ineffiency of pg_restore

pg_restore give admin a lot of flexabilities, it can restore any object
from a single dump file. But if the table is very big, e.g. 2M records,
it will take a long time to restore because it use INSERTs.
I believe that if we can have following feature:
1. quick restore(use COPY not INSERTs),
2. restore related objects, e.g. when we restore a table, we can choose
   to restore its indexes, triggers and related sequence etc.

that will be great, I expect that we have them in next version of release.


test log:

su-2.05$ time psql urldb<jurloldmail.out
You are now connected as new user jliang.
CREATE

real    0m0.662s
user    0m0.045s
sys    0m0.015s
su-2.05$ time pg_restore -cvt jurloldmail -d urldb urldb.out.020810
pg_restore: connecting to database for restore
pg_restore: dropping TABLE jurloldmail
pg_restore: connecting to database urldb as user jliang
pg_restore: creating TABLE jurloldmail
pg_restore: restoring data for table jurloldmail

real    0m6.880s
user    0m0.296s
sys    0m0.028s



10 times slower!!!




Jie Liang