Обсуждение: Script terminates even though ON_ERROR_STOP not set
I have an admin script that drops a schema and then recreates it (quick way to refresh from backup). However, if the schema doesn't exist, the drop errors, as you would expect. The next SQL command, the schema create, is never executed. I do not have ON_ERROR_STOP set so why is this happening? I don't want to split down to two calls of psql if possible, which is the present work-around. Would prefer to solve the problem. Thanks, Brad
On Fri, Jan 14, 2005 at 12:03:02PM +0000, Bradley Kieser wrote: > I have an admin script that drops a schema and then recreates it (quick > way to refresh from backup). However, if the schema doesn't exist, the > drop errors, as you would expect. The next SQL command, the schema > create, is never executed. How do you know the create is never executed? Are you looking for the schema and not finding it? How are you looking? Or is there an error message? What is it? Are you doing the drop and create in the same transaction? If so, then the drop statement's failure is preventing subsequent statements in that transaction from executing. If that's not the problem then please post a self-contained test case so we can see exactly what you're doing. > I do not have ON_ERROR_STOP set so why is this happening? How do you know ON_ERROR_STOP isn't set? Could it be set in a .psqlrc file? What does \set show? -- Michael Fuhr http://www.fuhr.org/~mfuhr/
Michael Fuhr wrote: >On Fri, Jan 14, 2005 at 12:03:02PM +0000, Bradley Kieser wrote: > > > >>I have an admin script that drops a schema and then recreates it (quick >>way to refresh from backup). However, if the schema doesn't exist, the >>drop errors, as you would expect. The next SQL command, the schema >>create, is never executed. >> >> > >How do you know the create is never executed? Are you looking for >the schema and not finding it? How are you looking? Or is there >an error message? What is it? > > Looking for the schema and not finding it. >Are you doing the drop and create in the same transaction? If so, > > I have auto-commit on. >then the drop statement's failure is preventing subsequent statements >in that transaction from executing. If that's not the problem then >please post a self-contained test case so we can see exactly what >you're doing. > > Hmm, maybe I need to explicitly call commit? Will try and get back to you. > > >>I do not have ON_ERROR_STOP set so why is this happening? >> >> > >How do you know ON_ERROR_STOP isn't set? Could it be set in >a .psqlrc file? What does \set show? > > >
Okay, Michael gets that trophy! Thanks Michael, it needed a commit. However, this is odd. I get a "No transaction in progress" error coming out on the commit, but at least it works, so I am happy! Thank you! This works: psql -U$dba $db << END drop schema $repschema cascade; commit; create schema $repschema authorization $repuser; END Tis doesn't: psql -U$dba $db << END drop schema $repschema cascade; create schema $repschema authorization $repuser; END Brad Michael Fuhr wrote: >On Fri, Jan 14, 2005 at 12:03:02PM +0000, Bradley Kieser wrote: > > > >>I have an admin script that drops a schema and then recreates it (quick >>way to refresh from backup). However, if the schema doesn't exist, the >>drop errors, as you would expect. The next SQL command, the schema >>create, is never executed. >> >> > >How do you know the create is never executed? Are you looking for >the schema and not finding it? How are you looking? Or is there >an error message? What is it? > >Are you doing the drop and create in the same transaction? If so, >then the drop statement's failure is preventing subsequent statements >in that transaction from executing. If that's not the problem then >please post a self-contained test case so we can see exactly what >you're doing. > > > >>I do not have ON_ERROR_STOP set so why is this happening? >> >> > >How do you know ON_ERROR_STOP isn't set? Could it be set in >a .psqlrc file? What does \set show? > > >
Michael, I didn't see the last comment in my previous replies. I know that the variable isn't set because I embedded \set into the psql commands so that they would output what is set. Thanks, Brad Michael Fuhr wrote: >On Fri, Jan 14, 2005 at 12:03:02PM +0000, Bradley Kieser wrote: > > > >>I have an admin script that drops a schema and then recreates it (quick >>way to refresh from backup). However, if the schema doesn't exist, the >>drop errors, as you would expect. The next SQL command, the schema >>create, is never executed. >> >> > >How do you know the create is never executed? Are you looking for >the schema and not finding it? How are you looking? Or is there >an error message? What is it? > >Are you doing the drop and create in the same transaction? If so, >then the drop statement's failure is preventing subsequent statements >in that transaction from executing. If that's not the problem then >please post a self-contained test case so we can see exactly what >you're doing. > > > >>I do not have ON_ERROR_STOP set so why is this happening? >> >> > >How do you know ON_ERROR_STOP isn't set? Could it be set in >a .psqlrc file? What does \set show? > > >
On Fri, Jan 14, 2005 at 06:11:27PM +0000, Bradley Kieser wrote: > Okay, Michael gets that trophy! Thanks Michael, it needed a commit. > However, this is odd. I get a "No transaction in progress" error coming > out on the commit, but at least it works, so I am happy! What version of the client and server are you using? For the client run "psql --version"; for the server execute "SELECT version();". Where do you have autocommit set? PostgreSQL 7.4 removed autocommit from the server and added it to the client; I wonder if you're using differing client and server versions that are causing confusion. -- Michael Fuhr http://www.fuhr.org/~mfuhr/
$ psql --version
psql (PostgreSQL) 7.4.1
contains support for command-line editing
# SELECT version();
version
---------------------------------------------------------------------------------------------------------
PostgreSQL 7.4.1 on i686-pc-linux-gnu, compiled by GCC gcc (GCC) 3.3.2
(Mandrake Linux 10.0 3.3.2-6mdk)
(1 row)
Michael Fuhr wrote:
>On Fri, Jan 14, 2005 at 06:11:27PM +0000, Bradley Kieser wrote:
>
>
>
>>Okay, Michael gets that trophy! Thanks Michael, it needed a commit.
>>However, this is odd. I get a "No transaction in progress" error coming
>>out on the commit, but at least it works, so I am happy!
>>
>>
>
>What version of the client and server are you using? For the client
>run "psql --version"; for the server execute "SELECT version();".
>Where do you have autocommit set? PostgreSQL 7.4 removed autocommit
>from the server and added it to the client; I wonder if you're using
>differing client and server versions that are causing confusion.
>
>
>