From: Gary Stainburn <gary.stainburn@ringways.co.uk>
Organization: Ringways Garages Ltd
To: pgsql-admin@lists.postgresql.org
Subject: Re: upgrade from FC9 / PG8.3 to C7 / PG 9.6 - utf8
Date: Fri, 18 Jan 2019 16:42:45 +0000
User-Agent: KMail/1.9.10
References: <201901181429.51278.gary.stainburn@ringways.co.uk>
 <201901181530.42190.gary.stainburn@ringways.co.uk>
 <6ea28c2f-9916-1f74-9262-9448e12c62a9@gmail.com>
In-Reply-To: <6ea28c2f-9916-1f74-9262-9448e12c62a9@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <201901181642.45463.gary.stainburn@ringways.co.uk>
Precedence: bulk

On Friday 18 January 2019 16:23:25 Ron wrote:
> On 1/18/19 9:30 AM, Gary Stainburn wrote:
> [snip]
>
> > I have done the import using the pg_dumpall done on the 8.3 system. From
> > the comments earlier, would I be better off doing a pg_dumpall of the 8.3
> > system using the binarys from the 10.6 system?
>
> How much data do you have, and how large of a window do you have to migrate
> the data?
>
> For example, we moved from 8.4 to 9.6 (that was frozen while 10 was still
> in beta) and it would have taken almost five days to dump the data using
> the single-threaded 8.4 pg_dump, and just as long to do a single-threaded
> restore.
>
> OTOH, the elapsed time using the 9.6 pg_dump with 8 threads was 14 hours,
> and the restore only 8.33 hours.
My data is nowhere near that scale. Mine took less than 40 minutes to run 
pg_dumpall even using thr 8.3 pg_dumpall.

I have just had a look at the output from the 10.6 import and I am seeing a 
large number of errors like:

psql:2019-01-18_1556.sql:8918192: ERROR:  could not access 
file "$libdir/pgcrypto": No such file or directory
psql:2019-01-18_1556.sql:8918195: ERROR:  function 
public.pgp_sym_encrypt(text, text, text) does not exist


Anyone know why I'm getting this and what I need to do to fix it? I wasn't 
getting this (AFAIK) when importing to 9.6