 | |
#91
 
| |||
| |||
|
Re: Nested Transactions -
10-02-2008
, 06:58 AM
|
You are right, I did not think of that - all statements in a BEGIN ... EXCEPTION block that were executed before the exception occurred will not be rolled back. In autocommit mode there is also an active transaction; all statements in a function run in that same transaction. So basically what you would like are SAVEPOINT and ROLLBACK TO SAVEPOINT statements in a function, right? Maybe you could ask the pgsql-hackers list if this would be a good feature to add. |
wrong: the savepoint in a BEGIN ... EXCEPTION block is actually set at
the BEGIN.
So if any statement in the block causes an exception that is caught,
*all* the statements within the block will be undone.
So PostgreSQL actually does implement subtransactions the way you want
them.
See the mailing list archives which also contain examples:
http://archives.postgresql.org/pgsql...0/msg00061.php
http://archives.postgresql.org/pgsql...0/msg00068.php
Yours,
Laurenz Albe
#92
| |||
| |||
|
|
You are right, I did not think of that - all statements in a BEGIN ... EXCEPTION block that were executed before the exception occurred will not be rolled back. In autocommit mode there is also an active transaction; all statements in a function run in that same transaction. So basically what you would like are SAVEPOINT and ROLLBACK TO SAVEPOINT statements in a function, right? Maybe you could ask the pgsql-hackers list if this would be a good feature to add. |
wrong: the savepoint in a BEGIN ... EXCEPTION block is actually set at
the BEGIN.
So if any statement in the block causes an exception that is caught,
*all* the statements within the block will be undone.
So PostgreSQL actually does implement subtransactions the way you want
them.
See the mailing list archives which also contain examples:
http://archives.postgresql.org/pgsql...0/msg00061.php
http://archives.postgresql.org/pgsql...0/msg00068.php
Yours,
Laurenz Albe
#93
| |||
| |||
|
|
You are right, I did not think of that - all statements in a BEGIN ... EXCEPTION block that were executed before the exception occurred will not be rolled back. In autocommit mode there is also an active transaction; all statements in a function run in that same transaction. So basically what you would like are SAVEPOINT and ROLLBACK TO SAVEPOINT statements in a function, right? Maybe you could ask the pgsql-hackers list if this would be a good feature to add. |
wrong: the savepoint in a BEGIN ... EXCEPTION block is actually set at
the BEGIN.
So if any statement in the block causes an exception that is caught,
*all* the statements within the block will be undone.
So PostgreSQL actually does implement subtransactions the way you want
them.
See the mailing list archives which also contain examples:
http://archives.postgresql.org/pgsql...0/msg00061.php
http://archives.postgresql.org/pgsql...0/msg00068.php
Yours,
Laurenz Albe
#94
| |||
| |||
|
|
You are right, I did not think of that - all statements in a BEGIN ... EXCEPTION block that were executed before the exception occurred will not be rolled back. In autocommit mode there is also an active transaction; all statements in a function run in that same transaction. So basically what you would like are SAVEPOINT and ROLLBACK TO SAVEPOINT statements in a function, right? Maybe you could ask the pgsql-hackers list if this would be a good feature to add. |
wrong: the savepoint in a BEGIN ... EXCEPTION block is actually set at
the BEGIN.
So if any statement in the block causes an exception that is caught,
*all* the statements within the block will be undone.
So PostgreSQL actually does implement subtransactions the way you want
them.
See the mailing list archives which also contain examples:
http://archives.postgresql.org/pgsql...0/msg00061.php
http://archives.postgresql.org/pgsql...0/msg00068.php
Yours,
Laurenz Albe
#95
| |||
| |||
|
|
You are right, I did not think of that - all statements in a BEGIN ... EXCEPTION block that were executed before the exception occurred will not be rolled back. In autocommit mode there is also an active transaction; all statements in a function run in that same transaction. So basically what you would like are SAVEPOINT and ROLLBACK TO SAVEPOINT statements in a function, right? Maybe you could ask the pgsql-hackers list if this would be a good feature to add. |
wrong: the savepoint in a BEGIN ... EXCEPTION block is actually set at
the BEGIN.
So if any statement in the block causes an exception that is caught,
*all* the statements within the block will be undone.
So PostgreSQL actually does implement subtransactions the way you want
them.
See the mailing list archives which also contain examples:
http://archives.postgresql.org/pgsql...0/msg00061.php
http://archives.postgresql.org/pgsql...0/msg00068.php
Yours,
Laurenz Albe
#96
| |||
| |||
|
|
You are right, I did not think of that - all statements in a BEGIN ... EXCEPTION block that were executed before the exception occurred will not be rolled back. In autocommit mode there is also an active transaction; all statements in a function run in that same transaction. So basically what you would like are SAVEPOINT and ROLLBACK TO SAVEPOINT statements in a function, right? Maybe you could ask the pgsql-hackers list if this would be a good feature to add. |
wrong: the savepoint in a BEGIN ... EXCEPTION block is actually set at
the BEGIN.
So if any statement in the block causes an exception that is caught,
*all* the statements within the block will be undone.
So PostgreSQL actually does implement subtransactions the way you want
them.
See the mailing list archives which also contain examples:
http://archives.postgresql.org/pgsql...0/msg00061.php
http://archives.postgresql.org/pgsql...0/msg00068.php
Yours,
Laurenz Albe
#97
| |||
| |||
|
|
You are right, I did not think of that - all statements in a BEGIN ... EXCEPTION block that were executed before the exception occurred will not be rolled back. In autocommit mode there is also an active transaction; all statements in a function run in that same transaction. So basically what you would like are SAVEPOINT and ROLLBACK TO SAVEPOINT statements in a function, right? Maybe you could ask the pgsql-hackers list if this would be a good feature to add. |
wrong: the savepoint in a BEGIN ... EXCEPTION block is actually set at
the BEGIN.
So if any statement in the block causes an exception that is caught,
*all* the statements within the block will be undone.
So PostgreSQL actually does implement subtransactions the way you want
them.
See the mailing list archives which also contain examples:
http://archives.postgresql.org/pgsql...0/msg00061.php
http://archives.postgresql.org/pgsql...0/msg00068.php
Yours,
Laurenz Albe
 |
« Previous Thread
|
Next Thread »
| Thread Tools | |
| Display Modes | |
Linear Mode
| |
Powered by vBulletin Version 3.5.3
Copyright ©2000 - 2012, Jelsoft Enterprises Ltd.
Copyright ©2000 - 2012, Jelsoft Enterprises Ltd.