Re: Question about pg_upgrade from 9.2 to X.X

Поиск
Список
Период
Сортировка
От Perumal Raj
Тема Re: Question about pg_upgrade from 9.2 to X.X
Дата
Msg-id CALvqh4qSarv=4Rw7SFx8D3mNeR50ta6wNb+P13G2w6+Oz_EJ-g@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Question about pg_upgrade from 9.2 to X.X  (Sergei Kornilov <sk@zsrv.org>)
Ответы Re: Question about pg_upgrade from 9.2 to X.X  (Sergei Kornilov <sk@zsrv.org>)
Список pgsql-hackers
Thank you very much Sergei,

Yes, i want to get rid of old extension, Could you please share the query to find extension which is using pg_reorg.

Regards,




On Thu, Feb 28, 2019 at 10:27 AM Sergei Kornilov <sk@zsrv.org> wrote:
Hello

pgsql-hackers seems wrong list for such question.

> could not load library "$libdir/hstore": ERROR:  could not access file "$libdir/hstore": No such file or directory
> could not load library "$libdir/adminpack": ERROR:  could not access file "$libdir/adminpack": No such file or directory
> could not load library "$libdir/uuid-ossp": ERROR:  could not access file "$libdir/uuid-ossp": No such file or directory
>
> Observation : the above Libraries are present in 9.2 whereas its mising in 10.7. So i decided to go with lower version.

This is contrib modules. They can be shipped in separate package, postgresql10-contrib.x86_64 for example (in centos repo)

> Second  i tried to attempt to upgrade from 9.2.24 to 9.6.12,9.4,9.3 but its failed with following error during Check Mode.
>
> could not load library "$libdir/pg_reorg":
> ERROR:  could not access file "$libdir/pg_reorg": No such file or directory
>
> Observation : In this case , pg_reorg is not present on both Source and Target . But strange its failing.

This is 3rd-party extension. Best way would be drop this extension on old cluster and perform upgrade. pg_reorg is abandoned for years, pg_repack is live fork if you need such tool.

regards, Sergei

В списке pgsql-hackers по дате отправления:

Предыдущее
От: Alvaro Herrera
Дата:
Сообщение: Re: partitioned tables referenced by FKs
Следующее
От: Tom Lane
Дата:
Сообщение: Re: Bloom index cost model seems to be wrong