| From: | PG Bug reporting form <noreply(at)postgresql(dot)org> |
|---|---|
| To: | pgsql-bugs(at)lists(dot)postgresql(dot)org |
| Cc: | keiko713(at)gmail(dot)com |
| Subject: | BUG #15112: Unable to run pg_upgrade with earthdistance extension |
| Date: | 2018-03-14 23:12:26 |
| Message-ID: | 152106914669.1223.5104148605998271987@wrigleys.postgresql.org |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-bugs pgsql-hackers |
The following bug has been logged on the website:
Bug reference: 15112
Logged by: Keiko Oda
Email address: keiko713(at)gmail(dot)com
PostgreSQL version: 10.3
Operating system: Ubuntu
Description:
This is similar to the bug report from 2016
(https://www.postgresql.org/message-id/20161015001424.1413.93990@wrigleys.postgresql.org)
where earthdistance extension is assuming that it's executed with public
schema in search_path.
With 10.3 release, Postgres has tighten the search_path during pg_upgrade
only to pg_catalog. This is problematic as it's now impossible to run
pg_upgrade with earthdistance extension (it fails with create index with
ll_to_earth function).
Prior to this, we were at least able to workaround by adding public schema
in search_path with pg_upgrade.
As it's recommended in the release note, earthdistance should adjust these
functions to not assume anything about what search path they are invoked
under.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Mark Magnus | 2018-03-15 04:15:03 | Postgres 10.3 delete on partition table cannot execute, out of memory. |
| Previous Message | Tom Lane | 2018-03-14 21:53:37 | Re: BUG #15111: c between x and x gives bad planning |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Mark Wong | 2018-03-14 23:14:38 | Re: Re: Re: [GSOC 18] Performance Farm Project——Initialization Project |
| Previous Message | Dave Page | 2018-03-14 21:33:00 | Re: Re: Re: [GSOC 18] Performance Farm Project——Initialization Project |