pg7.3b1

Started by Laurette Cisnerosover 23 years ago7 messages

laurette@nextbus.com

over 23 years ago

I am so glad that postgres now keeps track of relationships between rule,
views, functions, tables, etc. I've had to re-do all my creation and drop
scripts but this is definitely for the better.

During my testing of my scripts, I have come across this message:
psql:/u1/cvs73/DataBase/Config/Schema/logconfig.sql:142: WARNING: Relcache reference leak: relation "positions" has refcnt 1 instead of 0

What does this indicate?

Thanks,

--
Laurette Cisneros
The Database Group
(510) 420-3137
NextBus Information Systems, Inc.
www.nextbus.com
-----------------------------------
NextBus say:
Riders prefer to arrive just minute
before bus than just minute after.

Rod Taylor

rbt@rbt.ca

over 23 years ago

In reply to: Laurette Cisneros (#1)

Re: pg7.3b1

On Thu, 2002-09-26 at 16:46, Laurette Cisneros wrote:

I am so glad that postgres now keeps track of relationships between rule,
views, functions, tables, etc. I've had to re-do all my creation and drop
scripts but this is definitely for the better.

During my testing of my scripts, I have come across this message:
psql:/u1/cvs73/DataBase/Config/Schema/logconfig.sql:142: WARNING: Relcache reference leak: relation "positions" has refcnt 1 instead of 0

What does this indicate?

Someone (probably me) made a mistake and forgot to release a cache
handle.

Do you happen to have a sequence of commands that can reproduce this?

--
Rod Taylor

Laurette Cisneros

laurette@nextbus.com

over 23 years ago

In reply to: Rod Taylor (#2)

Re: pg7.3b1

I'll see if I can pare down my scripts (they are long) to reproduce this
easier.

L.
On 26 Sep 2002, Rod Taylor wrote:

On Thu, 2002-09-26 at 16:46, Laurette Cisneros wrote:

I am so glad that postgres now keeps track of relationships between rule,
views, functions, tables, etc. I've had to re-do all my creation and drop
scripts but this is definitely for the better.

During my testing of my scripts, I have come across this message:
psql:/u1/cvs73/DataBase/Config/Schema/logconfig.sql:142: WARNING: Relcache reference leak: relation "positions" has refcnt 1 instead of 0

What does this indicate?

Someone (probably me) made a mistake and forgot to release a cache
handle.

Do you happen to have a sequence of commands that can reproduce this?

Laurette Cisneros

laurette@nextbus.com

over 23 years ago

In reply to: Rod Taylor (#2)

Re: pg7.3b1

Ok, finally had time to narrow this down.

Here's the simplified script that will reproduce this (this sequence
reroduces on my system using 7.3b2):

\echo BEGIN tst.sql

create table pp
( x integer
, i text
);

create view p as
select * from pp where i is null;

comment on view p is
'This is a comment.';

create rule p_ins as on insert to p do instead
insert into pp
values ( new.x
, null
);

comment on rule p_ins is 'insert to p goes to pp';

\echo END tst.sql

On 26 Sep 2002, Rod Taylor wrote:

On Thu, 2002-09-26 at 16:46, Laurette Cisneros wrote:

I am so glad that postgres now keeps track of relationships between rule,
views, functions, tables, etc. I've had to re-do all my creation and drop
scripts but this is definitely for the better.

During my testing of my scripts, I have come across this message:
psql:/u1/cvs73/DataBase/Config/Schema/logconfig.sql:142: WARNING: Relcache reference leak: relation "positions" has refcnt 1 instead of 0

What does this indicate?

Someone (probably me) made a mistake and forgot to release a cache
handle.

Do you happen to have a sequence of commands that can reproduce this?

--
Laurette Cisneros
The Database Group
(510) 420-3137
NextBus Information Systems, Inc.
www.nextbus.com
------------------------------
Do you know where your bus is?

Rod Taylor

rbt@rbt.ca

over 23 years ago

In reply to: Laurette Cisneros (#4)

1 attachment(s)

Re: [HACKERS] pg7.3b1

Same rule patch, right email this time :)

Lock on the rule relation wasn't removed after adding the comment.

On Wed, 2002-10-02 at 19:03, Laurette Cisneros wrote:

Ok, finally had time to narrow this down.

Here's the simplified script that will reproduce this (this sequence
reroduces on my system using 7.3b2):

\echo BEGIN tst.sql

create table pp
( x integer
, i text
);

create view p as
select * from pp where i is null;

comment on view p is
'This is a comment.';

create rule p_ins as on insert to p do instead
insert into pp
values ( new.x
, null
);

comment on rule p_ins is 'insert to p goes to pp';

\echo END tst.sql

On 26 Sep 2002, Rod Taylor wrote:

On Thu, 2002-09-26 at 16:46, Laurette Cisneros wrote:

I am so glad that postgres now keeps track of relationships between rule,
views, functions, tables, etc. I've had to re-do all my creation and drop
scripts but this is definitely for the better.

During my testing of my scripts, I have come across this message:
psql:/u1/cvs73/DataBase/Config/Schema/logconfig.sql:142: WARNING: Relcache reference leak: relation "positions" has refcnt 1 instead of 0

What does this indicate?

Someone (probably me) made a mistake and forgot to release a cache
handle.

Do you happen to have a sequence of commands that can reproduce this?

