Thread: Transaction with in function
I know that postgresql does not support nested transactions,
I looking for solution for commit, rollback with in function. May be using the save point.
I were to use save point in function , when it would get commit.?
Thanks in advance .
Regards
Ravi Katkar
Ravi Katkar <Ravi.Katkar@infor.com> wrote: > I know that postgresql does not support nested transactions, > > I looking for solution for commit, rollback with in function. May > be using the save point. That is essentially the same as nested transactions, isn't it? Just using SQL standard syntax? > I were to use save point in function , when it would get commit.? Each savepoint creates a subtransaction which will be included in the top level transaction unless it or an intermediate savepoint is rolled back. Nothing from the top level transaction is visible to other transactions until the commit of the top level transaction. If this and the docs don't make it clear, please post again: http://www.postgresql.org/docs/8.4/interactive/sql-savepoint.html If you do post again, please include your PostgreSQL version and an explanation of what problem you're trying to solve. More specific questions tend to elicit more useful responses: -Kevin
Hi ,
What is the Oracle’s Long data type equivalent data type in Postgresql ?
Regards,
Ravi katkar
Στις Friday 28 May 2010 11:45:17 ο/η Ravi Katkar έγραψε: > Hi , > > What is the Oracle's Long data type equivalent data type in Postgresql ? > int8 i guess > Regards, > Ravi katkar > > > > -- Achilleas Mantzios
> -----Original Message----- > From: Achilleas Mantzios [mailto:achill@matrix.gatewaynet.com] > Sent: Friday, May 28, 2010 5:03 AM > To: pgsql-admin@postgresql.org > Subject: Re: Oracle's Long data type equivalent data type in > Postgresql ? > > Στις Friday 28 May 2010 11:45:17 ο/η Ravi Katkar έγραψε: > > Hi , > > > > What is the Oracle's Long data type equivalent data type > in Postgresql ? > > > > int8 i guess > > > Regards, > > Ravi katkar > > > > > > > > > > > > -- > Achilleas Mantzios > No, it's bytea (and not int8). Igor Neyman