Re: BUG #9616: Materialized view with indexes unable to load from pg_dump
| От | Tom Lane |
|---|---|
| Тема | Re: BUG #9616: Materialized view with indexes unable to load from pg_dump |
| Дата | |
| Msg-id | 9030.1395154727@sss.pgh.pa.us обсуждение |
| Ответ на | BUG #9616: Materialized view with indexes unable to load from pg_dump (jesse.denardo@myfarms.com) |
| Ответы |
Re: BUG #9616: Materialized view with indexes unable to load
from pg_dump
|
| Список | pgsql-bugs |
jesse.denardo@myfarms.com writes:
> I have created several materialized views on one database. To copy data to a
> new database, I dump the schema(s) of choice using pg_dump and pipe the
> result to psql. The exact command looks like:
> $ pg_dump -b -n <schema> -O -x -h localhost -U <user> <db> | psql -h <other
> host> -d <other db> -U <other user>
> The materialized views get copied and restored just fine with one exception:
> The one mat view that has indexes. Here is the result of \d on this mat
> view:
I can't reproduce this here; it seems clear that there's some triggering
condition other than the mere presence of indexes. Could you submit a
*self contained* example (preferably, a SQL script that makes a database
that pg_dump misbehaves on)?
The output looks a little bit like pg_dump might be trying to break some
circular dependency involving the matview. But it's not obvious how
you'd have gotten into such a situation, and we certainly haven't been
shown all the moving parts.
regards, tom lane
В списке pgsql-bugs по дате отправления: