Christian Fowler wrote:
db=# \encoding
UNICODE
the last four columns in the table are:
full_name_nd | character varying(200) |
mod_date | timestamp without time zone |
pc_char | character varying(4)|
dim_char | character varying(16) |
[EMAIL PROTECTED] root]#
Joshua -
well, i ran a --inserts dump as you suggested, and it died upteen millions
(and many hours) in at:
INSERT 59235383 1
invalid command \033',
cannot allocate memory for output buffer
Weiping,
db=# \encoding
UNICODE
the last four columns in the table are:
full_name_nd | character varying
Christian Fowler wrote:
I'm running a large database ( data dir is ~9gigs ) on Fedora Core 2
with 7.4.5-PGDG rpm's
When using the -Fc dump method + pg_restore, I get:
-bash-2.05b$ pg_restore -Fc -d foo -L backup.list
/tmp/02\:43-postgresql_database-foo-backup
pg_restore: ERROR: invalid input sy
When using pg_dump + psql:
I get a similar error. In one table that has about 5.4 million rows,
the dump has several incomplete rows in the large block of copy data.
It seems to attempt to fill the copy with data from the *next* line.
ERROR: invalid input syntax for type timestamp: "4"
CONTEXT
I'm running a large database ( data dir is ~9gigs ) on Fedora Core 2 with
7.4.5-PGDG rpm's
When using the -Fc dump method + pg_restore, I get:
-bash-2.05b$ pg_restore -Fc -d foo -L backup.list
/tmp/02\:43-postgresql_database-foo-backup
pg_restore: ERROR: invalid input syntax for type timestamp:
