Re: BUG #15446: Crash on ALTER TABLE

That still happens in postgresql 11.1

сб, 27 окт. 2018 г. в 18:39, Dmitry Molotkov <aldarund(at)gmail(dot)com>:

> I tried to reproduce it with DDL but no luck. I even tried to backup with
> the state before crash and then restore but there wasn't error in such case
> too.
> I`m using windows, so i cant really follow your instruction for taking
> dump.
> I guess the easiest would be to recreate it via that python .
>
> сб, 27 окт. 2018 г. в 17:46, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>:
>
>> On 2018-Oct-27, Dmitry Molotkov wrote:
>>
>> > I just checked with that statement and pgadmin and it still crash the
>> db.
>> > Its not related to my pc, since i initially encountered it on CI with
>> > totally different os and setup, and then reproduced it locally.
>> > https://i.imgur.com/qn1cyOz.png here what happen if i execute that
>> alter
>> > query in pgadmin
>>
>> It probably depends on other things in the table -- maybe indexes, or
>> foreign keys, or something else. If you can reproduce in psql or
>> pgadmin starting from an empty database and some DDL, please submit
>> that.
>>
>> If not, can you attach a debugger to the process before it crashes, and
>> get a core file? There are some instructions here:
>>
>> https://wiki.postgresql.org/wiki/Getting_a_stack_trace_of_a_running_PostgreSQL_backend_on_Linux/BSD
>>
>> Thanks
>>
>> --
>> Álvaro Herrera https://www.2ndQuadrant.com/
>> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
>>
>