Postgres Pro
Downloads
Home   >   mailing lists

ERROR: missing chunk number 0 for toast value - Mailing list pgsql-hackers

From Rushabh Lathia
Subject ERROR: missing chunk number 0 for toast value
Date
Msg-id CAGPqQf2bG_u8reY3ehL5pcWzw+4C6c7Kn0OsfHnOvHs3hDsihw@mail.gmail.com
Whole thread Raw
Responses Re: ERROR: missing chunk number 0 for toast value
List pgsql-hackers
Tree view
Hi All,

Test case:

drop table if exists t;
create table t(c text);
insert into t values ('x'), (repeat(md5('abcdefghijklmnop'), 10000));
select pg_column_size(c), pg_column_size(c || '') FROM t;

CREATE OR REPLACE FUNCTION copy_toast_out() RETURNS VOID AS $$ 
declare
        v text;
BEGIN  
        SELECT c INTO v FROM t WHERE c <> 'x';
    Select 1/0;
Exception
        When Others Then
        PERFORM pg_sleep(30); -- go run "TRUNCATE t" in a 2nd session                                                                                                          
        raise notice 'length :%', length(v || ''); -- force detoast                                                                                                          
END;
$$ language plpgsql;

postgres=# select copy_toast_out();
ERROR:  missing chunk number 0 for toast value 16390 in pg_toast_16384
CONTEXT:  PL/pgSQL function copy_toast_out() line 10 at RAISE

Analysis:

The basic problem here is that if the lock is released on table before
extracting toasted value, and in meantime someone truncates the table,
this error can occur.  Here error coming with PL block contains an Exception
block (as incase there is an exception block, it calls RollbackAndReleaseCurrentSubTransaction).

Do you think we should detoast the local variable before  RollbackAndReleaseCurrentSubTransaction ? Or any other options ?

Regards,
Rushabh Lathia

pgsql-hackers by date:

Previous
From: Magnus Hagander
Date:
Subject: Re: Fixing pg_basebackup with tablespaces found in $PGDATA
Next
From: Robert Haas
Date:
Subject: Re: preserving forensic information when we freeze