Обсуждение: Invalid page feader on Solaris 10

Hi!

We have a Sun Fire V480 server 4GB of memory with Solaris 10 operation system and PostgreSQL 8.1.3 (from Sun).
The dump of my database is about 2 GB if we are lucky.
If not, we have one of error messages:

 - pg_dump: ERROR:  invalid memory alloc request size 2623204397

 - pg_dump: ERROR:  could not access status of transaction 2423378468
   DETAIL:  could not open file "pg_clog/0907": No such file or directory

 - pg_dump: ERROR:  invalid page header in block 16147 of relation "jn_szll"

What is the problem?

Some line of postgresql.conf: (the others are default)

shared_buffers = 50000 #1000                    # min 16 or max_connections*2, 8KB each
#temp_buffers = 1000                    # min 100, 8KB each
#max_prepared_transactions = 5          # can be 0 or more
# note: increasing max_prepared_transactions costs ~600 bytes of shared memory
# per transaction slot, plus lock space (see max_locks_per_transaction).
work_mem = 81920 #1024                  # min 64, size in KB
maintenance_work_mem = 16384            # min 1024, size in KB
#max_stack_depth = 2048                 # min 100, size in KB

# - Free Space Map -

max_fsm_pages = 80000 #20000                    # min max_fsm_relations*16, 6 bytes each
max_fsm_relations = 4000 #1000          # min 100, ~70 bytes each

autovacuum = on                         # enable autovacuum subprocess?
autovacuum_naptime = 600                # time between autovacuum runs, in secs



--
Üdvözlettel:
Rózsahegyi László

=?ISO-8859-2?Q?R=F3zsahegyi_L=E1szl=F3?= <laszlo.rozsahegyi@rool.hu> writes:
> The dump of my database is about 2 GB if we are lucky.
> If not, we have one of error messages:

>  - pg_dump: ERROR:  invalid memory alloc request size 2623204397

>  - pg_dump: ERROR:  could not access status of transaction 2423378468
>    DETAIL:  could not open file "pg_clog/0907": No such file or directory

>  - pg_dump: ERROR:  invalid page header in block 16147 of relation "jn_szll"

> What is the problem?

All of those look like corrupt-data symptoms to me.  I wonder whether
you've got hardware problems (flaky disk, flaky memory, bad cable ...)
How repeatable are the errors?  Do they always occur for just one table,
or is the problem widespread?

            regards, tom lane