--
Laurette Cisneros
The Database Group
(510) 420-3137
NextBus Information Systems, Inc.
www.nextbus.com
------------------------------
Do you know where your bus is?

--
Rod Taylor

Attachments:

rule.patchtext/x-patch; charset=ISO-8859-1; name=rule.patchDownload

Index: src/backend/commands/comment.c
===================================================================
RCS file: /projects/cvsroot/pgsql-server/src/backend/commands/comment.c,v
retrieving revision 1.60
diff -c -r1.60 comment.c
*** src/backend/commands/comment.c	2002/09/04 20:31:14	1.60
--- src/backend/commands/comment.c	2002/10/03 00:12:23
***************
*** 559,565 ****
  	}
  
  	/* Check object security */
- 
  	aclcheck = pg_class_aclcheck(reloid, GetUserId(), ACL_RULE);
  	if (aclcheck != ACLCHECK_OK)
  		aclcheck_error(aclcheck, rulename);
--- 559,564 ----
***************
*** 568,575 ****
  	classoid = get_system_catalog_relid(RewriteRelationName);
  
  	/* Call CreateComments() to create/drop the comments */
- 
  	CreateComments(ruleoid, classoid, 0, comment);
  }
  
  /*
--- 567,575 ----
  	classoid = get_system_catalog_relid(RewriteRelationName);
  
  	/* Call CreateComments() to create/drop the comments */
  	CreateComments(ruleoid, classoid, 0, comment);
+ 
+ 	heap_close(relation, AccessShareLock);
  }
  
  /*

Bruce Momjian

pgman@candle.pha.pa.us

over 23 years ago

In reply to: Rod Taylor (#5)

Re: [HACKERS] pg7.3b1

Your patch has been added to the PostgreSQL unapplied patches list at:

http://candle.pha.pa.us/cgi-bin/pgpatches

I will try to apply it within the next 48 hours.

---------------------------------------------------------------------------

Rod Taylor wrote:

Same rule patch, right email this time :)

Lock on the rule relation wasn't removed after adding the comment.

On Wed, 2002-10-02 at 19:03, Laurette Cisneros wrote:

Ok, finally had time to narrow this down.

Here's the simplified script that will reproduce this (this sequence
reroduces on my system using 7.3b2):

\echo BEGIN tst.sql

create table pp
( x integer
, i text
);

create view p as
select * from pp where i is null;

comment on view p is
'This is a comment.';

create rule p_ins as on insert to p do instead
insert into pp
values ( new.x
, null
);

comment on rule p_ins is 'insert to p goes to pp';

\echo END tst.sql

On 26 Sep 2002, Rod Taylor wrote:

On Thu, 2002-09-26 at 16:46, Laurette Cisneros wrote:

I am so glad that postgres now keeps track of relationships between rule,
views, functions, tables, etc. I've had to re-do all my creation and drop
scripts but this is definitely for the better.

During my testing of my scripts, I have come across this message:
psql:/u1/cvs73/DataBase/Config/Schema/logconfig.sql:142: WARNING: Relcache reference leak: relation "positions" has refcnt 1 instead of 0

What does this indicate?

Someone (probably me) made a mistake and forgot to release a cache
handle.

Do you happen to have a sequence of commands that can reproduce this?

--
Laurette Cisneros
The Database Group
(510) 420-3137
NextBus Information Systems, Inc.
www.nextbus.com
------------------------------
Do you know where your bus is?

--
Rod Taylor

[ Attachment, skipping... ]

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

Bruce Momjian

pgman@candle.pha.pa.us

over 23 years ago

In reply to: Rod Taylor (#5)

Re: [HACKERS] pg7.3b1

Patch applied with Tom's fix for heap_close(NoLock). Thanks.

---------------------------------------------------------------------------

Rod Taylor wrote:

Same rule patch, right email this time :)

Lock on the rule relation wasn't removed after adding the comment.

On Wed, 2002-10-02 at 19:03, Laurette Cisneros wrote:

Ok, finally had time to narrow this down.

Here's the simplified script that will reproduce this (this sequence
reroduces on my system using 7.3b2):

\echo BEGIN tst.sql

create table pp
( x integer
, i text
);

create view p as
select * from pp where i is null;

comment on view p is
'This is a comment.';

create rule p_ins as on insert to p do instead
insert into pp
values ( new.x
, null
);

comment on rule p_ins is 'insert to p goes to pp';

\echo END tst.sql

On 26 Sep 2002, Rod Taylor wrote:

On Thu, 2002-09-26 at 16:46, Laurette Cisneros wrote:

I am so glad that postgres now keeps track of relationships between rule,
views, functions, tables, etc. I've had to re-do all my creation and drop
scripts but this is definitely for the better.

During my testing of my scripts, I have come across this message:
psql:/u1/cvs73/DataBase/Config/Schema/logconfig.sql:142: WARNING: Relcache reference leak: relation "positions" has refcnt 1 instead of 0

What does this indicate?

Someone (probably me) made a mistake and forgot to release a cache
handle.

Do you happen to have a sequence of commands that can reproduce this?

--
Laurette Cisneros
The Database Group
(510) 420-3137
NextBus Information Systems, Inc.
www.nextbus.com
------------------------------
Do you know where your bus is?

--
Rod Taylor

[ Attachment, skipping... ]

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073