On Sat, Sep 14, 2013 at 6:54 PM, Craig James <cjames@xxxxxxxxxxxxxx> wrote:
=# select * from pg_stat_activity where pid <> pg_backend_pid();On Sat, Sep 14, 2013 at 11:36 AM, bricklen <bricklen@xxxxxxxxx> wrote:
On Sat, Sep 14, 2013 at 11:28 AM, Craig James <cjames@xxxxxxxxxxxxxx> wrote:
I'm trying to do a pg_dump of a database, and it more-or-less just sits there doing nothing.
What is running in the db? Perhaps there is something blocking the pg_dump? What does the output of the following query look like?select * from pg_stat_activity where pid <> pg_backend_pid()
datid | datname | pid | usesysid | usename | application_name | client_addr | client_hostname | client_port | backend_start
| xact_start | query_start | state_change | waiting | state |
query
--------+------------+-------+----------+----------+------------------+-------------+-----------------+-------------+------------------------------
-+-------------------------------+------------------------------+-------------------------------+---------+--------+-------------------------------
-------------------------------------------------------------------------
231308 | emolecules | 13312 | 10 | postgres | pg_dump | | | -1 | 2013-09-14 18:37:08.752938-07
| 2013-09-14 18:37:08.783782-07 | 2013-09-14 18:39:43.74618-07 | 2013-09-14 18:39:43.746181-07 | f | active | COPY orders.chmoogle_thesaurus
(thesaurus_id, version_id, normalized, identifier, typecode) TO stdout;
I don't have any solutions at the moment, but three things come to mind:
1). Try without "--blobs",
2). Does "strace -p <pid of pg_dump process>" show anything unusual? Futex? Lots of semops?
3). Does "pg_dump -f your_file.out -U postgres -Fc emolecules" act any different than redirecting STDOUT to a file?