I am trying to load a data base using pg_restore. We are migrating to
a new server and I want to see if I can make the load go faster if I
load with pg_restore instead of psql.
The old server is:
CentOS upgraded to 5.4 64bit with the stock PostgreSQL 8.1.
The dump file was created with:
pg_dump --format=c --file=ec.restore ec
where ec is the name of the database. It is encoded SQL_ASCII. I
plan on converting to UTF-8 but I can't do that quite yet.
The new server is
CentOS 5.4 64 bit with PostgreSQL 8.4 installed from the YUM repo on
the PostgreSQL site.
The restore command is:
createdb -E SQL_ASCII -T template0 ec
pg_restore -dec -j2 ec.restore
I get this error:
pg_restore: [custom archiver] dumping a specific TOC data block out of
order is not supported without ID on this input stream (fseek
required)
pg_restore: [archiver] worker process failed: exit code 1
It seems to run if I ommit the -j2 option. (Though as of this moment
is has not completed). Is there a known problem with using parallel
loading in 8.4 from a file created with an 8.1 database?
--
Bryan White
