become part of regression tests or something like this.
good idea ?
ps : I can try to build something like this, but I'm a bit short of time. So I
can only do this during Christmas holiday.
>
> In the JobMedia table, MySQL has a Stripe column which is not present in
> either PostgreSQL or SQLite.
>
> Recommdation: delete it from the table definition. I do not see it
> referenced in the code. No need to drop it. We could issue a note
> saying users could drop if they wish.
>
--
Bruno Friedmann
Ioda-Net Sàrl
2830 Vellerat - Switzerland
Tél : ++41 32 435 7171
Fax : ++41 32 435 7172
gsm : ++41 78 802 6760
www.ioda-net.ch
Centre de Formation et de Coaching En Ligne
www.cfcel.com
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Bacula-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-devel
